The effect of socket design on the reliability of EMG signal transduction in trans-radial myoelectric prostheses

Background: The majority of trans-radial myoelectric prosthesis users experience some level of difficulty and disruption in the control of their prosthesis everyday (1). Previous studies by Head (1) and Chadwell et al. (2) have suggested that the fit and/or design of the socket could be a contributing factor in the disruption of prosthesis control by introducing uncertainty and motion artefacts into the control system. However, there is currently no definitive prescription criterion that determines the effect that trans-radial socket design has on myoelectric control. Aims: To investigate the effect of socket design on: the reliability of voluntary-initiated prosthetic prehensor activation and The avoidance of unwanted prosthetic prehensor activation Methods: Six participants with a trans-radial limb absence and previous experience of using a myoelectric prosthesis were recruited for the study. The user performance of three different trans-radial socket types, the UK Hybrid, Longitudinal Compression and Münster socket, were evaluated against a baseline no-socket ‘ideal’ condition (in which the electrodes are held firmly against the skin). The participants ability to control myoelectric prosthesis activation was assessed via the use of a reaction time test with the forearm in 2 orientations, each being 45° either side of the horizontal, as described by Chadwell et al (2). The frequency of unplanned activations caused by motion artefacts during these movements was determined by a goniometer situated on the prosthetic forefinger that measured the aperture of the hand, and recorded any unwanted activation. Each participant also completed a questionnaire which evaluated each socket via the following: comfort; ease of control; and overall preference. Results: The UK Hybrid had the least impact on voluntary activation of the hand for the Open Function (Difference in Spread= 15.83 (Standard Deviation of Reaction Time Socket – Standard Deviation of Reaction Time No Socket)), and also had a considerably lower standard deviation in reaction times than the no-socket condition in the close function. All three sockets caused unwanted prehensor activations in at least 65% of trials, in comparison to 30.56% of ‘ideal’ no-socket condition trials, in the functional movement tasks. Summary: The socket that produced experimental data closest to that of the ‘ideal’ was the UK Hybrid. The results suggest that the fit of the socket, rather than the specific design, is more important when creating a reliable electrode-socket interface. All participants cited the most comfortable socket as their overall preference, although this finding should be interpreted with caution