Virtual Training of the Myosignal

OBJECTIVE To investigate which of three virtual training methods produces the largest learning effects on discrete and continuous myocontrol. The secondary objective was to examine the relation between myocontrol and manual motor control tests. DESIGN A cohort analytic study. SETTING University laboratory. PARTICIPANTS 3 groups of 12 able-bodied participants (N = 36). INTERVENTIONS Participants trained the control over their myosignals on 3 consecutive days. Training was done with either myosignal feedback on a computer screen, a virtual myoelectric prosthetic hand or a computer game. Participants performed 2 myocontrol tests and 2 manual motor control tests before the first and after the last training session. They were asked to open and close a virtual prosthetic hand on 3 different velocities as a discrete myocontrol test and followed a line with their myosignals for 30 seconds as a continuous myocontrol test. The motor control tests were a pegboard and grip-force test. MAIN OUTCOME MEASURES Discrete myocontrol test: mean velocities. Continuous myocontrol test: error and error SD. Pegboard test: time to complete. Grip-force test: produced forces. RESULTS No differences in learning effects on myocontrol were found for the different virtual training methods. Discrete myocontrol ability did not significantly improve as a result of training. Continuous myocontrol ability improved significantly as a result of training, both on average control and variability. All correlations between the motor control and myocontrol test outcome measures were below .50. CONCLUSIONS Three different virtual training methods showed comparable results when learning myocontrol. Continuous myocontrol was improved by training while discrete myocontrol was not. Myocontrol ability could not be predicted by the manual motor control tests.