Clinical Decision-Making Using Artificial Intelligence

Clinical Decision-Making Using Artificial Intelligence Authors: Mehrdad Farrokhi ERIS Research Institute Saba Mehrtabar Tabriz University of Medical Sciences Khadijeh Harati ERIS Research Institute Tannaz Pourlak Tabriz University of Medical Sciences Erfan Ghadirzadeh Mazandaran University of Medical Sciences Horrieh Abbasmofrad Rahman Institute of Higher Education Reza Zahedpasha Golestan University of Medical Sciences Peyman Bashghareh Golestan University of Medical Sciences Kiana Bahmanipour Shiraz University of Medical Sciences Melika Hemmati ShahroudUniversity of Medical Sciences Sepideh Amin Afshari Islamic Azad University Hamedan Branch Marjan Lashgari Hamadan University of Medical Sciences CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 1 Masih Kavian Islamic Azad University, Tehran Dental Branch Zahra Tajik Islamic Azad University, Shahr-e-Rey Branch Amirali Mohammadi Tehran University of Medical Sciences Mohammad Mehdi Karimi Kenari BabolUniversity of Medical Sciences Hamid Askari Babol University of Medical Sciences Ali Amiri Guilan University of Medical Sciences Artin Rahimi Kurdistan University of Medical Sciences Siavash Ketabi Tehran University of Medical Sciences Kamyab Komaee Koma Southeast University Kiana Nouri Tehran Medical Sciences-Islamic Azad University Reyhaneh Mehrvar Shahid Beheshti University of Medical Sciences Naeimeh Hosseini Helmholtz Munich Javaneh Atighi Qazvin University of Medical Sciences DR. MEHRDAD FARROKHI 2 Maryam Haghani University of Tabriz Zahra Naseh Emory University Sheida Akhlaghitehrani University of Rome Tor Vergata Zohreh Kourehpaz Hassanalizad Islamic Azad University, Tabriz Roozbeh Roohinezhad Iran University of Medical Sciences Somayeh Hashemi Ali Abadi Tehran University of Medical Sciences Seyed Amirali Zakavi Ardabil University of Medical Sciences Mahdi Javadian University of Campania Luigi Vanvitelli Mohammad Ali Daliri Ojghaz University of Naples Federico II Mohammad - Alinezhad Taheri University of Naples Federico II Zahra Hamzehnejadi University of Naples Federico II Eros Cribello University of Naples Federico II Sayed Mohammadamin Tabatabaei University of Naples Federico II CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 3 Masoud Seifi University of Naples Federico II Niloofar Taheri Shahroud University of Medical Sciences Omid Fakharzadeh Moghadam Mashhad University of Medical Sciences Sanaz Amiri Marbini University Medical Center Hamburg-Eppendorf Saman Abdollahpour Shahid Beheshti University of Medical Sciences Kameliya Sanjabiyan Jundishapur University of Medical Sciences DR. MEHRDAD FARROKHI 4 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 5 Book Details: Publisher: PreferPub and Kindle Publication Date: December 2025 Language: English Dimensions: 5 x 0.39 x 8 inches © PreferPub and Kindle 2025 ISBN-13: 979-8279125142 This peer-reviewed book is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specially the rights of translation, reprinting, result of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. DR. MEHRDAD FARROKHI 6 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 7 Contents Chapter 1AI in Clinical Decision-Making in Internal Medicine 2AI in Clinical Decision-Making in Neurology and Neurosurgery 3AI in Clinical Decision-Making in Oncology and Hematology 4AI in Clinical Decision-Making in Dentistry and Oral Medicine 5AI in Clinical Decision-Making in Rehabilitation Medicine 6AI in Clinical Decision-Making in Surgery and Emergency Medicine 7AI in Clinical Decision-Making in Pediatrics and Obstetrics-Gynecology 8AI in Clinical Decision-Making in Psychiatry and Behavioral Sciences 9AI in Clinical Decision-Making in Radiology and Pathology 10AI in Clinical Decision-Making in Other Medical Specialties DR. MEHRDAD FARROKHI 8 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 9 DR. MEHRDAD FARROKHI 10 1AI IN CLINICAL DECISIONMAKING IN INTERNAL MEDICINE Background Internal medicine is a medical specialty that may have the highest number of patients among all hospital disciplines. It has been reported that admissions of patients under internal medicine specialty care account for approximately 17.4 percent of all hospital admissions. The increasing prevalence of multimorbidity and chronic conditions, in addition to population aging and unhealthy lifestyles, further elevates the importance of this specialty in the future. The holistic diagnostic nature of internal medicine, the presence of fragmented and modular health care systems, and the lack of inclusive clinical guidelines that adequately address increasing multimorbidity, together with the exponential growth of medical knowledge, necessitate a closer examination of current aspects of patient care in this field. In internal medicine, misdiagnosis rates can reach up to approximately 11.1 percent. With the availability of improved health care infrastructure, such as electronic health records, diverse imaging datasets, and longer-term patient follow-up data, artificial intelligence can offer potential alternatives to address these challenges. 11 Its utility in the early detection of diseases, including arrhythmic events, sepsis, lung injury, and in the advancement of personalized medicine, has been enabled through data analysis and pattern recognition applied to large-scale health care data. The development of clinical decision support tools began in the early 1970s with the creation of systems such as MYCIN, CASNET, and INTERNIST-I. However, limited computational resources, increasing data complexity, and the high level of maintenance required posed significant barriers to their integration into health care and routine clinical practice. The high maintenance demands were largely related to the difficulty of incorporating continuously evolving medical knowledge into models that relied primarily on decision trees and predefined rule-based structures. Nevertheless, these early systems established a foundational framework that informed and guided future developments in the field. Contemporary artificial intelligence algorithms, including neural network techniques that use interconnected neural nodes across multiple layered networks, enable the analysis of large and complex datasets, such as imaging data, without the need for explicit instructions regarding data structure. Current AI applications, such as DeepDR, which employs deep learning methods, have demonstrated the ability to DR. MEHRDAD FARROKHI 12 predict diabetic retinopathy several years before the appearance of overt clinical signs. Similarly, studies describing models such as FJCovNet have reported the use of deep learning approaches for the early detection of COVID-19, highlighting the potential of artificial intelligence to reduce the burden on health care systems during pandemics. Although a growing number of studies report the use of artificial intelligence algorithms across various areas of health care, several important concerns remain. These include the black box nature of many deep learning models, as well as their tendency to underperform in underrepresented patient populations, such as female patients or individuals belonging to certain ethnic groups. A comprehensive understanding of artificial intelligence and current advancements in this field will be valuable in supporting clinicians and regulatory bodies as they seek to address existing challenges and to develop concepts and techniques aimed at mitigating these limitations within the practice of internal medicine. Cardiology is a unique specialty within medicine, integrating a wide spectrum of biological data, including imaging, time-series signals such as pulse waves and electrocardiograms, and sound information obtained through cardiac auscultation. The emerging application of machine learning to these complex datasets offers considerable potential to enhance cardiovascular CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 13 research and to support clinical decision-making. The expanding scope of cardiovascular big data provides significant opportunities for advanced analysis through artificial intelligence and machine learning techniques. For clinicians, these technologies offer the potential for more precise and standardized image interpretation, enhanced diagnostic accuracy, improved risk prediction, and tailored treatment guidance throughout the disease course. At the health system level, artificial intelligence can contribute to improved efficiency, reduced medical errors, and better patient outcomes. For patients who are increasingly engaged in monitoring their own health data, artificial intelligence creates opportunities for education, empowerment, and the promotion of both primary and secondary cardiovascular prevention. When thoroughly validated and implemented in routine clinical practice, algorithms have the potential to ease clinicians’ cognitive burden by providing preliminary diagnostic support, minimizing diagnostic errors, and helping to prevent misdiagnosis. However, despite rapid technological progress, most current applications remain at the proof-of-concept stage, with limited translation into real-world cardiology practice. Ongoing concerns related to generalizability, interpretability, and ethical use continue to create a gap between technological innovation DR. MEHRDAD FARROKHI 14 and widespread clinical adoption. This review synthesizes current evidence on the application of artificial intelligence in cardiovascular clinical decision-making. Although many studies have explored artificial intelligence in cardiology, relatively few have examined its integration across diagnostic, prognostic, and therapeutic domains in clinical settings. By addressing this gap, the review outlines areas where artificial intelligence has demonstrated clinical value, highlights existing limitations, and identifies priorities for the safe and effective adoption of artificial intelligence in cardiology practice. AI for Diagnostic Decision-Making Diagnosing complex cardiovascular conditions such as heart failure involves interpretation of high-dimensional data that may exceed human cognitive capacity, making artificial intelligence particularly valuable. The ability of artificial intelligence to detect and integrate subtle patterns across multiple data sources supports more efficient and accurate clinical decisionmaking. This capability is evident across a range of diagnostic modalities, from advanced cardiac imaging techniques to commonly used laboratory investigations. In cardiac imaging, deep learning models trained using large echocardiographic datasets have demonstrated performance comparable to CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 15 conventional volume-driven techniques used by experts. These findings indicate that algorithmderived assessments of left ventricular systolic function, including left ventricular ejection fraction, can achieve reliability similar to expert interpretation. Such results suggest that artificial intelligence can automate labor-intensive measurements while maintaining diagnostic accuracy. Beyond imaging, artificial intelligence is also transforming the interpretation of biochemical markers. Machine learning models that integrate cardiac troponin concentrations measured at presentation or through serial testing with relevant clinical features can generate individualized risk scores for myocardial infarction. These scores function as clinical decision-support tools that move beyond rigid threshold values, enabling more refined and personalized risk stratification for patients presenting with suspected acute coronary syndromes. Electrocardiographic signal analysis represents one of the most active areas of artificial intelligence development in cardiology. Recent advancements in machine learning algorithms, signal noise reduction, feature extraction methods, and optimization strategies have significantly enhanced diagnostic accuracy for arrhythmia detection, with some models achieving accuracy rates approaching 95 DR. MEHRDAD FARROKHI 16 percent. With the widespread digitalization of electrocardiograms, artificial intelligence-driven analytical methods have become increasingly common, particularly for predicting cardiac arrhythmias. Highly effective machine learning approaches have been developed to classify electrocardiograms and forecast the onset of paroxysmal atrial fibrillation with very high sensitivity and specificity. In addition, the clinical utility of artificial intelligenceenhanced electrocardiography extends beyond rhythm interpretation. Deep learning algorithms have demonstrated strong performance in distinguishing between hypertensive heart disease, hypertrophic cardiomyopathy, and cardiac amyloidosis, which represent common causes of left ventricular hypertrophy. In comparative analyses, these algorithms have outperformed expert echocardiographic interpretation, highlighting the capacity of artificial intelligence to derive detailed and subtype-specific insights from a widely available and relatively inexpensive diagnostic test. The diagnostic applications of artificial intelligence also extend to physical examination and population screening. Artificial intelligenceenabled digital stethoscopes and computerassisted auscultation systems offer promising opportunities for early detection of heart murmurs and rheumatic heart disease, conditions CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 17 for developing equitable artificial intelligence systems. Another challenge involves limited interpretability of complex models such as deep neural networks, which may reduce clinician trust. The development of explainable artificial intelligence methods is critical to making algorithmic decisions transparent and clinically meaningful. Regulatory and validation barriers also persist. Although many artificial intelligencebased cardiology tools have received regulatory approval, there remains a need for prospective randomized studies to demonstrate consistent improvements in real-world patient outcomes. Legal and ethical questions surrounding accountability for adverse outcomes associated with artificial intelligence-assisted decisions require careful consideration, involving clinicians, healthcare institutions, and technology developers. Data privacy and security represent additional concerns, as the use of sensitive patient information for training and deployment of artificial intelligence systems demands strict adherence to data protection standards in order to maintain patient trust. Conclusion In summary, artificial intelligence demonstrates substantial transformative potential across cardiology by enhancing diagnostic accuracy, DR. MEHRDAD FARROKHI 30 improving prognostic assessment, and supporting individualized therapeutic strategies. However, successful integration into routine clinical practice depends on addressing critical challenges, including the need for rigorous validation, careful management of ethical and liability concerns, and reinforcement of artificial intelligence as a complement to, rather than a replacement for, clinical judgment. The future of cardiology lies in a collaborative partnership between human expertise and artificial intelligence-driven insights, with the shared goal of improving patient care and clinical outcomes. Accordingly, sustained collaboration among data scientists, clinicians, and regulatory authorities will be essential to translate the promise of artificial intelligence into meaningful clinical reality. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 31 DR. MEHRDAD FARROKHI 32 2AI IN CLINICAL DECISIONMAKING IN NEUROLOGY AND NEUROSURGERY Background Neurology and neurosurgery are among the most data-intensive medical specialties that require the rapid integration of multimodal information, such as neuroimaging, electrophysiology, genomics, and electronic health records, including radiology, operative, and consultation notes. Clinical decisions often rely on accurate and timely interpretation of these data, a process that frequently exceeds human cognitive capacity when performed under significant time constraints. Artificial intelligence has emerged as a transformative force in these domains and comprises machine learning, deep learning, natural language processing, and reinforcement learning. Recent systematic reviews and meta-analyses demonstrate that artificial intelligence is transitioning from proof-of-concept development to regulated clinical deployment, with FDAcleared platforms used in acute stroke triage and seizure detection, CE-marked solutions applied in neuroimaging, and artificial intelligence enhanced robotic guidance increasingly utilized in 33 neurosurgical practice. At the same time, major challenges remain related to methodological rigor, generalizability, interpretability, and ongoing regulatory monitoring, which continue to limit widespread and uniform adoption across clinical settings. This chapter summarizes the current state of artificial intelligence in clinical decision making in neurology and neurosurgery, evaluates key methodological strengths and existing drawbacks, and outlines translational, regulatory, and ethical imperatives that are essential for successful implementation in routine clinical practice. Methodological Foundations The foundation of artificial intelligence in neuroscience comprises several core methodological pillars. a. Supervised machine learning includes classification and regression models developed for tasks such as predicting stroke outcomes or estimating complications following spine surgery. b. Unsupervised machine learning focuses on clustering and representation learning techniques that aim to discover phenotypes in neurodegenerative disorders and to stratify heterogeneous conditions, such as multiple sclerosis. c. Deep learning approaches include convolutional neural networks and transformerDR. MEHRDAD FARROKHI 34 based vision models applied to neuroimaging and electroencephalography or time-series analysis, achieving performance comparable to fellowship-trained neuroradiologists and clinical neurophysiologists for specific and well-defined tasks. d. Natural language processing includes large language models combined with retrievalaugmented generation to structure, summarize, and analyze unstructured clinical notes, operative reports, and electronic health records, thereby enabling automated risk prediction, clinical summarization, and patient triage. e. Reinforcement learning includes offline or batch reinforcement learning with safe policy evaluation for adaptive therapy optimization, such as closed-loop deep brain stimulation parameter tuning under predefined safety constraints. f. Federated learning enables multi-institutional model training without centralizing raw data, thereby mitigating compliance risks under the European Union General Data Protection Regulation and the United States Health Insurance Portability and Accountability Act, while also improving model generalizability across institutions and populations. Extensions of this approach include secure aggregation and differential privacy techniques. g. Validation, calibration, and clinical utility CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 35 emphasize the importance of external validation, appropriate calibration methods such as Brier scores, calibration slopes, and scaling techniques, as well as decision-curve analysis to quantify net clinical benefit and clinical usefulness beyond traditional discrimination metrics. h. Uncertainty and reliability approaches use conformal prediction, deep ensembles, or Monte Carlo dropout techniques to quantify case-level predictive uncertainty and to support abstention strategies, triage policies, and human-in-the-loop clinical review. i. Robustness and distribution shift address variability related to clinical sites, imaging scanners, and patient populations through domain adaptation, data harmonization methods such as ComBat, targeted data augmentation, out-of-distribution detection, and test-time adaptation. j. Causal inference for decision support distinguishes associational relationships from causal effects using directed acyclic graphs, target trial emulation, and heterogeneous treatment effect models, such as causal forests or uplift modeling, to support individualized clinical decisions. k. Self-supervised and foundation models leverage contrastive learning and masked autoencoder pretraining for magnetic resonance imaging and electroencephalography, parameter-efficient DR. MEHRDAD FARROKHI 36 tuning strategies, and electronic health record pipelines powered by large language models with retrieval-augmented generation for retrievalgrounded reasoning. l. Explainability and model auditing apply methods such as Shapley values, integrated gradients, and gradient-weighted class activation mapping, incorporate counterfactual explanations, and promote the publication of model cards and dataset datasheets that specify intended use, limitations, and monitoring plans. m. Deployment and governance align with established software lifecycle and risk management standards, define change-control processes for machine learning software used as a medical device, monitor data and model drift, ensure appropriate model versioning, and maintain human-in-the-loop safeguards supported by structured post-market surveillance. Multimodal fusion strategies, including early, late, and intermediate fusion as well as cross-attention transformer approaches, integrate imaging data, physiologic time series, clinical variables, and narrative text. In addition, graph neural networks are used to model structural and functional connectivity as well as vascular graph representations. Applications in Neurology CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 37 Stroke and Acute Neurology Artificial intelligence has demonstrated the greatest level of maturity in the context of acute stroke care. Deep learning platforms such as Viz.ai and RapidAI automatically detect large vessel occlusions, with reported sensitivities exceeding 90 percent, and shorten time to treatment, such as door to needle or door to groin puncture intervals, by approximately 20 minutes. Regulatory clearances now extend to automated quantification of intracerebral hemorrhage and subdural hematoma. Despite measurable workflow improvements, prospective randomized evidence demonstrating consistent improvements in patient outcomes remains limited. Epilepsy and Seizure Detection Artificial intelligence has substantially accelerated electroencephalography interpretation. Convolutional neural networks achieve seizure prediction sensitivities ranging from approximately 75 to 85 percent when applied to intracranial electroencephalography. Commercial rapid electroencephalography platforms incorporate artificial intelligence analytics for near real time detection of seizures and status epilepticus at the bedside, thereby broadening access in emergency department and intensive care unit settings. DR. MEHRDAD FARROKHI 38 Multimodal fusion of magnetic resonance imaging, positron emission tomography, and electroencephalography improves epileptogenic zone localization and supports surgical decision making in refractory epilepsy. Movement Disorders Wearable sensor data streams analyzed using machine learning algorithms classify Parkinsonian motor states, including bradykinesia, dyskinesia, and freezing, with accuracy exceeding 85 percent and enable continuous remote monitoring through smartphone based systems. In addition, artificial intelligence can predict response to levodopa therapy and supports deep brain stimulation parameter optimization, including offline or batch reinforcement learning guided search strategies and connectivity informed targeting approaches. Neurodegenerative Disorders Deep learning applied to structural magnetic resonance imaging can detect Alzheimer related patterns of cerebral atrophy several years before clinical onset. Prediction of progression from mild cognitive impairment to Alzheimer disease achieves area under the curve values of approximately 0.85 to 0.90 across multiple models. Speech and language based natural language processing markers provide complementary early detection signals. Artificial CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 39 intelligence is also increasingly embedded in population level prevention and triage frameworks to prioritize cohorts at elevated risk. Emergency Neurology An ensemble artificial intelligence framework combining large language models, retrieval augmented generation, and machine learning classifiers has been shown to predict hospital admission and short term mortality across more than one thousand emergency neurology consultations. Model outputs closely aligned with expert clinical judgment and improved triage accuracy. These findings illustrate the role of artificial intelligence not only in single task diagnostic applications but also in systems level clinical decision support. Multiple Sclerosis and Neuroimmunology Artificial intelligence supports lesion detection and segmentation, quantification of new or enlarging T2 and gadolinium enhancing lesions, measurement of brain and spinal cord atrophy, and modeling of relapse risk and treatment response. These applications are often implemented within multimodal analytical pipelines that integrate imaging data with serum and cerebrospinal fluid biomarkers. Neuromuscular Disorders DR. MEHRDAD FARROKHI 40 and Muscle Ultrasound Automated classifiers applied to electromyography and nerve conduction study waveforms support differentiation between demyelinating and axonal neuropathies and can identify mixed patterns that warrant specialist review. Muscle ultrasound techniques quantify echogenicity, fasciculations, and architectural changes to phenotype myopathies and motor neuron disease. Beyond static imaging, dynamic cine magnetic resonance imaging of respiratory and bulbar musculature enables frame by frame assessment of contractile mechanics, such as diaphragmatic excursion and muscle thickening, thereby providing objective markers of neuromuscular respiratory weakness in conditions including Duchenne muscular dystrophy and Pompe disease. In clinical practice, multimodal fusion of electrophysiology, ultrasound, and cine magnetic resonance imaging can improve diagnostic confidence, guide therapeutic decisions such as ventilatory support or immunotherapy, and provide quantitative endpoints for longitudinal follow up. Sleep Medicine and Chronobiology Artificial intelligence enables automated scoring of polysomnography, including apnea hypopnea events, arousals, and sleep staging. It also supports detection of rapid eye movement behavior disorder as a prodromal marker CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 41 of synucleinopathies and facilitates validation of home based wearable devices for sleep monitoring. Cognitive and Language Biomarkers Speech and language based natural language processing features, tablet based cognitive tasks, and multimodal fusion with magnetic resonance imaging and positron emission tomography provide early detection and prognostic enrichment for Alzheimer disease and frontotemporal dementias. These approaches enhance sensitivity to subtle cognitive and linguistic changes that may precede overt clinical symptoms. Pediatric Neurology Artificial intelligence models in pediatric neurology target neonatal seizure detection using amplitude integrated electroencephalography and conventional electroencephalography, support rapid magnetic resonance imaging triage, and assist in classification of metabolic and epileptic encephalopathies. Particular emphasis is placed on addressing dataset shift, bias, and ethical safeguards when deploying these tools in pediatric populations. Neuro-ophthalmology and Optical Coherence Tomography Artificial intelligence applied to optical coherence DR. MEHRDAD FARROKHI 42 tomography and visual field testing detects optic neuritis, tracks disease progression in multiple sclerosis, and maps structure function relationships across visual pathways. These applications contribute to earlier diagnosis, improved monitoring, and a deeper understanding of visual system involvement in neurological disease. Headache and Autonomic Disorders Digital phenotyping using wearable devices and smartphone-based platforms supports migraine subtype prediction, trigger modeling, and attack forecasting. Current evidence in this area is still emerging, but available findings are promising and suggest that continuous passive data collection may enhance individualized headache management and improve understanding of autonomic dysfunction patterns over time. Neurocritical Care Non Stroke In neurocritical care settings outside of stroke, multimodal intensive care unit monitoring that includes electroencephalography, intracranial pressure measurements, brain tissue oxygenation, and detailed hemodynamic parameters feeds predictive models designed to detect secondary brain injury, generate early deterioration alerts, and guide sedation titration. These approaches aim to support timely interventions and improve physiologic stability in critically ill neurological CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 43 patients. Tele Neurology and Workflow Artificial Intelligence Artificial intelligence applications in tele neurology focus on triage optimization, referral prioritization, and capacity planning across health care systems. These tools have the potential to improve access to neurological care and enhance clinical workflow efficiency, provided that robust governance structures and strict data quality controls are implemented to ensure safety, equity, and reliability. Neurogenetics and Variant Interpretation Machine learning techniques assist in variant pathogenicity scoring and phenotype to genotype matching in neurodevelopmental and neurodegenerative syndromes. By integrating genomic data with detailed clinical phenotypes, these approaches support diagnostic interpretation, reduce uncertainty in rare or complex cases, and facilitate earlier and more precise genetic counseling and clinical decision making. Brain Computer Interfaces and Assistive Neurotechnology Advances in brain computer interfaces and assistive neurotechnology focus on decoding DR. MEHRDAD FARROKHI 44 intended speech and motor commands for patients with amyotrophic lateral sclerosis or locked in syndrome. Closed loop adaptive paradigms are progressing toward clinical readiness under strict ethical safeguards, with the goal of restoring communication and functional independence while ensuring patient safety and autonomy. Rare Disease Pattern Recognition Magnetic resonance imaging pattern classifiers assist in the differential diagnosis of leukodystrophies and metabolic or mitochondrial disorders by accelerating expert level pattern recognition. These tools help reduce diagnostic delay, particularly in rare diseases, and support earlier referral to specialized care and targeted therapeutic interventions. Applications in Neurosurgery Preoperative Planning Artificial intelligence systems integrate preoperative magnetic resonance imaging and computed tomography with anatomical atlases and tractography to support risk aware trajectory optimization for biopsy procedures, deep brain targets, and minimally invasive evacuation strategies. In intracerebral hemorrhage, end to end pipelines that perform automated hematoma segmentation and trajectory planning illustrate CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 45 how algorithmic planning can standardize case preparation and reduce planning times, with prospective evaluation currently underway. In neuro oncology, deep learning tools generate robust preoperative tumor segmentations that standardize volumetric assessment and facilitate surgical resection strategies. In functional neurosurgery, connectomic targeting approaches link subthalamic stimulation sites to normative connectivity fingerprints that predict clinical outcomes across independent cohorts, providing a quantitative complement to traditional coordinate based planning. While gains in planning accuracy and efficiency are consistently reported, prospective trials demonstrating reductions in surgical complications remain limited. Intraoperative Guidance Two major streams of intraoperative guidance have matured. First, label free optical histology using stimulated Raman histology combined with convolutional neural networks enables near real time intraoperative diagnosis at the point of care, typically within approximately one hundred and fifty seconds. These methods have been extended to rapid molecular screening, such as mutation status and chromosomal alterations, to inform intraoperative surgical decision making. Deep learning also supports real time segmentation of tumor margins and vascular or white matter DR. MEHRDAD FARROKHI 46 structures using magnetic resonance imaging and intraoperative ultrasound, with operating room deployable inference increasingly described. Second, computer vision models applied to intraoperative endoscopic and microscopic video provide anatomical and procedural phase recognition, supporting navigation, exposure strategies, and surgical team coordination. Phase recognition systems for endoscopic endonasal pituitary surgery are being developed to support training, operative coaching, and automated time and motion analytics. Robotic systems integrated with artificial intelligence based planning are also being evaluated to reduce stereotactic violations and the need for repositioning, although large prospective trials with definitive outcome endpoints remain necessary. Postoperative Prognostication Postoperative risk prediction models increasingly combine perioperative variables with early postoperative imaging findings, laboratory trends, and physiologic signals to predict complications such as pulmonary events and infections, as well as functional recovery trajectories. These models aim to enable targeted surveillance and early intervention. In spine surgery, machine learning based risk calculators are being incorporated into shared decision making frameworks to estimate patient reported outcome measures and CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 47 complication risk, thereby supporting expectation management and perioperative risk mitigation strategies. Neuro Oncology Artificial intelligence has been particularly transformative in neuro oncologic imaging and intraoperative diagnosis. Open benchmark initiatives for brain tumor segmentation and radiogenomic prediction have catalyzed reproducible approaches to preoperative volumetry and target delineation. Parallel clinical studies demonstrate that stimulated Raman histology and deep learning pipelines can deliver near real time histologic diagnosis and molecular screening from fresh tissue, effectively linking imaging, pathology, and intraoperative decision making. Contemporary systematic reviews in diagnostic neurosurgery report high diagnostic performance for tumor detection, grading, and molecular characterization, while also emphasizing heterogeneous imaging protocols and limited prospective validation as key barriers to routine integration into surgical practice. Cerebrovascular Neurosurgery Across computed tomography angiography, magnetic resonance angiography, and digital subtraction angiography, artificial intelligence models are now capable of detecting and segmenting intracranial aneurysms and DR. MEHRDAD FARROKHI 48 quantifying aneurysm morphology with high levels of accuracy. These capabilities facilitate standardized measurement and support consistent longitudinal follow up in clinical practice. However, rupture risk stratification remains an emerging and incompletely validated application. Most available studies are retrospective and single center in design, rely on inconsistent annotation strategies, and demonstrate limited external validation. As a result, claims of clinical superiority over established risk assessment scores remain premature. Priority areas for future progress include multi site benchmarking, harmonized labeling strategies, and prospective clinical evaluations prior to routine clinical adoption. Functional Neurosurgery and Neuromodulation Beyond preoperative planning, artificial intelligence supports postoperative programming and ongoing therapy adaptation in functional neurosurgery. Connectivity informed targeting approaches have been shown to predict deep brain stimulation response in Parkinson disease, while adaptive deep brain stimulation uses local field potential signals, such as beta band power, to modulate stimulation parameters in real time. Early studies suggest that these adaptive paradigms may improve motor outcomes while reducing energy consumption when compared CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 49 3AI IN CLINICAL DECISIONMAKING IN ONCOLOGY AND HEMATOLOGY Background Artificial intelligence has emerged as one of the most influential technological developments in modern medicine, with oncology and hematology standing among the specialties most profoundly affected. These fields are inherently data intensive, relying on complex diagnostic inputs such as imaging, histopathology, genomics, laboratory data, and longitudinal clinical records. At the same time, cancer and hematologic disorders are characterized by biological heterogeneity, rapidly evolving treatment options, and high stakes clinical decisions. This combination creates both an opportunity and a necessity for advanced computational tools that can support clinicians in navigating uncertainty and complexity. Artificial intelligence, particularly through machine learning, deep learning, natural language processing, and large language models, offers new approaches to clinical decision-making that aim to enhance diagnostic precision, refine risk stratification, personalize therapy, and improve patient outcomes while preserving the central role of clinician judgment. 62 In oncology and hematology, decision-making extends across the entire continuum of care, from early detection and diagnosis to prognostication, treatment selection, monitoring of response, and survivorship planning. Traditional clinical models and guidelines remain foundational, but they often struggle to account for the full dimensionality of modern biomedical data or to adapt rapidly to emerging evidence. Artificial intelligence systems are increasingly positioned as decision support tools that can integrate diverse data sources, identify subtle patterns beyond human perception, and generate probabilistic insights that complement established clinical frameworks. Understanding how these tools are developed, validated, and implemented is essential for their responsible and effective use in cancer and blood disorder care. Foundations of Artificial Intelligence in Oncology and Hematology Artificial intelligence refers to computational systems designed to perform tasks that typically require human intelligence, such as learning, reasoning, pattern recognition, and decision support. In clinical oncology and hematology, the most relevant forms of artificial intelligence are machine learning and deep learning. Machine learning involves algorithms that learn relationships from data rather than relying CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 63 on explicit rule-based programming. These algorithms can be supervised, unsupervised, or reinforcement based, depending on whether labeled outcomes are provided, latent patterns are discovered, or sequential decision optimization is required. Deep learning is a subset of machine learning that uses multi-layer neural networks to model complex, non-linear relationships. Deep learning has been particularly impactful in image analysis, including radiology and digital pathology, as well as in genomics and speech or text processing. Natural language processing enables artificial intelligence systems to extract structured information from unstructured clinical text, such as pathology reports, progress notes, and discharge summaries. Large language models extend this capability by generating coherent summaries, answering clinical queries, and assisting with documentation, although their use in clinical settings requires careful oversight. In oncology and hematology, these methods are rarely used in isolation. Instead, they are combined into integrated systems that process multimodal data, reflecting the reality that cancer and hematologic disorders are diagnosed and managed using converging streams of biological and clinical information. The clinical relevance of artificial intelligence depends not only on algorithmic sophistication but also on the quality DR. MEHRDAD FARROKHI 64 of input data, the clarity of the clinical question being addressed, and the rigor of validation processes. AI in Cancer Screening and Early Detection Early detection remains one of the most effective strategies for improving cancer outcomes, and artificial intelligence has shown considerable promise in enhancing screening programs. In imaging-based screening, deep learning algorithms have demonstrated high performance in detecting abnormalities in mammography, lowdose computed tomography for lung cancer, prostate magnetic resonance imaging, and colonoscopy imaging. These systems are trained on large datasets annotated by expert clinicians and can identify subtle features associated with early malignancy that may be difficult to detect consistently by human readers alone. In breast cancer screening, artificial intelligence systems have been developed to assist radiologists by prioritizing high-risk images, reducing false positives, and lowering recall rates without compromising sensitivity. Similar advances have been observed in lung cancer screening, where deep learning models analyze computed tomography scans to identify suspicious nodules and estimate malignancy risk. In colorectal cancer screening, artificial intelligence-assisted CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 65 colonoscopy systems provide real-time feedback to endoscopists, improving adenoma detection rates and reducing the likelihood of missed lesions. Beyond imaging, artificial intelligence is increasingly applied to non-invasive screening approaches, including liquid biopsies and bloodbased biomarkers. Machine learning models can analyze circulating tumor DNA, cell-free RNA, proteomic profiles, and metabolomic signatures to detect early cancer signals. These approaches hold particular promise for cancers that lack effective population screening methods and for patients at elevated genetic or environmental risk. Diagnostic DecisionMaking and Pathology Accurate diagnosis is the cornerstone of oncology and hematology, and artificial intelligence has made significant inroads in both solid tumor pathology and hematopathology. Digital pathology enables the conversion of glass slides into high-resolution whole-slide images, creating a foundation for computational analysis. Deep learning algorithms can identify tumor regions, classify histologic subtypes, grade malignancies, and detect features such as lymphovascular invasion or mitotic activity. In solid tumors, artificial intelligence systems have demonstrated performance comparable to DR. MEHRDAD FARROKHI 66 expert pathologists in tasks such as distinguishing benign from malignant lesions, identifying tumor margins, and predicting molecular alterations directly from histologic images. These capabilities are particularly valuable in settings where access to subspecialty pathology expertise is limited. In hematology, artificial intelligence is applied to peripheral blood smears, bone marrow aspirates, and flow cytometry data to assist in the classification of leukemias, lymphomas, and myelodysplastic syndromes. Artificial intelligence also plays a growing role in integrating pathology with molecular diagnostics. Machine learning models can predict genetic mutations, chromosomal rearrangements, and expression profiles from histologic or cytologic images, potentially reducing the need for additional testing or guiding targeted molecular analysis. While these approaches are not intended to replace molecular diagnostics, they offer a complementary layer of information that can streamline workflows and support diagnostic confidence. Radiology and ImagingBased Decision Support Imaging is central to cancer staging, treatment planning, and response assessment. Artificial intelligence has transformed oncologic imaging by enabling automated segmentation CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 67 of tumors and organs at risk, extraction of quantitative imaging features, and longitudinal tracking of disease burden. Radiomics, which involves the extraction of high-dimensional features from imaging data, uses machine learning to associate imaging patterns with clinical outcomes, molecular characteristics, and treatment response. In radiation oncology, artificial intelligence supports contouring of target volumes and normal tissues, reducing inter-observer variability and planning time. In medical oncology, imagingbased artificial intelligence tools assist in staging, detection of metastatic disease, and assessment of treatment response beyond conventional criteria. For example, machine learning models can detect subtle changes in tumor texture or shape that precede measurable size reduction, providing early indicators of therapeutic effectiveness. In hematologic malignancies, imaging plays a role in staging lymphomas and evaluating organ involvement. Artificial intelligence-assisted imaging analysis can improve consistency in response assessment and help differentiate residual disease from treatment-related changes. These tools are particularly valuable in clinical trials and longitudinal monitoring, where standardized assessment is essential. Prognostic Modeling and DR. MEHRDAD FARROKHI 68 Risk Stratification Prognostication is a critical component of clinical decision-making in oncology and hematology, influencing treatment intensity, surveillance strategies, and patient counseling. Traditional prognostic models are often based on a limited number of clinical and laboratory variables. Artificial intelligence enables the development of more comprehensive risk models that incorporate diverse data sources, including genomics, imaging, laboratory trends, and patient-reported outcomes. In solid tumors, machine learning models have been developed to predict overall survival, progression-free survival, and risk of recurrence following surgery or systemic therapy. These models can support shared decision-making by providing individualized risk estimates that complement guideline-based recommendations. In hematology, artificial intelligence-based prognostic tools have shown promise in diseases such as acute myeloid leukemia, multiple myeloma, and chronic lymphocytic leukemia, where outcomes depend on complex interactions between genetic, clinical, and treatment-related factors. Artificial intelligence can also assist in predicting treatment-related toxicity and complications. Models that analyze baseline characteristics and early treatment data can estimate the CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 69 risk of adverse events such as neutropenia, cardiotoxicity, or immune-related toxicity. These insights enable clinicians to tailor supportive care strategies and adjust therapy proactively. Precision Oncology and Treatment Selection Precision medicine aims to match patients with therapies most likely to be effective based on the biological characteristics of their disease. Artificial intelligence is central to this effort, as it can integrate genomic data, transcriptomics, proteomics, and clinical variables to support treatment selection. Machine learning models analyze large datasets from clinical trials, realworld evidence, and molecular profiling studies to identify patterns of response and resistance. In oncology, artificial intelligence supports decision-making in targeted therapy and immunotherapy by predicting which patients are likely to benefit from specific agents. For example, models can analyze tumor mutational burden, gene expression signatures, and microenvironment features to estimate response to immune checkpoint inhibitors. In hematology, artificial intelligence assists in selecting optimal induction regimens, consolidation strategies, and transplant eligibility, particularly in complex diseases with multiple therapeutic pathways. Clinical decision support platforms increasingly DR. MEHRDAD FARROKHI 70 incorporate artificial intelligence to generate ranked treatment options based on patientspecific data and current evidence. These systems are designed to assist multidisciplinary tumor boards by synthesizing vast amounts of information into actionable insights, while leaving final decisions to clinicians. AI in Clinical Trials and Drug Development Artificial intelligence has implications beyond individual patient care, extending into clinical research and drug development. In oncology and hematology, clinical trials are often expensive, lengthy, and limited by restrictive eligibility criteria. Artificial intelligence can improve trial design by identifying suitable patient populations, optimizing endpoints, and predicting enrollment feasibility. Machine learning models can analyze electronic health records to identify patients who meet trial criteria, facilitating recruitment and improving representation. Artificial intelligence also supports adaptive trial designs by enabling real-time analysis of emerging data and informing protocol modifications. In drug development, artificial intelligence accelerates target discovery, predicts drug toxicity, and identifies synergistic drug combinations, potentially shortening the timeline from discovery to clinical application. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 71 Ethical, Regulatory, and Practical Considerations Despite its promise, the integration of artificial intelligence into oncology and hematology raises important ethical, regulatory, and practical challenges. Data quality and bias remain central concerns, as models trained on nonrepresentative datasets may perpetuate disparities in care. Ensuring diversity in training data and conducting fairness assessments are essential to equitable deployment. Interpretability is another critical issue. Clinicians must understand the rationale behind artificial intelligence-generated recommendations to trust and appropriately use these tools. Explainable artificial intelligence approaches aim to provide insights into model behavior, highlighting relevant features and uncertainties. Regulatory frameworks increasingly require evidence of safety, effectiveness, and transparency before clinical deployment. Data privacy and security are particularly sensitive in oncology and hematology, where genomic and longitudinal health data are integral to care. Robust governance structures, secure data handling, and compliance with regulatory standards are necessary to maintain patient trust. From a practical perspective, successful implementation depends on workflow DR. MEHRDAD FARROKHI 72 integration, clinician training, and institutional support. The Future of AI in Oncology and Hematology The future of artificial intelligence in oncology and hematology lies in deeper integration, multimodal learning, and collaborative human and machine intelligence. Multimodal models that combine imaging, molecular data, clinical text, and patient-reported outcomes are expected to provide more holistic insights into disease behavior and treatment response. Large language models may further reduce administrative burden and support knowledge retrieval, provided that safeguards are in place. Digital twins, which are virtual representations of individual patients, represent an emerging concept with potential applications in treatment simulation and outcome prediction. As computational power and data availability continue to grow, artificial intelligence systems may become increasingly adaptive, learning from ongoing clinical practice while remaining subject to rigorous oversight. Conclusion Artificial intelligence is reshaping clinical decision-making in oncology and hematology by augmenting diagnostic accuracy, refining CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 73 prognostication, personalizing treatment, and supporting research innovation. Rather than replacing clinicians, artificial intelligence serves as a powerful adjunct that helps manage complexity and uncertainty in cancer and blood disorder care. Realizing its full potential requires careful attention to data quality, validation, ethics, and human-centered design. When implemented responsibly, artificial intelligence has the capacity to improve outcomes, enhance efficiency, and support more equitable and precise care for patients facing some of the most challenging diseases in medicine. DR. MEHRDAD FARROKHI 74 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 75 4AI IN CLINICAL DECISIONMAKING IN DENTISTRY AND ORAL MEDICINE Background Artificial intelligence is increasingly reshaping healthcare delivery, and dentistry and oral medicine are no exception. These fields combine visually rich diagnostics, procedural precision, longitudinal patient management, and preventive care, making them well suited for computational support. Dental and oral health professionals routinely interpret radiographs, photographs, histopathology, clinical notes, and laboratory data while balancing patient preferences, cost considerations, and evolving evidence. Clinical decision-making in dentistry and oral medicine therefore involves complexity that extends beyond isolated diagnostic acts and into comprehensive care planning. Artificial intelligence offers tools to manage this complexity by augmenting human expertise with data-driven insights, pattern recognition, and predictive modeling. Dentistry has historically relied on clinician experience, visual assessment, and manual measurements. While these remain essential, modern dental practice now generates large 76 volumes of digital data, including cone beam computed tomography scans, intraoral radiographs, digital impressions, electronic dental records, and chairside photographs. Oral medicine further integrates histopathology, microbiology, immunology, and systemic disease associations. Artificial intelligence enables the integration and analysis of these diverse data sources, supporting more consistent diagnoses, earlier detection of disease, personalized treatment planning, and improved monitoring of outcomes. Importantly, artificial intelligence in dentistry and oral medicine is not intended to replace the dentist, oral medicine specialist, or dental team. Instead, it functions as a clinical decision support system that enhances accuracy, efficiency, and consistency while preserving professional judgment and patient-centered care. Understanding how artificial intelligence is applied, validated, and ethically integrated is critical for its responsible adoption in oral healthcare. Foundations of Artificial Intelligence in Dentistry and Oral Medicine Artificial intelligence refers to computational systems capable of performing tasks that typically require human intelligence, such as learning from data, recognizing patterns, and supporting CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 77 decisions. In dentistry and oral medicine, the most commonly used forms of artificial intelligence include machine learning, deep learning, natural language processing, and large language models. Machine learning algorithms learn relationships between inputs and outputs from data. Supervised learning uses labeled examples, such as radiographs annotated for caries or periodontal bone loss, to train predictive models. Unsupervised learning identifies hidden patterns within unlabeled data, such as clustering patients based on disease progression or risk profiles. Reinforcement learning, while less common in dentistry, has potential applications in optimizing treatment sequences or behavioral interventions. Deep learning, a subset of machine learning, uses multi-layer neural networks to model complex and non-linear relationships. Convolutional neural networks are particularly important in dental imaging because they excel at analyzing visual data such as radiographs, photographs, and scans. Natural language processing allows artificial intelligence systems to extract meaningful information from unstructured clinical text, including dental notes, referral letters, and pathology reports. Large language models extend these capabilities by summarizing records, assisting documentation, and supporting clinical communication, though their use requires strict safeguards. DR. MEHRDAD FARROKHI 78 These methods are increasingly combined into multimodal systems that reflect real-world dental and oral medicine workflows. The effectiveness of artificial intelligence depends not only on algorithm choice but also on data quality, clinical relevance, validation rigor, and thoughtful integration into practice. AI in Dental Imaging and Radiographic Interpretation Imaging is central to diagnosis and treatment planning in dentistry and oral medicine. Artificial intelligence has made its most visible impact in this area by assisting with interpretation of dental radiographs, cone beam computed tomography scans, and clinical photographs. In caries detection, deep learning algorithms analyze bitewing and periapical radiographs to identify early enamel and dentin lesions. These systems can detect subtle radiolucencies that may be overlooked, particularly in earlystage disease or high-volume clinical settings. Artificial intelligence-assisted caries detection has been shown to improve sensitivity and consistency while reducing inter-clinician variability. Importantly, these tools support early intervention and preventive care rather than overtreatment. Periodontal assessment is another major application. Machine learning models can CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 79 quantify alveolar bone loss from radiographs, classify periodontal disease severity, and track changes over time. Automated bone level measurements provide objective baselines that support diagnosis, prognosis, and monitoring of treatment response. In implant dentistry, artificial intelligence assists in evaluating bone quality, anatomical landmarks, and proximity to critical structures, enhancing safety and planning accuracy. Cone beam computed tomography interpretation is complex and time-consuming. Artificial intelligence systems can segment anatomical structures, identify pathologies such as cysts or impacted teeth, and flag incidental findings. These tools are particularly valuable in oral and maxillofacial radiology and oral surgery, where comprehensive evaluation is essential. In oral medicine, imaging-based artificial intelligence supports detection of potentially malignant disorders and oral cancer. Algorithms trained on clinical photographs and radiographs can identify suspicious lesions, support referral decisions, and prioritize high-risk cases. While not a substitute for biopsy and specialist evaluation, these systems can enhance early detection and access to care. AI in Oral Pathology and Histopathology DR. MEHRDAD FARROKHI 80 Oral pathology plays a critical role in diagnosing diseases of the oral cavity, jaws, and salivary glands. The transition to digital pathology has enabled artificial intelligence to assist in histopathologic interpretation. Deep learning models can analyze whole-slide images to identify dysplasia, malignancy, inflammatory patterns, and infectious organisms. In oral squamous cell carcinoma, artificial intelligence systems have demonstrated the ability to differentiate benign, dysplastic, and malignant tissues, as well as to grade tumors and assess margins. These capabilities support pathologists by reducing workload, improving consistency, and highlighting areas of concern. In salivary gland pathology, where diagnosis can be particularly challenging due to morphological overlap, artificial intelligence offers additional pattern recognition support. Artificial intelligence also contributes to prognostication by extracting features from histopathology images that correlate with outcomes such as recurrence or survival. When combined with clinical and molecular data, these insights support risk stratification and treatment planning in oral oncology. AI in Diagnosis and Management of Oral Diseases Oral medicine encompasses a wide range CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 81 improved clinical outcomes and higher adherence rates have been reported. These systems illustrate how feedback loops and predictive analytics can personalize the timing, intensity, and duration of rehabilitation interventions. Digital Twins and Predictive Simulation An emerging frontier in rehabilitation medicine is the development of artificial intelligence driven digital twins, which are dynamic virtual representations of individual patients. These models integrate sensor data and clinical information to simulate rehabilitation environments and predict functional outcomes. When combined with adaptive rehabilitative robots and artificial intelligence controlled exoskeletons, digital twins enable bidirectional collaboration between humans and machines to restore mobility. Intelligent actuators guided by patient specific biomechanical models can automatically adjust assistance levels, supporting optimized and individualized recovery trajectories. Cognitive and Neurobehavioral Rehabilitation Artificial intelligence is also advancing cognitive and behavioral rehabilitation. Predictive analytics applied to traumatic brain injury datasets DR. MEHRDAD FARROKHI 94 can identify early risk of cognitive decline. Natural language processing tools analyze emotional tone and engagement during cognitive behavioral therapy sessions, while virtual reality platforms that incorporate adaptive artificial intelligence adjust task complexity to match individual cognitive profiles. These approaches have demonstrated improvements in memory, executive function, and behavioral outcomes, highlighting the role of artificial intelligence in addressing cognitive and psychological recovery alongside motor rehabilitation. Cross Cutting Themes Evidence Quality and Reporting Standards Rigorous validation remains a fundamental requirement for safe clinical translation of artificial intelligence tools. Limitations such as lack of external validation, overfitting, and model opacity pose risks to real world deployment. In addition to traditional performance metrics, calibration analysis and decision curve evaluation are essential to ensure clinical reliability. Embedding these standards into rehabilitation research design supports responsible and transparent use of artificial intelligence. Regulation and Liability Regulatory frameworks increasingly classify CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 95 medical artificial intelligence systems as high risk technologies, requiring structured conformity assessments, traceable datasets, and continuous human oversight. Regulatory approval pathways for artificial intelligence enabled rehabilitation technologies emphasize post market monitoring and accountability. Together with evolving product liability legislation, these requirements clarify responsibility when algorithmic errors occur and reinforce the importance of ongoing system evaluation. Ethics and Human Factors The adoption of artificial intelligence in rehabilitation introduces ethical and psychosocial challenges, including ambiguity in patient consent, concerns regarding data privacy, algorithmic bias, and the potential for clinical deskilling. Transparency, explainability, and inclusive data representation are critical safeguards. Human factors research consistently indicates that patients and clinicians view artificial intelligence as a supportive adjunct to professional judgment rather than a replacement, provided that accountability remains with human decision-makers. Continuous education and participatory design approaches are essential for maintaining trust, usability, and clinical competence. DR. MEHRDAD FARROKHI 96 Future Directions Future developments in artificial intelligence for rehabilitation will emphasize multimodal data fusion, continuous model monitoring, and the use of large foundation models. Integrating biomechanics, imaging, speech, and genomic data will enable more comprehensive patient representations. Federated and continual learning frameworks will allow models to adapt to evolving populations while reducing algorithmic bias. Digital twin ecosystems are expected to incorporate real time data from wearables, home monitoring systems, and electronic health records to refine personalized rehabilitation pathways. Adaptive robotic platforms and exoskeletons will increasingly operate in coordination with clinician supervised artificial intelligence to deliver precise and responsive therapy. Ethical governance structures and equitable digital infrastructure will be essential to support scalability and global implementation. Conclusion Artificial intelligence is reshaping rehabilitation medicine by transforming raw multimodal data into personalized and actionable clinical intelligence. Current applications include automated functional assessment, predictive outcome modeling, cognitive and behavioral CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 97 rehabilitation, digital twin simulation, and continuous telemonitoring. These innovations offer substantial gains in precision, efficiency, and access to care. Their long term success depends on rigorous clinical validation, strong ethical governance, and thoughtful human centered integration. Rather than replacing clinical expertise, artificial intelligence extends it by enabling clinicians to deliver individualized and data informed recovery strategies. As predictive analytics, adaptive robotics, and regulatory frameworks continue to evolve, rehabilitation medicine is entering a new era focused on precision recovery and sustainable patient centered care. DR. MEHRDAD FARROKHI 98 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 99 6AI IN CLINICAL DECISIONMAKING IN SURGERY AND EMERGENCY MEDICINE Background Time-critical care in surgery and emergency medicine demands rapid, high-consequence clinical decisions made under uncertainty and significant cognitive load. Within this context, artificial intelligence, including machine learning, deep learning, natural language processing, and large language models, has progressed from pilot projects to early clinical implementations of AI-based decision support systems. These technologies offer advanced pattern recognition across multimodal data sources, including electronic health records, imaging, physiologic waveforms, and clinical text. Such capabilities support earlier detection of patient deterioration, more consistent triage and diagnosis, improved risk stratification, enhanced operational flow, and more individualized perioperative care. This chapter brings together current evidence and practical guidance for clinicians and clinical researchers on integrating artificial intelligence into decision-making across operative and emergency settings. It defines core methodological approaches, reviews validated 100 applications in perioperative and emergency workflows, and examines cross-cutting issues including evidence quality, generalizability and bias, data governance, regulation and liability, and human factors. The chapter concludes with practical recommendations and near-term research priorities. For readers who are new to clinical artificial intelligence, brief and pragmatic definitions are provided throughout. Surgeons and emergency physicians routinely confront diagnostic and judgment errors, which rank immediately after technical complications among the leading contributors to preventable adverse events. These challenges arise in timepressured and information-dense environments. Enhancing human reasoning with continuously updated predictions derived from live electronic health record streams and imaging data holds promise for reducing variability and improving outcomes, provided that these systems are interpretable, well calibrated, and appropriately embedded within clinical workflows. In emergency medicine, the literature has expanded rapidly, documenting hundreds of AI-based clinical decision support tools. However, only a small proportion of models progress to advanced testing or sustained clinical implementation. This gap between promising models and real-world impact highlights the need for methodologically robust design, transparent reporting, and rigorous evaluation standards. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 101 Recent introductory frameworks have been developed to support nonexpert clinicians by clarifying artificial intelligence concepts and terminology and by addressing system limitations such as hallucinations, bias, and the need for explainability, which are essential prerequisites for responsible clinical integration. Methodological Foundations for Clinical AI Core methods and what they mean clinically Supervised machine learning methods, including regularized regression, gradient boosting, and neural networks, learn mappings from labeled data. Unsupervised learning identifies latent structures within unlabeled datasets. Deep learning introduces multiple processing layers to extract complex features, while large language models analyze and generate clinical text. In emergency and surgical settings, these approaches are used for risk prediction, triage, imaging analysis, and documentation support. However, their probabilistic outputs and potential for erroneous or fabricated responses require ongoing human oversight. Discrimination is not enough: calibration and net benefit Beyond performance metrics such as accuracy or DR. MEHRDAD FARROKHI 102 area under the receiver operating characteristic curve, clinically useful models must demonstrate strong calibration, meaning predicted risk aligns closely with observed outcomes, and meaningful decision-curve utility across clinically relevant thresholds. Stepwise implementation work in perioperative oncology demonstrates model development, validation, workflow integration, and impact evaluation, illustrating how translation from research to bedside practice should be approached. Reporting, appraisal, and trial standards The use of standardized reporting and appraisal frameworks for artificial intelligence studies has become increasingly important. These frameworks emphasize clear problem specification, transparent data provenance, appropriate handling of missing data and model drift, and explicit reporting of model thresholds and calibration. Adherence to these standards supports reproducibility, facilitates assessment of bias and applicability, and strengthens confidence in AI-based clinical decision support systems. Safety and software lifecycle Risk management and software development processes for clinical artificial intelligence systems should align with established medical device safety and lifecycle standards. These processes CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 103 include systematic hazard identification, implementation of risk controls, continuous monitoring, and structured software lifecycle management to ensure safety and reliability throughout deployment. Regulatory context (US/EU) In the United States, regulatory oversight of artificial intelligence enabled medical devices reflects evolving expectations for software classified as a medical device. In the European Union, current regulations classify many surgical and emergency department artificial intelligence systems as high risk, requiring conformity assessment, human oversight, and post-market monitoring. Updated product liability frameworks extend responsibility to software and artificial intelligence systems, reinforcing accountability when adverse events occur. Applications in Surgery Preoperative prediction and shared decision-making Preoperative machine learning models increasingly outperform traditional risk scores for predicting complications such as infection, venous thromboembolism, and readmission. By integrating comorbidities, functional status, tumor characteristics, laboratory results, imaging findings, and prior healthcare utilization, these DR. MEHRDAD FARROKHI 104 models support more refined risk stratification. Registry-derived tools applied to colorectal cancer populations have demonstrated strong predictive performance for one-year mortality and have shown reductions in postoperative complications when implemented within risktiered perioperative care pathways. These examples illustrate how predictive outputs can be translated into actionable care bundle decisions rather than remaining isolated within analytic dashboards. Intraoperative guidance by computer vision and robotics Intraoperative computer vision systems can recognize anatomical structures and surgical phases, identify deviations from expected workflows, and provide context-aware prompts. When combined with robotic platforms and augmented reality systems, these tools support standardization of complex procedural steps and reduce cognitive burden. Skill assessment analytics further contribute to surgical training and performance feedback through motion tracking and error detection. As system latency and overlay accuracy improve, human factors design becomes increasingly important in determining how, when, and to whom recommendations are presented. Postoperative surveillance and CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 105 early complication detection Artificial intelligence driven postoperative surveillance systems analyze vital signs, laboratory data, and clinical notes to identify early signals of complications such as sepsis, hemorrhage, or respiratory failure. Their effectiveness depends on appropriate alert calibration, clear escalation protocols, and strategies to mitigate alarm fatigue. Surgeons’ mixed attitudes toward AI reflecting both curiosity and caution International surveys of trauma and emergency surgeons demonstrate interest in artificial intelligence alongside mistrust and limited familiarity. Many clinicians continue to prioritize clinical guidelines and multidisciplinary consultation over algorithmic outputs. These findings underscore the need for foundational artificial intelligence education within surgical training programs and professional society curricula. Applications in Emergency Medicine Triage, surge prediction, and operational flow Artificial intelligence assisted triage models that integrate vital signs, demographic variables, and free-text clinical notes frequently outperform DR. MEHRDAD FARROKHI 106 conventional triage scales in predicting hospital admission, intensive care unit transfer, and timecritical outcomes. These capabilities enable earlier escalation of care and proactive management of staffing and bed capacity during periods of high demand. Sepsis and shock: from detection to action Early warning systems that analyze continuously streaming physiologic and laboratory data support earlier recognition and treatment of sepsis and shock. Successful deployment depends on integration with nursing workflows, clinician training, and clearly defined escalation pathways, as demonstrated in real-world implementation programs. AI applications in diagnostic support using imaging and clinical text Imaging Deep learning systems applied to head computed tomography, chest radiography, and point-ofcare ultrasound increasingly assist interpretation in emergency settings. Recent syntheses report high diagnostic accuracy, although performance remains sensitive to domain shift and labeling quality. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 107 Regulated exemplars Regulated artificial intelligence tools for vascular imaging have demonstrated significant improvements in time to specialist notification and workflow efficiency. Ongoing studies continue to evaluate their impact on patientcentered outcomes and clinical effectiveness. ECG and Vitals Models that predict cardiac arrest and clinical decompensation using electrocardiography and continuous physiologic monitoring show considerable promise. However, these systems require careful site-specific calibration and clearly defined clinician-in-the-loop thresholds to ensure safe and effective use. Without appropriate local adaptation and oversight, model performance may vary, potentially limiting clinical reliability and acceptance. Natural Language Processing and Large Language Models Free-text triage notes and clinical documentation provide rich contextual information that can enhance early risk estimation and disposition decisions. Large language models can support clinicians by summarizing documentation and extracting salient clinical features. Nevertheless, robust safeguards are required to prevent hallucinations and to ensure protection of DR. MEHRDAD FARROKHI 108 sensitive and protected health information. Clear governance and validation processes are essential before widespread clinical adoption. Prognostic and Disposition Decision Support Emergency department models that estimate readmission risk, clinical deterioration, or the need for hospital admission can complement physician gestalt, particularly in geriatric and cardiac patient populations. Despite this potential, generalizability across settings remains inconsistent, and prospective evidence demonstrating real-world clinical impact is still limited. These tools should therefore be applied cautiously and evaluated continuously within clinical workflows. Education and Simulation Large language models can support medical education by grading examinations, generating explanations, and facilitating continuous professional development. Scoping reviews highlight both the promise of these applications and the importance of supervision to avoid over-reliance, propagation of misinformation, or erosion of critical thinking skills. Educational use of these systems should emphasize transparency and active learner engagement. Bridging the Translation CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 109 Gap between Research and Clinical Adoption Comprehensive scoping reviews demonstrate that the vast majority of emergency department artificial intelligence clinical decision support studies remain in early or preclinical phases. Key barriers include limited access to highquality data, performance drift over time, challenges in sociotechnical integration, and the substantial costs associated with rigorous evaluation. Addressing these barriers is essential for translating promising research into sustained clinical impact. Cross-Cutting Themes Evidence Quality and Transparent Reporting High-quality evidence and transparent reporting are essential for moving artificial intelligence tools from feasibility studies to dependable clinical practice. Reports should include calibration assessments, decision-curve analyses, and external validation whenever feasible. Prospective and multicenter trials, although resource-intensive, remain the reference standard for establishing generalizability and demonstrating meaningful clinical benefit. Generalizability, Bias, and Fairness Artificial intelligence systems frequently DR. MEHRDAD FARROKHI 110 experience performance degradation when deployed across different hospitals, imaging platforms, or patient populations. Bias may arise from imbalances in age, sex, ethnicity, or language representation, as well as from temporal changes in case mix or clinical practice. Regular fairness audits, monitoring for domain shift, and scheduled recalibration are therefore essential. Clinician trust depends not only on statistical performance but also on interpretability and a clear understanding of how and when systems may fail. Limited artificial intelligence literacy among clinicians is associated with skepticism and underuse, highlighting the need for ongoing education. Data Governance and Privacy Reliable artificial intelligence performance depends on access to high-quality, wellannotated data and secure mechanisms for data sharing. Federated learning and secure aggregation approaches allow collaborative model development across institutions without transferring protected health information, addressing privacy and regulatory constraints. Standardized data models, consistent labeling practices, and detailed documentation of preprocessing steps are equally important to ensure traceability and reproducibility across sites. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 111 Regulation and Liability Before clinical implementation, it is critical to determine whether an algorithm qualifies as software as a medical device. In the United States, this requires compliance with regulatory frameworks governing artificial intelligence and machine learning enabled medical devices. Within the European Union, current regulations classify most surgical and emergency care artificial intelligence tools as high risk, requiring conformity assessment, human oversight, and post-market monitoring. Updated liability directives extend responsibility to artificial intelligence and software developers, with implications for healthcare institutions and vendors alike. Ethics and Human Factors Ethical deployment of artificial intelligence requires that clinicians remain central to decisionmaking processes. High-stakes recommendations must always be verified by a qualified professional, with clear documentation of responsibility and clinical rationale. Human factors design should support situational awareness by clarifying who receives alerts, in what sequence, and with what expected response. Systems that enhance team coordination and clinical communication, rather than replacing judgment, are more likely to sustain safety, trust, and long-term adoption. DR. MEHRDAD FARROKHI 112 Future Directions Despite notable progress, the safe and sustainable integration of artificial intelligence into surgical and emergency care remains at an early stage. Several priorities are expected to shape future development and evaluation efforts. Prospective Effectiveness Studies Future research must move beyond performance metrics such as area under the receiver operating characteristic curve to focus on patientcentered outcomes, workflow efficiency, and costeffectiveness. Randomized or quasi-experimental studies that embed artificial intelligence tools within real clinical environments are needed to determine tangible benefits. Structured guidelines for trial design and reporting can support consistent evaluation of outcomes, biases, and implementation processes. Multimodal Data Fusion An important frontier involves combining heterogeneous data streams, including electronic health records, imaging, physiologic waveforms, clinical text, and wearable sensor data, into unified predictive models. Multimodal systems have the potential to improve early detection of deterioration, refine perioperative and postoperative risk stratification, and personalize clinical decision support. Achieving these goals CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 113 infections, tumors, and congenital anomalies. In pediatric pathology, artificial intelligence supports analysis of histologic images to classify tumors, identify inflammatory patterns, and assess margins. These tools are particularly valuable in rare pediatric cancers, where expertise may be limited. Artificial intelligence also assists in reducing radiation exposure by optimizing imaging protocols and supporting decision-making regarding the necessity of studies. This aligns with the principle of minimizing harm in pediatric care. AI in Preventive Care and Population Health Preventive care is a cornerstone of pediatrics and obstetrics-gynecology. Artificial intelligence supports population health by identifying individuals at increased risk and guiding preventive interventions. In pediatrics, machine learning models analyze demographic, social, and clinical data to identify children at risk for developmental delays, obesity, or mental health conditions. These insights support early intervention and resource allocation. In obstetrics-gynecology, artificial intelligence assists in identifying populations at risk for adverse pregnancy outcomes or gynecologic DR. MEHRDAD FARROKHI 126 cancers. Population-level analytics support screening strategies, vaccination programs, and health education initiatives. Artificial intelligence also plays a role in addressing health disparities by identifying gaps in care and supporting targeted outreach. However, careful attention to bias and equity is essential to ensure that these tools promote, rather than undermine, fairness. AI in Education, Communication, and Clinical Support Artificial intelligence has important applications in education and clinical support for pediatrics and obstetrics-gynecology. Simulation systems powered by artificial intelligence provide feedback on clinical skills, decision-making, and procedural performance. These tools support training in high-risk scenarios such as neonatal resuscitation and obstetric emergencies. Large language models assist with documentation, patient education, and clinical communication. They can generate draft notes, summarize records, and support counseling by translating complex information into understandable language. In pediatrics and obstetrics-gynecology, effective communication is particularly important due to family involvement and sensitive topics. Clinical decision support systems integrate CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 127 artificial intelligence with guidelines and patient data to support evidence-based care. These systems assist clinicians in navigating complex protocols while maintaining individualized decision-making. Ethical, Legal, and Practical Considerations The integration of artificial intelligence into pediatrics and obstetrics-gynecology raises important ethical and practical issues. Data privacy is a major concern, particularly when dealing with sensitive information related to children, pregnancy, and reproductive health. Robust governance, secure data handling, and regulatory compliance are essential. Bias and fairness are critical considerations. Artificial intelligence models trained on nonrepresentative datasets may perpetuate disparities in maternal and child health. Ensuring diversity in data and conducting fairness assessments are necessary steps toward equitable deployment. Interpretability and transparency influence clinician trust and adoption. Clinicians must understand the limitations and uncertainty associated with artificial intelligence recommendations. Explainable artificial intelligence approaches support this understanding by highlighting relevant features and confidence levels. DR. MEHRDAD FARROKHI 128 From a practical perspective, successful implementation depends on workflow integration, clinician training, and institutional support. Artificial intelligence tools must align with clinical processes and enhance, rather than hinder, care delivery. Future Directions in Pediatrics and Obstetrics-Gynecology The future of artificial intelligence in pediatrics and obstetrics-gynecology lies in deeper integration, multimodal analysis, and collaborative human and machine intelligence. Systems that combine imaging, genomics, clinical text, physiologic monitoring, and social determinants of health will provide more comprehensive insights into patient care. Personalized care pathways, predictive monitoring, and telehealth integration are expected to expand. Artificial intelligence may play a growing role in remote monitoring of pregnancy, chronic pediatric conditions, and postpartum recovery, improving access to care and continuity. Ongoing research, validation, and ethical oversight will be essential as artificial intelligence systems evolve. Collaboration between clinicians, data scientists, patients, and policymakers will shape responsible and effective adoption. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 129 Conclusion Artificial intelligence is increasingly influencing clinical decision-making in pediatrics and obstetrics-gynecology by enhancing diagnosis, risk assessment, treatment planning, and monitoring across the continuum of care. Its applications span prenatal screening, labor and delivery, neonatal care, pediatric diagnosis, gynecology, and preventive health. When implemented responsibly, artificial intelligence supports clinicians in managing complexity and uncertainty while preserving human judgment and patient-centered care. The successful integration of artificial intelligence requires rigorous validation, ethical governance, transparency, and education. By embracing these principles, pediatrics and obstetrics-gynecology can harness artificial intelligence to improve outcomes, reduce disparities, and support safer and more personalized care for mothers, children, and families. DR. MEHRDAD FARROKHI 130 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 131 8AI IN CLINICAL DECISIONMAKING IN PSYCHIATRY AND BEHAVIORAL SCIENCES Background Artificial intelligence is rapidly transforming psychiatry and the behavioral sciences. Traditionally, mental health assessment relied heavily on subjective approaches such as self-report measures, clinical interviews, and behavioral observation. While these methods remain essential to clinical practice, computational tools now augment them by enhancing precision, consistency, and accessibility. Techniques such as machine learning and natural language processing are evolving from experimental research prototypes into powerful clinical decision-support systems. This ongoing evolution is redefining how psychiatric disorders are diagnosed, treated, and managed across global healthcare systems, while also addressing workforce shortages and the inherent complexity involved in diagnosing heterogeneous psychiatric conditions. Artificial intelligence contributes across all stages of mental healthcare by generating datadriven insights that complement, rather than replace, clinical expertise. One of its most 132 impactful roles lies in early detection and diagnostic support. By analyzing large and multimodal datasets, including electronic health records, neuroimaging data, speech patterns, and information from wearable devices, AI provides clinicians with comprehensive and objective insights. Natural language processing models can identify linguistic markers that are indicative of depression, anxiety, or cognitive decline. Smartphone-based applications that monitor sleep patterns, typing behavior, and social interaction are capable of detecting subtle behavioral changes that often precede acute mental health episodes, thereby enabling earlier and more timely intervention. Meta-analyses report diagnostic accuracies ranging from 80 to 85 percent, depending on model architecture, data quality, and clinical context. By capturing subtle behavioral and physiological signals that are frequently missed during routine clinical observation, AI enhances diagnostic confidence and reduces inter-clinician variability. Beyond diagnosis, artificial intelligence enables precision psychiatry by supporting the individualization of treatment selection. Predictive models that integrate genetic information, treatment history, and clinical characteristics can identify the most effective therapies and medication combinations for individual patients. Evidence demonstrates pooled effect sizes between 0.80 and 0.85 CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 133 for AI-assisted treatment prediction, highlighting meaningful improvements in therapeutic outcomes and clinical response. In addition, AIdriven chatbots and virtual assistants provide scalable and continuous mental health support using structured psychotherapeutic approaches such as cognitive behavioral therapy. Advanced technologies, including AI-enhanced virtual reality exposure therapy, are also being explored as innovative tools to expand access to evidencebased interventions and to enhance treatment engagement. Despite these significant advancements, the implementation of artificial intelligence in psychiatry raises important ethical and practical challenges. The sensitive nature of mental health data intensifies concerns related to privacy, confidentiality, data security, and informed consent. Algorithmic bias remains a critical issue, as models trained on non-representative datasets may inadvertently perpetuate existing health disparities. Transparency and explainability are therefore essential, as opaque algorithms can undermine clinician trust, hinder clinical adoption, and complicate issues of accountability. Ethical frameworks, such as the Integrated Ethical Approach for Computational Psychiatry, emphasize core principles including beneficence, autonomy, justice, and transparency to guide responsible development and deployment. Practical barriers to implementation must also be DR. MEHRDAD FARROKHI 134 carefully addressed. The deployment of AI systems in clinical psychiatry requires rigorous validation, strong cross-disciplinary collaboration, and supportive regulatory and policy frameworks. Clinicians may understandably exercise caution when relying on AI-generated recommendations without corroboration from another qualified expert. This concern underscores the importance of positioning artificial intelligence as a supportive clinical tool rather than a replacement for human judgment and professional reasoning. Ultimately, responsible adoption of artificial intelligence promises a future in which psychiatry is both technologically advanced and ethically grounded. When applied responsibly, transparently, and in alignment with clinical expertise, AI has the potential to improve diagnostic accuracy, personalize treatment strategies, and enhance overall patient outcomes, thereby empowering both clinicians and patients within modern mental healthcare systems. The rapid advancement of artificial intelligence has led to transformative changes in healthcare, and mental and behavioral health nursing is positioned at the forefront of this digital revolution. Psychiatric and behavioral disorders, including dementia, schizophrenia, depression, suicide risk, and psychological distress associated with chronic illness, present complex challenges for healthcare systems and district providers. Nurses, who are often the first point of CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 135 contact and the most continuous source of care, face increasing demands in managing these multifaceted conditions. The complexity of psychiatric care highlights the need for innovative tools that can support clinical judgment, continuity of care, and timely interventions. Artificial intelligence has the potential to increase diagnostic accuracy, support early detection, personalize treatment strategies, and improve patient outcomes through real-time monitoring and predictive analytics. Despite growing interest in integrating artificial intelligence into psychiatric and behavioral health nursing, its actual implementation in routine clinical practice remains limited. This chapter explores how artificial intelligence is currently being applied, the opportunities it offers for enhancing nursing practice, and the challenges that must be addressed for it to become a central component of mental health nursing care. AI in Dementia Care: Early Diagnosis of Behavioral and Psychological Symptoms One of the most important areas for artificial intelligence in psychiatric nursing involves the early diagnosis and management of behavioral and psychological symptoms of dementia. Review studies evaluating international research have proposed the use of artificial intelligence-based technologies to support nursing interventions DR. MEHRDAD FARROKHI 136 for individuals living with dementia. These studies have categorized artificial intelligence applications into three primary areas: predicting disease progression, screening and evaluating symptoms, and managing behavioral and psychological symptoms of dementia. Most existing research has focused on symptom management rather than early diagnosis or disease prediction. Findings suggest that artificial intelligence can serve as a reliable tool for evaluation, supervision, and planning of therapeutic interventions that assist nurses in providing effective dementia care. However, many of the currently available tools have primarily been examined in research settings and have not yet been widely implemented in everyday nursing practice. This gap highlights the need for further clinical validation and integration of artificial intelligence tools within dementia care environments. AI in Psychiatric Mental Health Nurse Practitioner Practice Research has demonstrated how artificial intelligence supports diagnostic accuracy in psychiatry and personalized psychological care. By analyzing large datasets through machine learning algorithms, artificial intelligence can identify subtle clinical patterns that are often missed by traditional assessment methods. For example, changes in mood, behavior, and early CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 137 indicators of relapse can be detected through data collected from wearable devices and mobile health applications. In addition, psychiatric mental health nurse practitioners are encouraged to adapt to advancements in machine learning while addressing ethical considerations associated with artificial intelligence. Ethical issues, including data privacy, algorithmic bias, and responsible data use, must be carefully considered and integrated into clinical education and professional training. The evaluation of realtime symptoms through artificial intelligence introduces new possibilities for proactive care. Passive data collected from smartphones, such as sleep patterns, sound analysis, and text-based emotional indicators, can reveal early signs of depression or mania, allowing nurses to intervene before a crisis occurs. These developments demonstrate how artificial intelligence can enhance continuous monitoring while supporting timely and preventive psychiatric care. AI Applications across Mental Health Disorders Scoping studies examining the application of artificial intelligence in psychiatric nursing have identified a growing focus on disorders such as dementia, schizophrenia, and autism. These studies report the use of technologies including machine learning and robotic learning across DR. MEHRDAD FARROKHI 138 multiple mental health conditions. Findings have identified three core areas of artificial intelligence application: personalized care planning, symptom monitoring, and risk assessment. For instance, artificial intelligence algorithms can predict suicide risk by analyzing electronic health records and social media data, thereby enabling earlier and more targeted nursing interventions. Machine learning models are also capable of detecting early symptoms of schizophrenia through analysis of speech patterns and behavioral data. Despite these advances, concerns remain regarding limited population diversity in artificial intelligence models, insufficient largescale studies, and the presence of data bias. These challenges underscore the importance of educating nurses in the ethical and equitable use of artificial intelligence to ensure fair and inclusive mental health care. AI in Psychological Distress of Oral Cancer Patients with oral cancer frequently experience anxiety, depression, and body image disturbances that are often identified during routine clinical care but may not receive adequate psychological support. Reviews of artificial intelligence applications in the psychological aspects of oral cancer highlight an underexplored area within psychiatric nursing. Artificial intelligence tools, including natural language processing CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 139 for emotional tone analysis, predictive models for identifying high-risk groups, and real-time digital health consultation systems, can provide psychological support for these patients. Wearable devices are also able to monitor stress levels and sleep disturbances, allowing nurses to intervene in a timely manner. This approach demonstrates how artificial intelligence can address both the psychological and physiological dimensions of chronic illness, enabling psychiatric nurses to deliver comprehensive, early, and holistic care for individuals living with oral cancer. AI in Prevention of Suicide and Behavioral Prediction The use of artificial intelligence in suicide prevention represents a significant area of psychiatric nursing research. Large-scale reviews indicate that behavioral prediction using machine learning has become an important focus within mental health studies. Findings highlight the potential of natural language processing and ecological momentary assessment to detect suicidal ideation. Artificial intelligence can analyze language patterns in social media posts, clinical narratives, and patient communications to identify expressions of hopelessness or suicidal thoughts. Ecological momentary assessment allows nurses to track emotional changes and behavioral DR. MEHRDAD FARROKHI 140 triggers throughout the day, providing realtime insights through mobile technologies. While these tools demonstrate strong potential, they are intended to reinforce therapeutic judgment and human communication rather than replace them. Ethical practice, clinical supervision, and culturally sensitive communication remain essential components of suicide prevention strategies that incorporate artificial intelligence. Observations of Nursing with Digital Assistance in Inpatient Psychiatry Wards The application of artificial intelligence in inpatient psychiatric settings has also been explored through studies examining digitally assisted nursing observations. In acute psychiatric wards, frequent safety checks can disrupt sleep and delay patient recovery. Digitally assisted monitoring systems using artificial intelligenceenabled sensors have been proposed as an alternative approach to maintaining patient safety while reducing unnecessary disturbances. Studies have reported successful implementation of these systems, with extended periods of monitoring achieved without adverse effects. Patient feedback has indicated improved experiences, reduced stress, and enhanced perceptions of dignity. Although these studies were preliminary and nonrandomized, they demonstrate the potential for CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 141 artificial intelligence to support safer inpatient environments while preserving patient comfort and respect. Conclusion: An Evolving Approach to AI in Psychiatric Nursing Artificial intelligence demonstrates strong potential to enhance mental and behavioral health nursing practice. It can support every stage of psychiatric care, from early identification of dementia symptoms and suicide risk to personalized care planning and real-time monitoring. By enabling proactive, individualized, and data-informed care, artificial intelligence has the capacity to transform psychiatric nursing while maintaining a human-centered approach. Several key considerations must guide this integration. Artificial intelligence technologies require robust clinical validation in real-world settings and across diverse populations. Nursing education programs should incorporate artificial intelligence literacy and ethical reasoning to prepare nurses for responsible use. Clear ethical frameworks are needed to guide data use, privacy protection, and accountability. Active nurse engagement in the design, implementation, and evaluation of artificial intelligence systems is essential to ensure relevance and effectiveness. As artificial intelligence continues to shape mental health care, psychiatric and mental health nurses are uniquely positioned to lead its ethical, DR. MEHRDAD FARROKHI 142 empathetic, and evidence-informed integration. Rather than replacing nursing practice, artificial intelligence should be embraced as a supportive tool that enhances the quality of psychiatric care, promotes better patient outcomes, and contributes to a sustainable future for mental health nursing. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 143 DR. MEHRDAD FARROKHI 144 9AI IN CLINICAL DECISIONMAKING IN RADIOLOGY AND PATHOLOGY Background In recent decades, the volume of medical imaging data, including radiologic scans and digital pathology slides, has expanded at an exponential rate. Accurate and timely interpretation of these images is essential not only for establishing diagnoses but also for guiding therapeutic decision making and patient management. However, human-related challenges, such as diagnostic errors, interpretive delays, and interobserver variability, have motivated researchers to investigate solutions based on artificial intelligence. Advances in deep learning, convolutional neural networks, and multimodal data integration now provide distinctive opportunities to enhance diagnostic accuracy, operational efficiency, and standardization across diverse healthcare systems. In the field of radiology, artificial intelligence has demonstrated value in two principal domains, namely interpretive and non-interpretive tasks. In interpretive applications, AI models can identify subtle abnormalities, such as pulmonary nodules, intracranial hemorrhages, and neoplastic 145 Deep learning algorithms analyze whole-slide images to identify regions of interest, classify tissue types, and detect pathologic features. In oncology, artificial intelligence assists in distinguishing benign from malignant lesions, grading tumors, and identifying invasive patterns. These capabilities support diagnostic accuracy and consistency, particularly in high-volume settings or where subspecialty expertise is limited. In hematopathology, artificial intelligence supports classification of blood smears and bone marrow samples by analyzing cellular morphology. In cytopathology, models assist in screening cervical cytology and fine needle aspiration samples. These tools reduce workload and help pathologists focus on complex or ambiguous cases. Artificial intelligence also contributes to quality assurance by detecting discrepancies, highlighting potential errors, and supporting peer review. By reducing variability and improving efficiency, artificial intelligence enhances the reliability of pathologic diagnoses that underpin clinical decision-making. AI in Molecular Pathology and Integrated Diagnostics Modern pathology increasingly integrates molecular diagnostics, including genomics, transcriptomics, and proteomics. Artificial DR. MEHRDAD FARROKHI 158 intelligence plays a critical role in managing and interpreting these complex datasets. Machine learning models analyze genomic data to identify mutations, predict functional significance, and associate molecular profiles with clinical outcomes. In oncology, artificial intelligence supports identification of actionable mutations and selection of targeted therapies. In hematologic malignancies, models integrate cytogenetics, molecular markers, and clinical variables to refine diagnosis and prognosis. A particularly promising area is the prediction of molecular alterations directly from histopathology images. Deep learning models can infer mutation status, gene expression patterns, and molecular subtypes based on tissue morphology. While these predictions do not replace molecular testing, they provide complementary insights that can guide further evaluation and decision-making. Integrated diagnostics combines radiologic imaging, pathology, molecular data, and clinical information into unified decision support systems. Artificial intelligence enables this integration by identifying cross-modal patterns and generating comprehensive assessments that support personalized care. AI in Prognostication and Risk Stratification CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 159 Prognostic assessment is a critical component of clinical decision-making in radiology and pathology. Artificial intelligence enhances prognostication by incorporating highdimensional data and identifying patterns beyond traditional scoring systems. In radiology, radiomics and deep learning models extract quantitative features from images that correlate with disease aggressiveness, treatment response, and survival. These features support risk stratification in oncology, cardiovascular disease, and chronic conditions. In pathology, artificial intelligence derived features from histology images predict outcomes such as recurrence, metastasis, and treatment response. By providing individualized risk estimates, artificial intelligence supports shared decisionmaking and personalized treatment planning. Clinicians can use these insights to tailor therapy intensity, surveillance strategies, and supportive care. AI in Education and Clinical Decision Support Artificial intelligence has important implications for education and clinical support in radiology and pathology. Training programs increasingly use artificial intelligence powered simulation tools that provide feedback on image interpretation and diagnostic reasoning. These tools support DR. MEHRDAD FARROKHI 160 competency-based education and continuous professional development. Clinical decision support systems integrate artificial intelligence with guidelines, evidence, and patient data to support diagnostic and management decisions. These systems assist clinicians in navigating complex information and staying current with rapidly evolving knowledge. Large language models can summarize literature, draft reports, and support interdisciplinary communication, provided appropriate safeguards are in place. Ethical, Legal, and Regulatory Considerations The integration of artificial intelligence into radiology and pathology raises significant ethical, legal, and regulatory challenges. Data quality and bias are central concerns. Models trained on non-representative datasets may perform poorly in diverse populations, potentially exacerbating health disparities. Ensuring diversity in training data and conducting fairness assessments are essential steps toward equitable deployment. Interpretability and transparency influence clinician trust and adoption. Radiologists and pathologists must understand the limitations and uncertainty associated with artificial intelligence outputs. Explainable artificial intelligence approaches aim to provide insight into model CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 161 behavior and decision logic. Regulatory frameworks increasingly address artificial intelligence as a medical device, requiring evidence of safety, effectiveness, and ongoing monitoring. Data privacy and security are particularly important given the sensitive nature of imaging and pathology data. Robust governance structures are required to maintain patient trust and comply with regulatory standards. From a practical perspective, successful implementation depends on workflow integration, training, and institutional support. Artificial intelligence tools must enhance rather than disrupt clinical practice. Future Directions in Radiology and Pathology The future of artificial intelligence in radiology and pathology lies in deeper integration, multimodal analysis, and collaborative human and machine intelligence. Systems that combine imaging, pathology, molecular data, and clinical context will provide more comprehensive and actionable insights. Large language models may play an increasing role in documentation, education, and interdisciplinary communication. Digital twins and predictive simulation models may support personalized diagnosis and treatment planning. DR. MEHRDAD FARROKHI 162 Continued research, validation, and ethical oversight will be essential as these technologies evolve. Conclusion Artificial intelligence is transforming clinical decision-making in radiology and pathology by enhancing diagnostic accuracy, efficiency, and consistency. Its applications span image interpretation, workflow optimization, digital pathology, molecular diagnostics, and prognostication. When implemented responsibly, artificial intelligence augments human expertise and supports more precise and personalized care. The successful integration of artificial intelligence requires rigorous validation, transparency, ethical governance, and human-centered design. Radiologists and pathologists remain central to clinical decision-making, with artificial intelligence serving as a powerful tool to manage complexity and improve patient outcomes. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 163 DR. MEHRDAD FARROKHI 164 10AI IN CLINICAL DECISIONMAKING IN OTHER MEDICAL SPECIALTIES Artificial Intelligence in Clinical Decision-Making in Anesthesia Background The integration of artificial intelligence into anesthesiology represents a pivotal evolution in perioperative medicine, fundamentally reshaping the paradigm of clinical decision making throughout the entire perioperative period. Artificial intelligence technologies, including machine learning, deep learning, and natural language processing, are increasingly used to enhance and optimize therapeutic strategies, improve diagnostic accuracy, and ameliorate outcomes in complex anesthetic care. In this context, artificial intelligence offers substantial transformative potential as a powerful augmentation of clinical practice, while not serving as a replacement for human decision making capabilities. The Complexity of Anesthetic Decision-Making Anesthesiologists operate within highly complex clinical environments, often managing 165 multiple patients with significant comorbidities, interpreting dynamic hemodynamic parameters, and simultaneously evaluating anesthetic depth. In such demanding and rapidly changing settings that require continuous real-time interpretation of physiologic data, even the most experienced clinicians may experience cognitive overload that can compromise clinical judgment. Artificial intelligence technologies that utilize machine learning algorithms are capable of detecting subtle physiological patterns that may not be readily apparent to clinicians, thereby supporting improved patient safety and decision making. Predictive Analytics in Preoperative Assessment Traditional perioperative risk assessment tools, such as the American Society of Anesthesiologists Physical Status Classification and commonly used cardiac risk indices, rely heavily on subjective clinical judgment and are constrained by a limited number of categorical variables. Artificial intelligence models fundamentally transform this paradigm by integrating comprehensive patient data derived from electronic health records, including pharmacological histories, imaging findings, and laboratory results, to predict perioperative risks more accurately. Machine learning models demonstrate enhanced accuracy in predicting complications such as respiratory failure, myocardial infarction, acute DR. MEHRDAD FARROKHI 166 kidney injury, and postoperative delirium. These predictive tools enable more precise and targeted preoperative optimization strategies and empower anesthesiologists to deliver personalized and data driven risk assessments. Intraoperative Monitoring and Management Artificial intelligence systems represent a paradigm shift in intraoperative monitoring and management. AI assisted technologies enable optimization of anesthetic depth monitoring through advanced electroencephalographic signal processing, facilitate early forecasting of intraoperative hypotension, and support precise ventilation management based on pulmonary mechanics. These algorithms analyze continuous streams of physiological data, including plethysmographic variability, arterial waveform morphology, and spectral entropy measurements, to enhance real time decision support and reduce the likelihood of intraoperative crises. By applying machine learning techniques to arterial pressure waveforms, artificial intelligence algorithms can predict hypotensive episodes before they occur, allowing clinicians to initiate timely therapeutic interventions such as fluid resuscitation or administration of vasoactive agents, thereby reducing adverse events. In addition, AI assisted systems support precise titration of anesthetic agents, minimizing unnecessary drug exposure CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 167 while maintaining an adequate depth of anesthesia. Postoperative Care and Recovery Beyond the operating room, artificial intelligence models assist anesthesiologists in predicting postoperative complications, including nausea and vomiting, identifying patients at risk for delayed emergence from anesthesia, and assessing postoperative pain severity using multimodal data. By analyzing patient specific demographic characteristics, physiological parameters, surgical factors, and anesthetic exposures, artificial intelligence tools support Enhanced Recovery After Surgery protocols and contribute to optimization of postoperative care pathways. This data driven approach not only improves clinical outcomes but also has the potential to reduce overall healthcare costs through more efficient resource utilization. Challenges and Implementation Barriers Despite the promising capabilities of artificial intelligence, the implementation of these systems in anesthesiology faces several significant challenges. Algorithmic opacity, often referred to as the black box phenomenon, limits the ability of clinicians to understand and explain the reasoning behind artificial intelligence recommendations. This limitation raises concerns DR. MEHRDAD FARROKHI 168 regarding clinical accountability and medico legal responsibility, particularly when AI driven recommendations diverge from established clinical practices. The risk of algorithmic bias also represents a critical consideration, especially as artificial intelligence systems are integrated with hospital information systems that may reflect existing inequities in healthcare data. Additional barriers include regulatory approval pathways, the need for standardized validation frameworks, and the requirement for clinician education and training in algorithmic literacy. Ethical and Practical Considerations Although artificial intelligence offers substantial advantages in clinical decision making, its use in anesthesia raises important ethical and practical considerations. Issues related to patient data privacy, informed consent for AI assisted care, algorithmic bias, and the need for transparency must be carefully addressed. Successful implementation of artificial intelligence systems depends on the development of transparent, fair, and accountable tools that preserve patient centered care while respecting physician patient autonomy. Ethical integration requires ongoing evaluation and governance to ensure that artificial intelligence serves the best interests of patients and clinicians alike. Conclusion CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 169 Artificial intelligence offers significant opportunities to enhance clinical decision making, reduce preventable complications, and improve patient safety throughout the perioperative period. However, the incorporation of artificial intelligence into anesthesiology requires thoughtful consideration of ethical, technical, and regulatory challenges. The future of anesthesiology depends on effective collaboration between clinicians, data scientists, and ethicists to coordinate efforts and align goals. The objective is not to replace clinical judgment, but to augment it through advanced computational intelligence. The coming decade is likely to witness the continued integration of artificial intelligence across all aspects of anesthesiology, including the development and deployment of predictive perioperative decision support tools. AI in Clinical DecisionMaking in Public Health and Preventive Medicine Background Public health and preventive medicine focus on protecting and improving the health of populations rather than treating disease at the individual level alone. These disciplines aim to prevent illness, prolong life, and promote health through organized efforts that include surveillance, screening, vaccination, health DR. MEHRDAD FARROKHI 170 education, environmental protection, and policy development. Clinical decision-making in public health and preventive medicine therefore differs from traditional bedside medicine. It requires balancing individual risk with population benefit, working with incomplete or delayed data, and making decisions that often have social, economic, and ethical implications. Artificial intelligence has emerged as a powerful tool that can transform decision-making in public health and preventive medicine. The increasing availability of large-scale data from electronic health records, registries, laboratory systems, wearable devices, environmental sensors, genomic databases, and social determinants of health has created both opportunity and complexity. Traditional analytic methods often struggle to integrate these diverse data sources or to generate timely and actionable insights. Artificial intelligence offers advanced computational approaches capable of processing large, heterogeneous datasets, identifying patterns, forecasting trends, and supporting evidencebased decisions at both the clinical and population levels. Importantly, artificial intelligence in public health is not limited to prediction or automation. Its value lies in supporting human expertise, enhancing situational awareness, reducing uncertainty, and enabling more precise and CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 171 equitable preventive strategies. When applied responsibly, artificial intelligence can strengthen disease surveillance, improve risk stratification, optimize screening programs, guide resource allocation, and support policy decisions that ultimately improve population health outcomes. Foundations of Artificial Intelligence in Public Health and Preventive Medicine Artificial intelligence refers to computational systems designed to perform tasks that typically require human intelligence, such as learning from data, recognizing patterns, reasoning, and supporting decisions. In public health and preventive medicine, the most relevant forms of artificial intelligence include machine learning, deep learning, natural language processing, and large language models. Machine learning involves algorithms that learn relationships between input variables and outcomes without explicit rule-based programming. Supervised learning uses labeled data, such as confirmed disease cases or vaccination outcomes, to train predictive models. Unsupervised learning identifies hidden structures within data, such as clustering populations by risk profiles or detecting emerging patterns in disease incidence. Reinforcement learning, though less commonly used in public DR. MEHRDAD FARROKHI 172 health practice, has potential applications in optimizing intervention strategies over time. Deep learning is a subset of machine learning that uses multi-layer neural networks to model complex, non-linear relationships. It is particularly useful for processing highdimensional data such as satellite imagery, genomic data, and time-series signals. Natural language processing allows artificial intelligence systems to extract structured information from unstructured text, including clinical notes, public health reports, social media posts, and policy documents. Large language models extend these capabilities by summarizing information, generating reports, and supporting communication, although their use requires careful oversight to ensure accuracy and ethical compliance. In public health and preventive medicine, artificial intelligence systems often integrate multiple methods and data types to reflect the complexity of population health decision-making. The effectiveness of these systems depends on data quality, representativeness, transparency, and alignment with public health goals. AI in Disease Surveillance and Early Outbreak Detection Disease surveillance is a cornerstone of public health, enabling early detection of outbreaks, CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 173 monitoring of disease trends, and evaluation of intervention effectiveness. Artificial intelligence has significantly enhanced surveillance capabilities by enabling real-time analysis of large and diverse data streams. Traditional surveillance systems often rely on delayed reporting from healthcare facilities and laboratories. Artificial intelligence complements these systems by integrating alternative data sources, such as electronic health records, pharmacy sales, laboratory orders, travel data, environmental sensors, and digital traces from social media or search queries. Machine learning models can identify abnormal patterns or deviations from baseline trends that may signal emerging outbreaks. During infectious disease outbreaks, artificial intelligence supports early warning systems that detect increases in symptom clusters, hospital admissions, or laboratory confirmations. These systems can operate at local, national, and global levels, providing public health authorities with timely insights that support rapid response. Artificial intelligence has been applied to monitor influenza, respiratory infections, vector-borne diseases, and emerging pathogens, demonstrating its potential to reduce detection delays and improve situational awareness. Beyond detection, artificial intelligence supports outbreak characterization by identifying high-risk DR. MEHRDAD FARROKHI 174 populations, predicting geographic spread, and estimating disease burden. These insights inform targeted interventions, such as vaccination campaigns, travel advisories, and resource mobilization. AI in Epidemiology and Risk Modeling Epidemiologic analysis underpins preventive medicine by identifying risk factors, estimating disease burden, and evaluating interventions. Artificial intelligence enhances epidemiologic decision-making by modeling complex relationships between biological, behavioral, environmental, and social determinants of health. Machine learning models can analyze large datasets to identify non-linear interactions and high-dimensional risk profiles that may be missed by traditional statistical methods. For example, artificial intelligence can integrate demographic data, comorbidities, environmental exposures, and behavioral factors to predict individual and population-level risk for chronic diseases such as cardiovascular disease, diabetes, and cancer. In preventive medicine, risk prediction models support personalized prevention strategies by identifying individuals or groups who may benefit most from targeted interventions. Artificial intelligence enables dynamic risk assessment that updates predictions as new data become available, CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 175 supporting adaptive preventive care. Artificial intelligence is also applied in causal inference and policy evaluation, helping to estimate the impact of interventions on population health outcomes. While these applications require careful interpretation and validation, they offer new tools for evidenceinformed decision-making. AI in Screening Programs and Preventive Interventions Screening and early detection are central to preventive medicine. Artificial intelligence supports decision-making in screening programs by improving risk stratification, optimizing screening intervals, and enhancing diagnostic accuracy. In cancer screening, artificial intelligence assists in analyzing imaging and laboratory data to identify early disease and reduce false positives and false negatives. Machine learning models can stratify populations by risk, enabling more personalized screening strategies that balance benefit and harm. Similar approaches are applied in screening for cardiovascular disease, metabolic disorders, and infectious diseases. Artificial intelligence also supports evaluation of screening program effectiveness by analyzing uptake, outcomes, and disparities. These insights inform program design and resource allocation, DR. MEHRDAD FARROKHI 176 ensuring that preventive interventions reach those who need them most. In vaccination programs, artificial intelligence assists in predicting coverage gaps, identifying populations at risk of under-immunization, and modeling the impact of different vaccination strategies. These tools support planning and evaluation of immunization campaigns, particularly in complex or resource-limited settings. AI in Environmental and Occupational Health Environmental and occupational health are key components of public health, focusing on the prevention of disease related to environmental exposures and workplace hazards. Artificial intelligence plays an increasingly important role in analyzing complex environmental data and supporting preventive decision-making. Machine learning models analyze data from air quality monitors, water systems, satellite imagery, and climate sensors to assess exposure risks and predict health impacts. These models support early warning systems for pollution-related health events, heat waves, and natural disasters. By identifying vulnerable populations and highrisk areas, artificial intelligence informs targeted interventions and policy decisions. In occupational health, artificial intelligence CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 177 28. Karuppan Perumal MK, Rajan Renuka R, Kumar Subbiah S, Manickam Natarajan P. Artificial intelligence-driven clinical decision support systems for early detection and precision therapy in oral cancer: a mini review. Front Oral Health. 2025;6:1592428. 29. Khanagar SB, Al-Ehaideb A, Vishwanathaiah S, Maganur PC, Patil S, Naik S, et al. Scope and performance of artificial intelligence technology in orthodontic diagnosis, treatment planning, and clinical decision-making - A systematic review. J Dent Sci. 2021;16(1):482-92. 30. Kindle RD, Badawi O, Celi LA, Sturland S. Intensive Care Unit Telemedicine in the Era of Big Data, Artificial Intelligence, and Computer Clinical Decision Support Systems. Crit Care Clin. 2019;35(3):483-95. 31. Knop M, Weber S, Mueller M, Niehaves B. Human Factors and Technological Characteristics Influencing the Interaction of Medical Professionals With Artificial Intelligence-Enabled Clinical Decision Support Systems: Literature Review. JMIR Hum Factors. 2022;9(1):e28639. 32. Lee TH, Chen JJ, Cheng CT, Chang CH. Does Artificial Intelligence Make Clinical Decision Better? A Review of Artificial Intelligence and Machine Learning in Acute Kidney Injury Prediction. Healthcare (Basel). 2021;9 )12(. 33. León-Domínguez U. Towards an DR. MEHRDAD FARROKHI 190 artificial intelligence clinical decision-support system based on immersive virtual reality for neurocognitive assessment. Ergonomics. 2025:1-18. 34. Li Y, Zhang T, Yang Y, Gao Y. Artificial intelligence-aided decision support in paediatrics clinical diagnosis: development and future prospects. J Int Med Res. 2020;48(9):300060520945141. 35. Liao X, Yao C, Zhang J, Liu LZ. Recent advancement in integrating artificial intelligence and information technology with real-world data for clinical decision-making in China: A scoping review. J Evid Based Med. 2023;16(4):534-46. 36. Lin X, Liang C, Liu J, Lyu T, Ghumman N, Campbell B. Artificial Intelligence-Augmented Clinical Decision Support Systems for Pregnancy Care: Systematic Review. J Med Internet Res. 2024;26:e54737. 37. Magrabi F, Ammenwerth E, McNair JB, De Keizer NF, Hyppönen H, Nykänen P, et al. Artificial Intelligence in Clinical Decision Support: Challenges for Evaluating AI and Practical Implications. Yearb Med Inform. 2019;28(1):128-34. 38. Mahadevaiah G, Rv P, Bermejo I, Jaffray D, Dekker A, Wee L. Artificial intelligence-based clinical decision support in modern medical physics: Selection, acceptance, commissioning, CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 191 and quality assurance. Med Phys. 2020;47(5):e228e35. 39. Mikkonen K, Tuunainen S, Oikarinen A, Jansson M, Woo B, Zhou W, et al. Artificial Intelligence Technologies Supporting Nurses' Clinical Decision-Making: A Systematic Review. J Clin Nurs. 2025. 40. Montani S, Striani M. Artificial Intelligence in Clinical Decision Support: a Focused Literature Survey. Yearb Med Inform. 2019;28(1):120-7. 41. Montomoli J, Hilty MP, Ince C. Artificial intelligence in intensive care: moving towards clinical decision support systems. Minerva Anestesiol. 2022;88(12):1066-72. 42. Naderian S, Soleimanzadeh F, Nikniaz L, Sanaie S, Sadeghi-Ghyassi F, Samad-Soltani T. A Systematic Review of Artificial Intelligence-Based Clinical Decision Support Systems in Prostate Cancer Management. Healthc Technol Lett. 2025;12(1):e70026. 43. Nimri R, Phillip M. Enhancing Care in Type 1 Diabetes with Artificial Intelligence Driven Clinical Decision Support Systems. Horm Res Paediatr. 2025;98(4):384-95. 44. Oei SP, Bakkes T, Mischi M, Bouwman RA, van Sloun RJG, Turco S. Artificial intelligence in clinical decision support and the prediction of adverse events. Front Digit Health. 2025;7:1403047. DR. MEHRDAD FARROKHI 192 45. Ogut E. Artificial Intelligence in Clinical Medicine: Challenges Across Diagnostic Imaging, Clinical Decision Support, Surgery, Pathology, and Drug Discovery. Clin Pract. 2025;15 )9(. 46. Ouanes K, Farhah N. Effectiveness of Artificial Intelligence (AI) in Clinical Decision Support Systems and Care Delivery. J Med Syst. 2024;48(1):74. 47. Parsons CS, Zuiderwijk A, Orchard NA, Oosterhoff JHF, de Reuver M. Task-Technology Fit of Artificial Intelligence-based clinical decision support systems: a review of qualitative studies. BMC Med Inform Decis Mak. 2025;25(1):397. 48. Patel M, Nanji KC. Artificial Intelligence in Perioperative Medication-Related Clinical Decision Support. Anesthesiol Clin. 2025;43(3):587-602. 49. Pedersen M, Verspoor K, Jenkinson M, Law M, Abbott DF, Jackson GD. Artificial intelligence for clinical decision support in neurology. Brain Commun. 2020;2(2):fcaa096. 50. Peek N, Capurro D, Rozova V, van der Veer SN. Bridging the Gap: Challenges and Strategies for the Implementation of Artificial Intelligencebased Clinical Decision Support Systems in Clinical Practice. Yearb Med Inform. 2024;33(1):103-14. 51. Ramgopal S, Sanchez-Pinto LN, Horvat CM, Carroll MS, Luo Y, Florin TA. Artificial intelligencebased clinical decision support in pediatrics. Pediatr Res. 2023;93(2):334-41. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 193 52. Reicher L, Lutsker G, Michaan N, Grisaru D, Laskov I. Exploring the role of artificial intelligence, large language models: Comparing patient-focused information and clinical decision support capabilities to the gynecologic oncology guidelines. Int J Gynaecol Obstet. 2025;168(2):419-27. 53. Ren SQ, Chen JM, Cai C. Translational artificial intelligence in gastrointestinal and hepatic disorders: Advancing intelligent clinical decision-making for diagnosis, treatment, and prognosis. World J Gastroenterol. 2025;31(36):110742. 54. Ryan S, Heaney-Huls K, Kawamoto K, Lobach D, Desai PJ, CdsiC Implementation A, et al. Clinical Decision Support Innovation Collaborative (CDSiC) Reports. Artificial Intelligence-Supported Patient-Centered Clinical Decision Support: A Summary of Considerations. Rockville (MD): Agency for Healthcare Research and Quality (US); 2025. 55. Sáez C, Ferri P, García-Gómez JM. Resilient Artificial Intelligence in Health: Synthesis and Research Agenda Toward Next-Generation Trustworthy Clinical Decision Support. J Med Internet Res. 2024;26:e50295. 56. Safarian A, Mirshahvalad SA, Nasrollahi H, Jung T, Pirich C, Arabi H, et al. Impact of [(18)F]FDG PET/CT Radiomics and Artificial Intelligence in DR. MEHRDAD FARROKHI 194 Clinical Decision Making in Lung Cancer: Its Current Role. Semin Nucl Med. 2025;55(2):156-66. 57. Seely AJE, Newman K, Ramchandani R, Herry C, Scales N, Hudek N, et al. Roadmap for the evolution of monitoring: developing and evaluating waveform-based variability-derived artificial intelligence-powered predictive clinical decision support software tools. Crit Care. 2024;28(1):404. 58. Semerci ZM, Yardımcı S. Empowering Modern Dentistry: The Impact of Artificial Intelligence on Patient Care and Clinical Decision Making. Diagnostics (Basel). 2024;14 )12(. 59. Sexton DJ, Judge C. Assessments of Generative Artificial Intelligence as Clinical Decision Support Ought to be Incorporated Into Randomized Controlled Trials of Electronic Alerts for Acute Kidney Injury. Mayo Clin Proc Digit Health. 2024;2(4):606-10. 60. Shaikh F, Dehmeshki J, Bisdas S, RoettgerDupont D, Kubassova O, Aziz M, et al. Artificial Intelligence-Based Clinical Decision Support Systems Using Advanced Medical Imaging and Radiomics. Curr Probl Diagn Radiol. 2021;50(2):262-7. 61. Singla B, Afridi S, Vayolipoyil S, Ahmed T, Afzaal S, Saleem K, et al. The Evolving Role of Artificial Intelligence in Medical Science: Advancing Diagnostics, Clinical Decision-Making, CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 195 and Research. Cureus. 2025;17(9):e91514. 62. Sokol K, Fackler J, Vogt JE. Artificial intelligence should genuinely support clinical reasoning and decision making to bridge the translational gap. NPJ Digit Med. 2025;8(1):345. 63. Sperti M, Cardaci C, Bruno F, Shah STH, Panagiotopoulos K, Kassem K, et al. Artificial Intelligence-based Approaches for Characterizing Plaque Components From Intravascular Optical Coherence Tomography Imaging: Integration Into Clinical Decision Support Systems. Rev Cardiovasc Med. 2025;26(7):39210. 64. Teoman AS, Serefoglu EC. Artificial Intelligence-Based Clinical Decision-Making in Erectile Dysfunction: a Narrative Review. Curr Urol Rep. 2024;26(1):22. 65. Tun HM, Rahman HA, Naing L, Malik OA. Trust in Artificial Intelligence-Based Clinical Decision Support Systems Among Health Care Workers: Systematic Review. J Med Internet Res. 2025;27:e69678. 66. van der Ven WH, Veelo DP, Wijnberge M, van der Ster BJP, Vlaar APJ, Geerts BF. One of the first validations of an artificial intelligence algorithm for clinical use: The impact on intraoperative hypotension prediction and clinical decision-making. Surgery. 2021;169(6):1300-3. 67. Vasey B, Nagendran M, Campbell B, Clifton DA, Collins GS, Denaxas S, et al. DR. MEHRDAD FARROKHI 196 Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. Nat Med. 2022;28(5):924-33. 68. Wang L, Chen X, Zhang L, Li L, Huang Y, Sun Y, et al. Artificial intelligence in clinical decision support systems for oncology. Int J Med Sci. 2023;20(1):79-86. 69. Weissman GE. Evaluation and Regulation of Artificial Intelligence Medical Devices for Clinical Decision Support. Annu Rev Biomed Data Sci. 2025;8(1):81-99. 70. Wu M, Du X, Gu R, Wei J. Artificial Intelligence for Clinical Decision Support in Sepsis. Front Med (Lausanne). 2021;8:665464. 71. Xu Q, Xie W, Liao B, Hu C, Qin L, Yang Z, et al. Interpretability of Clinical Decision Support Systems Based on Artificial Intelligence from Technological and Medical Perspective: A Systematic Review. J Healthc Eng. 2023;2023:9919269. 72. Yan D, Zheng Q, Chang K, Hua R, Liu Y, Xue J, et al. Artificial intelligence in traditional Chinese medicine: from systems biological mechanism discovery, real-world clinical evidence inference to personalized clinical decision support. Chin J Nat Med. 2025;23(11):1310-28. 73. Yeo M, Kok HK, Kutaiba N, Maingard J, Thijs V, Tahayori B, et al. Artificial intelligence in CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 197 clinical decision support and outcome prediction - applications in stroke. J Med Imaging Radiat Oncol. 2021. 74. Zeng J, Shufean MA. Molecular-based precision oncology clinical decision making augmented by artificial intelligence. Emerg Top Life Sci. 2021;5(6):757-64. 75. Zhang C, Lam BD, Lucas F, Foy BH. Machine Learning and Artificial Intelligence-Based Clinical Decision Support for Modern Hematology. Clin Lab Med. 2025;45(4):691-705. 76. Abbasgholizadeh Rahimi S, Cwintal M, Huang Y, Ghadiri P, Grad R, Poenaru D, et al. Application of Artificial Intelligence in Shared Decision Making: Scoping Review. JMIR Med Inform. 2022;10(8):e36199. 77. Abdekhoda M, Madiseh FR. Artificial Intelligence Applications in Decision-Making for Disease Management: A scoping review. Sultan Qaboos Univ Med J. 2025;25(1):441-9. 78. Ajmal CS, Yerram S, Abishek V, Nizam VPM, Aglave G, Patnam JD, et al. Innovative Approaches in Regulatory Affairs: Leveraging Artificial Intelligence and Machine Learning for Efficient Compliance and Decision-Making. Aaps j. 2025;27(1):22. 79. Al Fryan LH, Shomo MI, Alazzam MB, Rahman MA. Processing Decision Tree Data Using Internet of Things (IoT) and Artificial Intelligence DR. MEHRDAD FARROKHI 198 Technologies with Special Reference to Medical Application. Biomed Res Int. 2022;2022:8626234. 80. Alenezi AM. Artificial Intelligence in Breast Cancer Diagnosis and Surgical Decision-Making: An Updated and Comprehensive Overview of Precision and Personalization in Current Evidence. Cancer Manag Res. 2025;17:2261-75. 81. Authors, Xie W, Butcher R. CADTH Horizon Scans. Artificial Intelligence Decision Support Tools for End-of-Life Care Planning Conversations: CADTH Horizon Scan. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2023. 82. Birla M, Rajan, Roy PG, Gupta I, Malik PS. Integrating Artificial IntelligenceDriven Wearable Technology in Oncology Decision-Making: A Narrative Review. Oncology. 2025;103(1):69-82. 83. Bivard A, Churilov L, Parsons M. Artificial intelligence for decision support in acute stroke - current roles and potential. Nat Rev Neurol. 2020;16(10):575-85. 84. Boreak N. Effectiveness of Artificial Intelligence Applications Designed for Endodontic Diagnosis, Decision-making, and Prediction of Prognosis: A Systematic Review. J Contemp Dent Pract. 2020;21(8):926-34. 85. Byerly S, Maurer LR, Mantero A, Naar L, An G, Kaafarani HMA. Machine Learning and Artificial CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 199 Intelligence for Surgical Decision Making. Surg Infect (Larchmt). 2021;22(6):626-34. 86. Caruso PF, Greco M, Ebm C, Angelotti G, Cecconi M. Implementing Artificial Intelligence: Assessing the Cost and Benefits of Algorithmic Decision-Making in Critical Care. Crit Care Clin. 2023;39(4):783-93. 87. Černevičienė J, Kabašinskas A. Review of Multi-Criteria Decision-Making Methods in Finance Using Explainable Artificial Intelligence. Front Artif Intell. 2022;5:827584. 88. Coelho H, Silva F, Correia M, Rodrigues PM. Artificial Intelligence in Patient Blood Management: A Systematic Review of Predictive, Diagnostic, and Decision Support Applications. J Clin Med. 2025;14 )23(. 89. Conte L, Decembrino N, Arribas C, Cucci F, De Nunzio G, Amodeo I, et al. Leveraging Artificial Intelligence for decision support in neonatal and pediatric pharmacotherapy: A scoping review. Semin Fetal Neonatal Med. 2025:101691. 90. Contreras I, Vehi J. Artificial Intelligence for Diabetes Management and Decision Support: Literature Review. J Med Internet Res. 2018;20(5):e10775. 91. Cresswell K, Callaghan M, Khan S, Sheikh Z, Mozaffar H, Sheikh A. Investigating the use of data-driven artificial intelligence in computerised decision support systems for health and social DR. MEHRDAD FARROKHI 200 care: A systematic review. Health Informatics J. 2020;26(3):2138-47. 92. Di Palma G, Scendoni R, De Benedictis A, Tambone V, De Micco F. Leveraging artificial intelligence for collaborative care planning: Innovations and impacts in shared decisionmaking - A systematic review. Open Med (Wars). 2025;20(1):20251232. 93. Ding Z, Fang W, Zhang J, Fang C, Sun Y. Artificial intelligence in wearable biosensing: Enhancing data analysis and decision-making. Prog Mol Biol Transl Sci. 2025;216:1-26. 94. Dunn N, Verma N, Dunn W. Artificial Intelligence for Predictive Diagnostics, Prognosis, and Decision Support in MASLD, Hepatocellular Carcinoma, and Digital Pathology. J Clin Exp Hepatol. 2026;16(1):103184. 95. Duran HT, Kingeter M, Reale C, Weinger MB, Salwei ME. Decision-making in anesthesiology: will artificial intelligence make intraoperative care safer? Curr Opin Anaesthesiol. 2023;36(6):691-7. 96. Egger K, Rijntjes M. [Big data and artificial intelligence for diagnostic decision support in atypical dementia]. Nervenarzt. 2018;89(8):875-84. 97. El-Kareh R, Sittig DF. Enhancing Diagnosis Through Technology: Decision Support, Artificial Intelligence, and Beyond. Crit Care Clin. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 201 2022;38(1):129-39. 98. Evangelista K, de Freitas Silva BS, Yamamoto-Silva FP, Valladares-Neto J, Silva MAG, Cevidanes LHS, et al. Accuracy of artificial intelligence for tooth extraction decision-making in orthodontics: a systematic review and meta-analysis. Clin Oral Investig. 2022;26(12):6893-905. 99. Fast NJ, Schroeder J. Power and decision making: new directions for research in the age of artificial intelligence. Curr Opin Psychol. 2020;33:172-6. 100. Fawaz A, Ferraresi A, Isidoro C. Systems Biology in Cancer Diagnosis Integrating Omics Technologies and Artificial Intelligence to Support Physician Decision Making. J Pers Med. 2023;13 )11(. 101. Froicu EM, Creangă-Murariu I, Afrăsânie VA, Gafton B, Alexa-Stratulat T, Miron L, et al. Artificial Intelligence and Decision-Making in Oncology: A Review of Ethical, Legal, and Informed Consent Challenges. Curr Oncol Rep. 2025;27(8):1002-12. 102. Giaccone P, D'Antoni F, Russo F, Ambrosio L, Papalia GF, d'Angelis O, et al. Prevention and management of degenerative lumbar spine disorders through artificial intelligence-based decision support systems: a systematic review. BMC Musculoskelet Disord. 2025;26(1):126. DR. MEHRDAD FARROKHI 202 103. Giacobbe DR, Vena A, Bassetti M. Role of artificial intelligence in ICU therapeutic decisionmaking for severe infections. Curr Opin Crit Care. 2025;31(5):547-53. 104. Gupta P, Pearce AK, Pham T, Miller M, Brunetti K, Heskett K, et al. Artificial intelligencedriven decision support for patients with acute respiratory failure: a scoping review. Intensive Care Med Exp. 2025;13(1):83. 105. Gurupur V, Wan TTH. Inherent Bias in Artificial Intelligence-Based Decision Support Systems for Healthcare. Medicina (Kaunas). 2020;56 )3(. 106. Higgins O, Short BL, Chalup SK, Wilson RL. Artificial intelligence (AI) and machine learning (ML) based decision support systems in mental health: An integrative review. Int J Ment Health Nurs. 2023;32(4):966-78. 107. Higgins O, Wilson RL. Integrating Artificial Intelligence (AI) With Workforce Solutions for Sustainable Care: A Follow Up to Artificial Intelligence and Machine Learning (ML) Based Decision Support Systems in Mental Health. Int J Ment Health Nurs. 2025;34(2):e70019. 108. Holz FG, Abreu-Gonzalez R, Bandello F, Duval R, O'Toole L, Pauleikhoff D, et al. Does realtime artificial intelligence-based visual pathology enhancement of three-dimensional optical coherence tomography scans optimise treatment CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 203 decision in patients with nAMD? Rationale and design of the RAZORBILL study. Br J Ophthalmol. 2023;107(1):96-101. 109. Hu M, Wang Y, Liu Y, Cai B, Kong F, Zheng Q, et al. Artificial Intelligence in Nursing DecisionMaking: A Bibliometric Analysis of Trends and Impacts. Nurs Rep. 2025;15 )6(. 110. Kahn CE, Jr. Artificial intelligence in radiology: decision support systems. Radiographics. 1994;14(4):849-61. 111. Khosravi M, Zare Z, Mojtabaeian SM, Izadi R. Artificial Intelligence and Decision-Making in Healthcare: A Thematic Analysis of a Systematic Review of Reviews. Health Serv Res Manag Epidemiol. 2024;11:23333928241234863. 112. Kumar AA, Vasudevan C. Artificial intelligence for medical decision making. J Assoc Physicians India. 1990;38(7):475-8. 113. Lareyre F, Yeung KK, Guzzi L, Di Lorenzo G, Chaudhuri A, Behrendt CA, et al. Artificial intelligence in vascular surgical decision making. Semin Vasc Surg. 2023;36(3):448-53. 114. Li J, Wu J, Zhao Z, Zhang Q, Shao J, Wang C, et al. Artificial intelligence-assisted decision making for prognosis and drug efficacy prediction in lung cancer patients: a narrative review. J Thorac Dis. 2021;13(12):7021-33. 115. Li Y, Chen D, Wu X, Yang W, Chen Y. A DR. MEHRDAD FARROKHI 204 narrative review of artificial intelligence-assisted histopathologic diagnosis and decision-making for non-small cell lung cancer: achievements and limitations. J Thorac Dis. 2021;13(12):7006-20. 116. Loftus TJ, Shickel B, Ozrazgat-Baslanti T, Ren Y, Glicksberg BS, Cao J, et al. Artificial intelligence-enabled decision support in nephrology. Nat Rev Nephrol. 2022;18(7):452-65. 117. Loftus TJ, Tighe PJ, Filiberto AC, Efron PA, Brakenridge SC, Mohr AM, et al. Artificial Intelligence and Surgical Decision-making. JAMA Surg. 2020;155(2):148-58. 118. Loushy I, Sperling MR. Artificial intelligence for epilepsy decision support. Epilepsia. 2025. 119. Lynn LA. Artificial intelligence systems for complex decision-making in acute care medicine: a review. Patient Saf Surg. 2019;13:6. 120. Manava P, Galster M, Heinen H, Stebner A, Lell M. [Artificial intelligencebased algorithms : Decision-making support for computed tomography of the chest]. Radiologe. 2020;60(10):952-8. 121. Marques M, Almeida A, Pereira H. The Medicine Revolution Through Artificial Intelligence: Ethical Challenges of Machine Learning Algorithms in Decision-Making. Cureus. 2024;16(9):e69405. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 205 122. Massalha S, Clarkin O, Thornhill R, Wells G, Chow BJW. Decision Support Tools, Systems, and Artificial Intelligence in Cardiac Imaging. Can J Cardiol. 2018;34(7):827-38. 123. McNair D. Artificial Intelligence and Machine Learning for Lead-to-Candidate DecisionMaking and Beyond. Annu Rev Pharmacol Toxicol. 2023;63:77-97. 124. Navarrete-Welton AJ, Hashimoto DA. Current applications of artificial intelligence for intraoperative decision support in surgery. Front Med. 2020;14(4):369-81. 125. Nzeako TR, Elendu C, Echefu G, Olanisa O, Kiladejo A, Bob-Manuel ED. Artificial intelligence in interventional cardiology: a review of its role in diagnosis, decision-making, and procedural precision. Ann Med Surg (Lond). 2025;87(9):5720-34. 126. Oehring R, Ramasetti N, Ng S, Roller R, Thomas P, Winter A, et al. Use and accuracy of decision support systems using artificial intelligence for tumor diseases: a systematic review and meta-analysis. Front Oncol. 2023;13:1224347. 127. Orzan F, Iancu Ş D, Dioşan L, Bálint Z. Textural analysis and artificial intelligence as decision support tools in the diagnosis of multiple sclerosis - a systematic review. Front Neurosci. 2024;18:1457420. DR. MEHRDAD FARROKHI 206 128. Pinton P. Impact of artificial intelligence on prognosis, shared decision-making, and precision medicine for patients with inflammatory bowel disease: a perspective and expert opinion. Ann Med. 2023;55(2):2300670. 129. Pozza A, Zanella L, Castaldi B, Di Salvo G. How Will Artificial Intelligence Shape the Future of Decision-Making in Congenital Heart Disease? J Clin Med. 2024;13 )10(. 130. Rasheed J, Jamil A, Hameed AA, Aftab U, Aftab J, Shah SA, et al. A survey on artificial intelligence approaches in supporting frontline workers and decision makers for the COVID-19 pandemic. Chaos Solitons Fractals. 2020;141:110337. 131. Reifs Jiménez D, Casanova-Lozano L, GrauCarrión S, Reig-Bolaño R. Artificial Intelligence Methods for Diagnostic and Decision-Making Assistance in Chronic Wounds: A Systematic Review. J Med Syst. 2025;49(1):29. 132. Ruiz NI, Cardona Salazar I, Naranjo Palacio LX, Agudelo Agudelo C, Ledesma Parra AM, Flores Rodriguez JC. Accuracy and Reliability of Artificial Intelligence in Surgical Decision-Making: A Literature Review. Cureus. 2025;17(10):e95337. 133. Saravi B, Hassel F, Ülkümen S, Zink A, Shavlokhova V, Couillard-Despres S, et al. Artificial Intelligence-Driven Prediction Modeling and Decision Making in Spine Surgery Using Hybrid CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 207 Machine Learning Models. J Pers Med. 2022;12 )4(. 134. Sardar P, Abbott JD, Kundu A, Aronow HD, Granada JF, Giri J. Impact of Artificial Intelligence on Interventional Cardiology: From Decision-Making Aid to Advanced Interventional Procedure Assistance. JACC Cardiovasc Interv. 2019;12(14):1293-303. 135. Sellin J, Pantel JT, Börsch N, Conrad R, Mücke M. [Short paths to diagnosis with artificial intelligence: systematic literature review on diagnostic decision support systems]. Schmerz. 2024;38(1):19-27. 136. Senghor AS, Bright TJ, Kakim S, Norris KC, Antwi HA, Cooper JK, et al. A communitybased approach to ethical decision-making in artificial intelligence for health care. JAMIA Open. 2025;8(4):ooaf076. 137. Sengul T, Sariköse S, Gul A. Ethical decision-making and artificial intelligence in nursing education: An integrative review. Nurs Ethics. 2025;32(8):2490-515. 138. Shah SP, Heiss JD. Artificial Intelligence as A Complementary Tool for Clincal DecisionMaking in Stroke and Epilepsy. Brain Sci. 2024;14 )3(. 139. Taber P, Armin JS, Orozco G, Del Fiol G, Erdrich J, Kawamoto K, et al. Artificial Intelligence and Cancer Control: Toward Prioritizing Justice, Equity, Diversity, and Inclusion (JEDI) in Emerging DR. MEHRDAD FARROKHI 208 Decision Support Technologies. Curr Oncol Rep. 2023;25(5):387-424. 140. Telecan T, Andras I, Crisan N, Giurgiu L, Căta ED, Caraiani C, et al. More than Meets the Eye: Using Textural Analysis and Artificial Intelligence as Decision Support Tools in Prostate Cancer Diagnosis-A Systematic Review. J Pers Med. 2022;12 )6(. 141. Threlkeld R, Ashiku L, Canfield C, Shank DB, Schnitzler MA, Lentine KL, et al. Reducing Kidney Discard With Artificial Intelligence Decision Support: the Need for a Transdisciplinary Systems Approach. Curr Transplant Rep. 2021;8(4):263-71. 142. Toffaha KM, Simsekler MCE, Omar MA. Leveraging artificial intelligence and decision support systems in hospital-acquired pressure injuries prediction: A comprehensive review. Artif Intell Med. 2023;141:102560. 143. Tyler NS, Jacobs PG. Artificial Intelligence in Decision Support Systems for Type 1 Diabetes. Sensors (Basel). 2020;20 )11(. 144. Zarkowsky DS, Stonko DP. Artificial intelligence's role in vascular surgery decisionmaking. Semin Vasc Surg. 2021;34(4):260-7. 145. Aamir A, Jamil Y, Bilal M, Diwan M, Nashwan AJ, Ullah I. Artificial Intelligence in Enhancing Syncope Management - An Update. Curr Probl Cardiol. 2024;49(1 Pt B):102079. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 209 220. Gorincour G, Monneuse O, Ben Cheikh A, Avondo J, Chaillot PF, Journe C, et al. Management of abdominal emergencies in adults using telemedicine and artificial intelligence. J Visc Surg. 2021;158(3s):S26-s31. 221. Gorris M, Hoogenboom SA, Wallace MB, van Hooft JE. Artificial intelligence for the management of pancreatic diseases. Dig Endosc. 2021;33(2):231-41. 222. Gou C, Zafar S, Hasnain Z, Aslam N, Iqbal N, Abbas S, et al. Machine and Deep Learning: Artificial Intelligence Application in Biotic and Abiotic Stress Management in Plants. Front Biosci (Landmark Ed). 2024;29(1):20. 223. Granata V, Fusco R, De Muzio F, Cutolo C, Grassi F, Brunese MC, et al. Risk Assessment and Cholangiocarcinoma: Diagnostic Management and Artificial Intelligence. Biology (Basel). 2023;12 )2(. 224. Granata V, Fusco R, Setola SV, Galdiero R, Maggialetti N, Silvestro L, et al. Risk Assessment and Pancreatic Cancer: Diagnostic Management and Artificial Intelligence. Cancers (Basel). 2023;15 )2(. 225. Gruson D, Dabla P, Stankovic S, Homsak E, Gouget B, Bernardini S, et al. Artificial intelligence and thyroid disease management: considerations for thyroid function tests. Biochem Med (Zagreb). 2022;32(2):020601. DR. MEHRDAD FARROKHI 222 226. Guan S, Liu D, Zhang Q. [Pediatric oral maxillofacial management and artificial intelligence]. Lin Chuang Er Bi Yan Hou Tou Jing Wai Ke Za Zhi. 2023;37(8):658-61. 227. Guan Z, Li H, Liu R, Cai C, Liu Y, Li J, et al. Artificial intelligence in diabetes management: Advancements, opportunities, and challenges. Cell Rep Med. 2023;4(10):101213. 228. Guerrisi A, Falcone I, Valenti F, Rao M, Gallo E, Ungania S, et al. Artificial Intelligence and Advanced Melanoma: Treatment Management Implications. Cells. 2022;11 )24(. 229. Gunasekeran DV, Ting DSW, Tan GSW, Wong TY. Artificial intelligence for diabetic retinopathy screening, prediction and management. Curr Opin Ophthalmol. 2020;31(5):357-65. 230. Guo W, Lv C, Guo M, Zhao Q, Yin X, Zhang L. Innovative applications of artificial intelligence in zoonotic disease management. Sci One Health. 2023;2:100045. 231. Gupta NS, Kumar P. Perspective of artificial intelligence in healthcare data management: A journey towards precision medicine. Comput Biol Med. 2023;162:107051. 232. Gutierrez L, Lim JS, Foo LL, Ng WY, Yip M, Lim GYS, et al. Application of artificial intelligence in cataract management: current and future directions. Eye Vis (Lond). 2022;9(1):3. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 223 233. Hameed BMZ, AVL SD, Raza SZ, Karimi H, Khanuja HS, Shetty DK, et al. Artificial Intelligence and Its Impact on Urological Diseases and Management: A Comprehensive Review of the Literature. J Clin Med. 2021;10 )9(. 234. Haverkamp W, Strodthoff N. [Artificial intelligence-enhanced electrocardiography : Will it revolutionize diagnosis and management of our patients?]. Herzschrittmacherther Elektrophysiol. 2024;35(2):104-10. 235. Hegazi M, Taverna G, Grizzi F. Is Artificial Intelligence the Key to Revolutionizing Benign Prostatic Hyperplasia Diagnosis and Management? Arch Esp Urol. 2023;76(9):643-56. 236. Hu Q, Li K, Yang C, Wang Y, Huang R, Gu M, et al. The role of artificial intelligence based on PET/CT radiomics in NSCLC: Disease management, opportunities, and challenges. Front Oncol. 2023;13:1133164. 237. Huang L, Huhulea EN, Abraham E, Bienenstock R, Aifuwa E, Hirani R, et al. The Role of Artificial Intelligence in Obesity Risk Prediction and Management: Approaches, Insights, and Recommendations. Medicina (Kaunas). 2025;61 )2(. 238. Isaksen JL, Baumert M, Hermans ANL, Maleckar M, Linz D. Artificial intelligence for the detection, prediction, and management of atrial fibrillation. Herzschrittmacherther DR. MEHRDAD FARROKHI 224 Elektrophysiol. 2022;33(1):34-41. 239. Issa IA, Youssef O, Issa T. Can artificial intelligence improve the diagnosis and management of patients with eosinophilic esophagitis? World J Gastroenterol. 2025;31(38):110999. 240. Ittoop SM, Jaccard N, Lanouette G, Kahook MY. The Role of Artificial Intelligence in the Diagnosis and Management of Glaucoma. J Glaucoma. 2022;31(3):137-46. 241. Ivanova S, Kuznetsov A, Zverev R, Rada A. Artificial Intelligence Methods for the Construction and Management of Buildings. Sensors (Basel). 2023;23 )21(. 242. Jacob M, Reddy RP, Garcia RI, Reddy AP, Khemka S, Roghani AK, et al. Harnessing Artificial Intelligence for the Detection and Management of Colorectal Cancer Treatment. Cancer Prev Res (Phila). 2024;17(11):499-515. 243. Jin K, Grzybowski A. Advancements in artificial intelligence for the diagnosis and management of anterior segment diseases. Curr Opin Ophthalmol. 2025;36(4):335-42. 244. Katebi M, Bahreini M, Bagherzadeh R, Pouladi S. Artificial Intelligence and Nursing Management: Opportunities, Challenges, and Ethical Considerations-A Scoping Review. J Nurs Manag. 2025;2025:2797535. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 225 245. Kaur I, Behl T, Aleya L, Rahman H, Kumar A, Arora S, et al. Artificial intelligence as a fundamental tool in management of infectious diseases and its current implementation in COVID-19 pandemic. Environ Sci Pollut Res Int. 2021;28(30):40515-32. 246. Kaushik AK, Dhau JS, Gohel H, Mishra YK, Kateb B, Kim NY, et al. Electrochemical SARS-CoV-2 Sensing at Point-ofCare and Artificial Intelligence for Intelligent COVID-19 Management. ACS Appl Bio Mater. 2020;3(11):7306-25. 247. Khalafi P, Morsali S, Hamidi S, Ashayeri H, Sobhi N, Pedrammehr S, et al. Artificial intelligence in stroke risk assessment and management via retinal imaging. Front Comput Neurosci. 2025;19:1490603. 248. Khan MJ, Karmakar A. Emerging Robotic Innovations and Artificial Intelligence in Endotracheal Intubation and Airway Management: Current State of the Art. Cureus. 2023;15(7):e42625. 249. Kim EN, Gowin K, Reb A, Sandhu D, Veguilla E, Zachariah F, et al. Artificial Intelligence in Supportive Oncology and Symptom Management Opportunities. Cancer J. 2025;31 )6(. 250. Kiwanuka F, Stevanin S, Ahtisham Y, Owusu B, Nurmeksela A, Kvist T. Nurse Leadership and Artificial Intelligence Integration in Nursing DR. MEHRDAD FARROKHI 226 Workforce Management: A Scoping Review. J Adv Nurs. 2025. 251. Kumar V, Gaddam M, Moustafa A, Iqbal R, Gala D, Shah M, et al. The Utility of Artificial Intelligence in the Diagnosis and Management of Pancreatic Cancer. Cureus. 2023;15(11):e49560. 252. Lan L, Sun W, Xu D, Yu M, Xiao F, Hu H, et al. Artificial intelligence-based approaches for COVID-19 patient management. Intell Med. 2021;1(1):10-5. 253. Levy-Mendelovich S, Glicksberg BS, Soffer S, Gendler M, Efros O, Klang E. Artificial Intelligence in Hemophilia Management: Revolutionizing Patient Care and Future Directions. Acta Haematol. 2025;148(5):546-55. 254. Li J, Huang J, Zheng L, Li X. Application of Artificial Intelligence in Diabetes Education and Management: Present Status and Promising Prospect. Front Public Health. 2020;8:173. 255. Li S, Yue R, Lu S, Luo J, Wu X, Zhang Z, et al. Artificial intelligence and machine learning in acute respiratory distress syndrome management: recent advances. Front Med (Lausanne). 2025;12:1597556. 256. Lim K, Heo TY, Yun J. Trends in the Approval and Quality Management of Artificial Intelligence Medical Devices in the Republic of Korea. Diagnostics (Basel). 2022;12 )2(. CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 227 257. Liu J, Liu Z, Liu C, Sun H, Li X, Yang Y. Integrating Artificial Intelligence in the Diagnosis and Management of Metabolic Syndrome: A Comprehensive Review. Diabetes Metab Res Rev. 2025;41(4):e70039. 258. Lococo F, Ghaly G, Flamini S, Campanella A, Chiappetta M, Bria E, et al. Artificial intelligence applications in personalizing lung cancer management: state of the art and future perspectives. J Thorac Dis. 2024;16(10):7096-110. 259. Lu MY, Chuang WL, Yu ML. The role of artificial intelligence in the management of liver diseases. Kaohsiung J Med Sci. 2024;40(11):962-71. 260. Luo J, Pan M, Mo K, Mao Y, Zou D. Emerging role of artificial intelligence in diagnosis, classification and clinical management of glioma. Semin Cancer Biol. 2023;91:110-23. 261. Mahdavi S, Anthony NM, Sikaneta T, Tam PY. Perspective: Multiomics and Artificial Intelligence for Personalized Nutritional Management of Diabetes in Patients Undergoing Peritoneal Dialysis. Adv Nutr. 2025;16(3):100378. 262. Maroju RG, Choudhari SG, Shaikh MK, Borkar SK, Mendhe H. Application of Artificial Intelligence in the Management of Drinking Water: A Narrative Review. Cureus. 2023;15(11):e49344. 263. Mateus N, Abade E, Coutinho D, Gómez DR. MEHRDAD FARROKHI 228 M, Peñas CL, Sampaio J. Empowering the Sports Scientist with Artificial Intelligence in Training, Performance, and Health Management. Sensors (Basel). 2024;25 )1(. 264. Maurya R, Chug I, Vudatha V, Palma AM. Applications of spatial transcriptomics and artificial intelligence to develop integrated management of pancreatic cancer. Adv Cancer Res. 2024;163:107-36. 265. Mayro EL, Wang M, Elze T, Pasquale LR. The impact of artificial intelligence in the diagnosis and management of glaucoma. Eye (Lond). 2020;34(1):1-11. 266. Medeiros HJS, Dabbagh A, Vlassakov K, Sabouri AS. Artificial Intelligence in Regional Anesthesia and Pain Management. Anesthesiol Clin. 2025;43(3):491-505. 267. Meier JM, Tschoellitsch T. Artificial Intelligence and Machine Learning in Patient Blood Management: A Scoping Review. Anesth Analg. 2022;135(3):524-31. 268. Mercurio M, Denami F, Vescio A, Familiari F, Longo UG, Galasso O, et al. Artificial Intelligence for the Diagnosis and Management of Patellofemoral Instability: A Comprehensive Review. Diagnostics (Basel). 2025;15 )22(. 269. Mina A. Big data and artificial intelligence in future patient management. How is it all started? Where are we at now? Quo tendimus? Adv CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 229 Lab Med. 2020;1(3):20200014. 270. Młynarska E, Bojdo K, Frankenstein H, Kustosik N, Mstowska W, Przybylak A, et al. Nanotechnology and Artificial Intelligence in Dyslipidemia Management-Cardiovascular Disease: Advances, Challenges, and Future Perspectives. J Clin Med. 2025;14 )3(. 271. Muzammil MA, Javid S, Afridi AK, Siddineni R, Shahabi M, Haseeb M, et al. Artificial intelligence-enhanced electrocardiography for accurate diagnosis and management of cardiovascular diseases. J Electrocardiol. 2024;83:30-40. 272. Nagarajan VD, Lee SL, Robertus JL, Nienaber CA, Trayanova NA, Ernst S. Artificial intelligence in the diagnosis and management of arrhythmias. Eur Heart J. 2021;42(38):3904-16. 273. Naik N, Roth B, Lundy SD. Artificial Intelligence for Clinical Management of Male Infertility, a Scoping Review. Curr Urol Rep. 2024;26(1):17. 274. Naser AM, Vyas R, Morgan AA, Kalaiger AM, Kharawala A, Nagraj S, et al. Role of Artificial Intelligence in the Diagnosis and Management of Pulmonary Embolism: A Comprehensive Review. Diagnostics (Basel). 2025;15 )7(. 275. Natekar A, Cohen F. Artificial Intelligence and Predictive Modeling in the Management and Treatment of Episodic Migraine. Curr Pain DR. MEHRDAD FARROKHI 230 Headache Rep. 2025;29(1):56. 276. Nguyen T, Ong J, Jonnakuti V, Masalkhi M, Waisberg E, Aman S, et al. Artificial intelligence in the diagnosis and management of refractive errors. Eur J Ophthalmol. 2025;35(4):1456-80. 277. Nishida N, Kudo M. Artificial Intelligence in Medical Imaging and Its Application in Sonography for the Management of Liver Tumor. Front Oncol. 2020;10:594580. 278. Tahavvori A, Chelan RJ, Aminoleslami S, Moghadam OF, Haghighi L, Abdian Y, et al. Large Language Models and ChatGPT in Medical Sciences: Foundations, Capabilities, and Challenges. Kindle. 2025;5(1):1-222. 279. Ramezanian M, Benis DS, Nikakhtar R, Gorjizadeh N, Asadi F, Bagherianlemraski M, et al. Artificial Intelligence in Genomic Medicine: Improving Diagnostic Accuracy and Treatment Outcomes. Kindle. 2025;5(1):1-215. 280. Rahmani E, Farrokhi M, Aghajan A, Gholampour G, Ghoodjani E, Shemshadigolafzani R, et al. AI-Driven Strategies for Improving Patient Quality of Life. Kindle. 2025;5(1):1-214. 281. Rahaeimehr R, Babakhani Z, Moghadam OF, Nasir SM, Safaei P, Abdollahi MAA, et al. Application of AI in Research and Data Science. Kindle. 2025;5(1):1-362. 282. Niakosari V, Mosaddeghi-Heris R, Hezarani CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 231 HB, Farrokhi M, Safaei P, Nikseresht H, et al. AI in Medical Imaging and Early Disease Detection. Kindle. 2025;5(1):1-203. 283. Louia S, Mosaddeghi-Heris R, Kamvar R, Zahmatkesh N, Damiri M, Esfahani MA, et al. Artificial Intelligence in Cancer Genomics: Transforming Diagnosis, Treatment, and Precision Medicine. Kindle. 2025;5(1):1-234. 284. Louia S, Moghadam OF, Chelan RJ, Taheri N, Amini F, Ahmadi S, et al. Role of Immunogenetics in the Etiology, Diagnosis, and Treatment of Diseases. Kindle. 2025;5(1):1-222. 285. Javadzadeh A, Shafiei D, Amlash RS, Mehrvar R, Sepehrian S, Shafiee A, et al. The Brain-Body Connection: Neuroscience’s Role Across Medical Sciences Disciplines. Kindle. 2025;5(1):1-210. 286. Hedayati F, Chelan RJ, Alijaniha M, Koma KK, Irajian P, Rajabi N, et al. Explainable Artificial Intelligence for Reducing the Global Cancer Burden. Kindle. 2025;5(1):1-195. 287. Harati K, Tahernejad M, Saddam SMS, Farshi M, Saeedfar M, Gheibi M, et al. The Future of Prosthetics and Organ Transplantation: A Therapeutic Approach Across Various Medical Disciplines. Kindle. 2025;5(1):1-193. 288. Harati K, Mosaddeghi-Heris R, Kiani K, Saligheh Rad M, Morovatshoar R, Kamali M, et al. The AI Revolution: Predicting and Managing DR. MEHRDAD FARROKHI 232 the Next Global Health Challenges and Emerging Disease Outbreaks. Kindle. 2025;5(1):1-326. 289. Harati K, Abbasmofrad H, Ebrahimi M, Hashemlu L, Chelan RJ, Hashemzadeh A, et al. Intelligent Patient Engagement: Education and Follow-Up through AI and Telemedicine. Kindle. 2025;5(1):1-185. 290. Gheibi M, Rajabloo Y, Alipour-Khabir Y, Azami P, Louia S, Bojnordi TE, et al. Artificial Intelligence in Biomarker Discovery: Applications Across Medical Specialties. Kindle. 2025;5(1):1-209. 291. Farrokhi M, Taheri N, Moghadam OF, Armoon M, Samimi S, Torkashvand N, et al. Artificial Intelligence for Hard-toTreat and Unknown-Origin Cancers. Kindle. 2025;5(1):1-296. 292. Farrokhi M, Ghalamkarpour N, Nouri S, Babaei M, Rajabloo Y, Sattari M, et al. Innovative Vaccination: A New Era in Cancer Prevention. Kindle. 2025;5(1):1-194. 293. Babaheidarian P, Soltanattar A, Sajadi SK, Rostamian L, Foroutani L, Soleymanpourshamsi T, et al. Robotics in Healthcare. Kindle. 2025;5(1):1-178. 294. Rahmani E, Bayat Z, Farrokhi M, Karimian S, Zahedpasha R, Sabzehie H, et al. Monkeypox: a comprehensive review of virology, epidemiology, transmission, diagnosis, prevention, treatment, CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 233 and artificial intelligence applications. Archives of Academic Emergency Medicine. 2024;12(1):e70. 295. Farrokhi M, Moeini A, Taheri F, Farrokhi M, Khodashenas M, Babaei M, et al. AI-assisted Screening and Prevention Programs for Diseases. Kindle. 2023;3(1):1-209. 296. Shayganfar A, Farrokhi M, Shayganfar S, Ebrahimian S. Associations between bone mineral density, trabecular bone score, and body mass index in postmenopausal females. Osteoporosis and sarcopenia. 2020;6(3):111-4. 297. Farrokhi M, Yarmohammadi B, Mangouri A, Hekmatnia Y, Bahramvand Y, Kiani M, et al. Screening performance characteristics of ultrasonography in confirmation of endotracheal intubation; a systematic review and metaanalysis. Archives of Academic Emergency Medicine. 2021;9(1):e68. 298. Farrokhi M, Khurshid M, Mohammadi S, Yarmohammadi B, Bahramvand Y, Nasrollahi E, et al. Comparison of ultrasound-accelerated versus conventional catheter-directed thrombolysis for deep vein thrombosis: A systematic review and meta-analysis. Vascular. 2022;30(2):365-74. 299. Mirahmadi A, Hosseini-Monfared P, Amiri S, Taheri F, Farokhi M, Minaei Noshahr R, et al. Cross‑cultural adaptation and validation of the Persian version of the new Knee Society Knee Scoring System (KSS). Journal of Orthopaedic DR. MEHRDAD FARROKHI 234 Surgery and Research. 2023;18(1):858. 300. Kazemi S-M, Khorram R, Fayyazishishavan E, Amani-Beni R, Haririan Y, Khameneh SMH, et al. Diagnostic Accuracy of Ottawa Knee Rule for Diagnosis of Fracture in Patients with Knee Trauma; a Systematic Review and Meta-analysis. Archives of Academic Emergency Medicine. 2023;11(1):e30. 301. Goodarzy B, Rahmani E, Farrokhi M, Tavakoli R, Fard AM, Ghaleh MR, et al. Diagnostic value of chest computed tomography scan for identification of foreign body aspiration in children: a systematic review and meta-analysis. Archives of Academic Emergency Medicine. 2024;13(1):e3. 302. Moteshakereh SM, Zarei H, Nosratpour M, Moshfegh MZ, Shirvani P, Mirahmadi A, et al. Evaluating the diagnostic performance of MRI for identification of meniscal ramp lesions in ACLdeficient knees: a systematic review and metaanalysis. JBJS. 2024;106(12):1117-27. 303. Farrokhi M, Manavi SP, Taheri F. Noninvasive monitoring of pH and oxygen using miniaturized electrochemical sensors. Journal of translational medicine. 2021;19(1):252. 304. Rahmani E, Fayyazishishavan E, Afzalian A, Varshochi S, Amani-Beni R, Ahadiat S-A, et al. Point-of-care ultrasonography for identification of skin and soft tissue abscess in adult and pediatric CLINICAL DECISION-MAKING USING ARTIFICIAL INTELL... 235 patients; a systematic review and meta-analysis. Archives of Academic Emergency Medicine. 2023;11(1):e49. DR. MEHRDAD FARROKHI 236 Proof