View Full Version : Myoelectric EMG Control Prosthetic Arm-Development & Analysis

See See
08-18-2002, 01:39 AM
Dear all,

I am an undergraduate from Biomedical Engineering, University of Malaya,
Malaysia. I am currently doing my thesis project under the subject of
rehabilitation engineering,
with the topic " Development of an biceps and triceps EMG signal-based
control system of a prosthetic arm".

I wish to ask that, does anyone know that what are the 3 muscles pair ( I
propose to use a combination pair of muscles (eg. biceps and supinator) )
that are most suitable for control site?
I have tried posterior deltoid, supinator, biceps, triceps & wirst flexor,
yet, the result not very appealing..maybe because I am not trained and still
finding help from a physiotherapist)
However, I wonder which three pairs are the most suitable one for arm
prosthesis control site?

Also, what is the lastest and greastest achivement in EMG control prosthetic
arm so far world-widely?
I know that so far, multifunctional prosthetic arm with high degree of
freedom is the main concern. However, there is much limitations that rather
hard to break through. Thus, what is the future of EMG control prosthetic
arm? In my opinion, maybe it will be surpassed by EEG control prosthetic
arm, which is more flexible and has a greater potential. Does anyone agree
or disagree? Or any other opinions?

Can anyone please provide me a statistical support regarding percentage of
usage of EMG control prosthetic arm in Europe as well as Asian countries?
This is because in Malaysia, there are relatively few amputees ( actually
very few) using myoelectric control prosthetic arm, I wish to know more
about the potential of this control method.

Can anyone suggest how to read a spectrogram ( rather fuzzy with a lot of
colours) in order to do good EMG features extraction, to gain the input
vectors for backpropagation classifier?

As features extraction is a very important step to get a good EMG signal
classification,I wish to ask that what is the most effective EMG features so
far? ( eg IAV, mean absolute value, zero crossing, PSD...) What is the
advantages and trade-offs of each features? For example,
IAV is good for small window application. I also do not wish to try
AutoRegressive method (AR) as my another coursemate is currently dealing
with it.

I am rather blur in the algorithm of EMG feature extraction. I have read
through some books and journals yet they do not explain detailly in this
part. Could anyone please give me an example of feature extraction algorithm
(code) in C or MATLAB?

Also, please suggest some good materials for EMG feature extraction. I have
searched hardly for a long time yet with less success. I really hope that
some good journals or books can assist me in this part.

As usual, I will post a summary at the end of this discussion for the
benefit of all.


Anxiously waiting your replies,

Matilda Cheng
Department of Biomedical Engineering
Faculty of Engineering
University of Malaya
Kuala Lumpur, Malaysia
603 79673279(Tel)
603 79560027(Fax)
Email : matildacheng@hotmail.com

__________________________________________________ _______________
Chat with friends online, try MSN Messenger: http://messenger.msn.com

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l