scieee Science in your language
[en] (orig)

Forearm-attachable EMG-based wireless controller

Abstract

This paper presents the development process starting from a simple electronic module for amplification of EMG signals. During development the tasks to be solved were extended towards the creation of an EMG-base controller that eventually can be connected to a PC. Among the main goals were the simplicity and low cost of production. There are similar devices commercially available, but the main focus was on the whole process of design, testing, fabrication and the exploration of application areas. It was important to use medical recommendations. The device matches the suggestion of SENIAM, except for the resolution of A/D conversion, which is permissible due to the fact that it was not designed for medical purposes. The final version is capable of detecting and amplifying EMG signals, with the appropriate filtering and digitizing procedures for further processing. With the development of wireless communication capabilities, the device which can be attached to the forearm can be used as a USB HID device with a PC. A number of additional applications are also discussed, pointing out the fact that this piece of technology offers a wide range of possibilities.

Read accessible full text

Forearm-attachable EMG-based wireless controller

Author: Bánfi, Tamás; Aradi, Petra
Publisher: DUPress
Year: 2016
Source: https://dea.lib.unideb.hu/bitstreams/66b14a06-7dc6-402f-8bab-484998cb81c9/download
Recen Inno a ions in Mecha onics (RIiM) Vol. 3. (2016). No. 1-2.
DOI: 10.17667/ iim.2016.1-2/18.
Fo ea m-a achable EMG-based wi eless con olle
Tamás Bán i BSc
Budapes Uni e si y o Technology and Economics
Facul y o Mechanical Enginee ing, Depa men o
Mecha onics, Op ics and Mechanical Enginee ing
In o ma ics
Budapes , Hunga y
ban i. [email protected]
Pe a A adi PhD
Budapes Uni e si y o Technology and Economics
Facul y o Mechanical Enginee ing, Depa men o
Mecha onics, Op ics and Mechanical Enginee ing
In o ma ics
a adi.p[email p o ec ed]
Abs ac — This pape p esen s he de elopmen p ocess
s a ing om a simple elec onic module o ampli ica ion o
EMG signals. Du ing de elopmen he asks o be sol ed we e
ex ended owa ds he c ea ion o an EMG-base con olle ha
e en ually can be connec ed o a PC. Among he main goals we e
he simplici y and low cos o p oduc ion. The e a e simila
de ices comme cially a ailable, bu he main ocus was on he
whole p ocess o design, es ing, ab ica ion and he explo a ion
o applica ion a eas. I was impo an o use medical
ecommenda ions. The de ice ma ches he sugges ion o
SENIAM, excep o he esolu ion o A/D con e sion, which is
pe missible due o he ac ha i was no designed o medical
pu poses. The inal e sion is capable o de ec ing and ampli ying
EMG signals, wi h he app op ia e il e ing and digi izing
p ocedu es o u he p ocessing. Wi h he de elopmen o
wi eless communica ion capabili ies, he de ice which can be
a ached o he o ea m can be used as a USB HID de ice wi h a
PC. A numbe o addi ional applica ions a e also discussed,
poin ing ou he ac ha his piece o echnology o e s a wide
ange o possibili ies.
Keywo ds—elec omyog aphy, a m mo emen , A duino,
STM32 mic ocon olle , signal p ocessing, USB HID con olle
I. INTRODUCTION
A. The pu pose o he wo k
Elec omyog aphy (EMG) is a widely-used medical
p ocess o examine he ac i i y o muscles. The de ice is
called elec omyog aph, he diag am being de ec ed is he
elec omyog am, which ep esen s he elec ical ac i i y o
muscles in mV o μV ange o e ime. Du ing he
measu emen elec odes a e a ached o he skin su ace o
needles a e inse ed in o he muscles. The second solu ion is
much mo e accu a e, bu i can sense he elec ic ac i i y o
jus a ew mo o uni s. This is an in asi e me hod, so medical
p o essionals use i mos ly o speci ic diagnos ic pu poses.
The nonin asi e o m is applied o mo emen de ec ion o as
a senso . The p esen ed de ice is based on he second me hod.
In o de o measu e he muscle ac i i y, a pai o elec odes is
a ached longi udinally o he skin su ace abo e he muscle.
The muscles in his applica ion a e he o ea m muscles,
esponsible o mo ing he w is and inge s. As su ace
de ec ion is he chosen me hod, he p o ound muscles canno
be examined. This is one o he easons, ha du ing
de elopmen and es ing he de ice he loca ion, o ien a ion
and unc ion o hese muscles mus be known [1], [2], [3]. To
gain mo e in o ma ion abou he a m’s mo ion, a 6 deg ee o
eedom accele ome e and gy oscope, an IMU (Ine ial
Measu emen Uni ) is used. This senso de ec s angula and
linea accele a ion. Fu he mo e, choosing he mos sui able
mic ocon olle is also an impo an s ep. I mus be able o
a end o se e al unc ions: digi izing 8 analog channels
conside ing he Nyquis -Shannon sampling heo em,
communica ing wi h he IMU and sending he da a o ano he
con olle wi elessly.
B. The cha ac e is ics o he EMG signal
Figu e 1 shows an elec omyog am measu ed by a medical
de ice. The elec odes a ached o he pa ien s’ o ea m de ec
muscles’ ac i i y while g abbing. Du ing he esea ch, he
pa ien ’s ask was o g ip a dynamome e , hen elax he
muscles. This p ocedu e had o be epea ed i e imes while
inc easing he o ce. The i s g aph ep esen s he
dynamome e ’s alues in kg, he second one shows he aw-,
he hi d one he il e ed EMG signals in mV, all o hem o e
ime.
Fig. 1. Elec omyog am [7]
I can be easily obse ed ha EMG signals a y, while he
g abbing o ce is cons an . Acco ding o he en elopes, i
would be eally di icul o de e mine he ix g abbing o ce.
Ano he obse a ion is ha he huge o ce one exe s, he
highe he g aph’s g adien is. This e ec is explained by he
s ia ed muscles’ a igue. Mo eo e , he en elope’s
smoo hness can be inc eased by educing he low-pass il e ’s
equency. Howe e , his would wo sen he eac ion ime.
Las , bu no leas , i is a di icul ask o di e en ia e low-
o ce g abbing signals om noise.
Recen Inno a ions in Mecha onics (RIiM) Vol. 3. (2016). No. 1-2.
DOI: 10.17667/ iim.2016.1-2/18.
Summing up he consequences o he measu emen , i is
clea ha du ing he p ocessing o EMG signals, one has o
deal wi h se e al di icul ies.
II. HARDWARE INTERFACE
A. Designing and building he ci cui
I was necessa y o design and build a ci cui o
measu ing and ampli ying he EMG signals. A i s , a es -
panel was buil o he de ec ion o only one EMG channel,
based on ideas om [8]. A common A duino panel was used
o digi iza ion o he measu ed EMG signal. Gaining
expe ience om his unc ioning ci cui , he inal panel, which
can measu e 8 channels was designed and c ea ed. I con ains
an accele ome e and gy oscope, as well as a wi eless adio
module. The A/D con e sion o signals was pe o med by an
STM32F4 Disco e y panel, because i s speed and esolu ion
a e much highe , han ha o he i s choice o A duino Mega.
The ol age supply was a 9V ba e y.
B. The panels in he de ice
The de eloped panels can be di ided in o wo g oups
based on hei unc ions: ins umen a ion ampli ie panels and
ope a ional ampli ie panel. The second one is also a shield o
he STM32. The ins umen a ion ampli ie de ec s he
po en ial di e ence be ween he wo ends o he muscle ha is
how he EMG signal is gene a ed. Mo eo e , hese panels
p o ide ampli ica ion by 20. As i ope a es om an
asymme ic powe supply, he signals ha e o be pu on o a 1,5
V o se be ween 0 and 3V. Then a band-pass il e was
applied, which emo es he unnecessa y high and low
equency componen s. By cu ing he low band, he d i and
he DC o se a e elimina ed. High- equency il e ing ex ac s
he noises and p e en s aliasing. Acco ding o SENIAM
(Su ace Elec oMyoG aphy o he Non-In asi e Assessmen
o Muscles) [5], o su ace elec odes 5-10 Hz high-passing
and 500 Hz low-passing should be used (Figu e 2). In he
cou se o he de elopmen , based on he ga he ed da a, 7.23
Hz high- and 482.29 Hz low-passing equencies we e se .
Fig. 2. Schema ic ou line o a sEMG eco ding sys em [6]
C. The inal PCBs
A e planning he ci cui s, NI Mul isim simula ion
so wa e we e used o es s. Figu e 3 shows he Bode diag am
esul s om he simula ion.
Fig. 3. Bode diag am esul o he simula ion
The PCBs we e designed wi h a p og am called DipT ace.
PCB e ching and solde ing he componen s we e he inal
manu ac u ing s eps. All in all, 8 smalle panel (Figu es 4 and
5), as well as a la ge shield we e made (Figu es 6 and 7).
Fig. 4. PCB design o he small ins umen a ion ampli ie ’s’ panel
Fig. 5. PCB o he small ins umen a ion ampli ie ’s’ panel
Recen Inno a ions in Mecha onics (RIiM) Vol. 3. (2016). No. 1-2.
DOI: 10.17667/ iim.2016.1-2/18.
Fig. 6. PCB design o he shield
Fig.7. PCB o he shield wi h he ope a ional ampli ie s, he wi eless adio
module and he IMU
III. SOFTWARE INTERFACE
Fo he compu e p ocessing o he signal, a digi al low-
pass il e is needed because o he i ing o he en elope.
Fi s , he signal mus be ec i ied: he p og am measu es he
EMG signals symme ic o 1.5V, hen subs ac s he mean
alue om hem, so ha he o se will be 0. A e ha
nega i e numbe s will be in e ed. As o he il e ype, he
usage o IIR (In ini e Impulse Response) is easonable,
because i p o ides highe p ocessing speed and esul s in a
smoo he en elope, han FIR (Fini e Impulse Response)
il e s. A 2nd o de Bu e wo h il e wi h 5 Hz cu was
chosen a e ex ensi e es ing.
A. Tes ing
The boa d was es ed wi h an A duino Mega a i s . To
display he signals, a ba e y-powe ed lap op was chosen, in
o de o elimina e mains dis u bance. The p og am sen he
aw and he il e ed EMG signals wi h se ial communica ion
o he PC, whe e hey we e eco ded. These esul s we e
isualized wi h Mic oso Excel. Figu e 8 shows he aw
(uppe ) and he il e ed (lowe ) EMG signals using 5 Hz cu .
The ho izon al axis shows ime in seconds, he e ical axis
shows he digi ized alue, whe e 1023 bi co esponds o 5V.
Fig. 8. The aw (uppe diag am) and he il e ed (lowe diag am) EMG signals
using 5 Hz cu
B. Final boa d
The inal de ice unc ions (Figu e 9) as a con olle
a ached wi elessly o a PC. I is able o handle 8 di e en
EMG signals and he da a o he IMU simul aneously. The
ci cui s a e loca ed in di e en plas ic boxes, which can be
a ached o he o ea m wi h an elas ic band. The STM32 and
i s shield a e in a bigge box. The nRF24L01+ communica ion
module p o ides he high ange wi eless communica ion wi h
he PC.
Fi s , he ansmi e p og am eads he signals o he 8
analog inpu s. Each o hem a e a e aged, o ming he
channels’ o se s. O se s a e sub ac ed om aw signals, so
hei mean alue will be 0. Nega i e signals a e in e ed, hen
a 2nd o de Bu e wo h il e p ocesses hem. These signals
and he IMU’s da a a e sen o he ecei e .
The da a a i e in an a ay o he ecei e , which
disassembles and pu s hem o di e en a iables. The
ecei e con e s IMU’s da a in o he i ems o X and Y axes.
Two EMG signals, gene a ed by a lexo and an ex enso
muscle, a e u ned in o a logical alue. I he muscle is ac i e,
a bu on is pushed. All he da a a e sha ed wi h he PC.
Fig. 9. The inal de ice
Recen Inno a ions in Mecha onics (RIiM) Vol. 3. (2016). No. 1-2.
DOI: 10.17667/ iim.2016.1-2/18.
IV. CONCLUSION
The i s expe imen al esul s came om es ing he panel
made o A duino. A e ha a p in ed ci cui boa d was
designed, buil and es ed, so ha inciden al e o s could be
ecognized and co ec ed. Then he digi al il e s app op ia e
o adjus ing he en elope we e selec ed. La e on he inal
de ice was made "mobile" wi h a 433 MHz communica ion
module. Du ing he p ojec , he e o s we e iden i ied and
co ec ed, so ha in he inal de ice hey we e omi ed.
Wi h he expe ience gained du ing de elopmen and
es ing, a much as e sys em was designed and c ea ed, ha is
capable o moni o ing 8 muscles, and can be placed on he
o ea m wi hou elec ode cables. STM32F4-Disco e y
handles digi iza ion, and includes a much as e
mic ocon olle han A duino, so ha he sampling equency
could ha e been inc eased. Fu he ad an age o STM32F4-
Disco e y is a 12 bi A/D con e e compa ed o A duino's 10
bi sys em, inc easing he esolu ion o he sampled signals.
Choosing he new, high pe o mance 2.4 GHz wi eless
communica ion ins ead o he 433MHz modules, he
ope a ional ange and da a speed inc eased conside ably. By
upg ading he de ice wi h a Six-Axis Mo ion T acking
De ice, he a m's posi ion and mo emen could be moni o ed
wi h mo e de ails. The ci cui boa d o he ecei e , which is
able o communica e se ially wi h he compu e was also
designed and buil . Fu he mo e, he inal de ice can be
connec ed o he compu e as an USB HID (Human In e ace
De ice). The e is no need o ins all a d i e , he de ice can be
used soon a e connec ion. This unc ion was es ed on a
ocke powe ed ai c a simula o game called HAWX, whe e
he igh e je s we e con olled by he de ice a ached o he
playe 's o ea m (Figu e 10).
Fig. 10. Playing wi h he simula o
V. FURTHER APPLICATION AND TASKS
A. Examina ion o muscle a igue
The o ea m muscles a e olun a ily mo eable s ia ed
muscles, which a e p one o a igue. This ea u e can be
de ec ed wi h an EMG boa d. All one has o do is o pu he
elec odes on o he biceps muscle. Then a hea ie weigh
should be li ed and hold as long as possible. Du ing he
exe cise he ampli ude o he EMG signal educes slowly. I is
also a good idea o use he de ice o wo kou ollow-up.
B. Follow-up o physio he apy ea men s
When someone needs o do physio he apy egula ly, he
e iciency could be supe ised, as well as he buildup o
s eng h. Fo ins ance, du ing he exe cise he pa ien should
epea edly push a small ubbe ball. The mic ocon olle
coun s one up a e he EMG signal eaches a ce ain analog
alue.
C. Con olling a d one o mobile obo
The 2,4 GHz communica ion module can be used o
con ol d ones o e en obo s because o i s qui e la ge ange.
Mo ing he gy oscope, i can di ec he con olled ehicle’s
mo emen . Wi h he s e ching o elaxing o muscles
di e en o de s can be gi en. One o i s ad an ages is ha he
hand emains ee.
D. Examining walk and balance
The inal de ice shows muscles’ elec ic ac i i y in a
sui able way, so a aching i o he leg, he ac i a ion le el can
be sensed while walking o balancing.
E. Redesign
I is wo h o examine he possibili ies o edesign he
de ice in a mo e compac o ma , o a wide ange o
applica ions.
REFERENCES
[1] A. Fonyó A ila: Medical Physiology (Az o osi éle an anköny e). 7 h
Ed.. Budapes , Medicina, 2014.
[2] M. Ré helyi, J. Szen ágo hai: Func ional Ana omy (Funkcionális
ana ómia). Band I. Budapes , Medicina, 2006.
[3] F ied ich Paulsen, Jens. Waschke: Sobo a A las o Human Ana omy
(Az embe ana ómiájának a lasza). 22nd Ed. Band I. Budapes ,
Medicina, 2012.
[4] William Rose: Elec omyog am analysis. 2014.
h p://www.udel.edu/biology/ osewc/kaap686/no es/EMG%20analysis.p
d
[5] SENIAM 8 Eu opean Recommenda ions o Su ace
Elec oMyoG aphy. Published by Roessingh Resea ch and De elopmen
b. . 1999.
[6] D.F. S egeman, H.J. He mens: S anda ds o su ace elec omyog aphy:
he Eu opean p ojec "Su ace EMG o non-in asi e assessmen o
muscles (SENIAM)". h p://www.med.uni-
jena.de/mo o ik/pd /s egeman.pd
[7] L. Dé á i: In es iga ion o muscle ac ion wi h elec omypg aphy (Az
izomműködés anulmányozása elek omiog á ia segí ségé el)
h p://de a i.web.el e.hu/gyak_jegyze /emg.pd
[8] Backya d B ains. Neu oscience o E e yone!
h ps://www.backya db ains.com/