One of the biggest concerns for control of robotic devices for rehabilitation of amputees is related to the sEMG signal quality. Depending on how clean is the signal, more efficient and effective is the response of the control in relation to the user needs. This article proposes a protocol for sEMG signal capture with the least amount of crosstalk without the use of filters on the forearm with four channels in healthy and amputee people. Extensor digitorum muscle (Channel 1), flexor digitorum superficilis (Channel 2), flexor carpi ulnaris (Channel 3) and flexor pollicislongus (Channel 4) were used. This protocol is used in a series of ten isometric motor tasks related to the movement of the fingers and wrist on the hand. The attenuates 90% of the noise generated by the grid at 59.97 to 60.05 Hz and their harmonics, together with identified unusual noise frequency of 258.1 Hz which was isolated in 75% for the four channels during all motor tasks. This allows the recognition of motor defects with the use of the signal obtained without the use of filters, allowing a lower computational overhead for processing the signal to control a myoelectric hand prosthesis.
展开▼