EMG-Based Hand-Wrist Control: Study B Mirrored (PSICON-B)

Worcester Polytechnic Institute, Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD)

Study assessing four-channel prosthesis controller, that compares contralateral (mirrored) EMG-force training to ipsilateral EMG-target training with both limb-absent and able-bodied subjects

System identification models relating forearm electromyogram (EMG) signals to phantom wrist radial-ulnar deviation force, pronation-supination moment and/or hand open-close force (EMG-force) are hampered by lack of supervised force/moment output signals in limb-absent subjects. In able-bodied and unilateral transradial limb-absent subjects, we studied three alternative supervised output sources in one degree of freedom (DoF) and 2-DoF target tracking tasks: (1) bilateral tracking with force feedback from the contralateral side (non-dominant for able-bodied/ sound for limb-absent subjects) with the contralateral force as the output, (2) bilateral tracking with force feedback from the contralateral side with the target as the output, and (3) dominant/limb-absent side unilateral target tracking without feedback and the target used as the output.

prosthesis, electromyography, EMG, sEMG, control, myo, regression, transradial, terminal device, electrodes

Ten unilateral transradial limb-absent subjects will each participate in one, half-day experimental session. Subjects will be seated and prepared in test apparatus seat (16 electrodes on the affected side, hand-wrist on the able side secured to load cells). Subjects will complete the 1-DoF and 2-DoF "dynamic" (force-varying) contractions (40-s duration, 0.75 Hz bandlimited, uniform random target). With 2 DoF contraction trials, hand Opn-Cls will always be one of the dimensions. The subject will then be released from the cuff and their able side not further involved in the experiment. The force feedback triangle cursor on the computer screen will be deleted such that only the target remains. Subjects will then repeat 1-DoF and 2-DoF trials in which the affected side attempts to produce hand-wrist effort that mimic movement of the target (with no feedback provided).

Ten able-bodied subjects, the electrodes will be mounted on the dominant arm for the EMG-force and EMG-target tracking trials. In addition, the non-dominant arm will also be constrained and measured in a second load cell. This load cell will not be used for feedback during the experiment, but will compare (RMS error, off-line) the dominant vs. non-dominant forces. For EMG-target tracking, the dominant hand will remain in the wrist cuff (to prevent flailing during contractions), with the screen feedback disabled. These subjects will repeat the trials with the electrodes moved to the non-dominant side. EMG-force tracking will be repeated using mirrored contractions. The three training methods (EMG-force ipsilateral, EMG-force contralateral mirrored, EMG-target on the dominant side) will be contrasted to help understand the source of errors when training with limb-absent subjects.

Test apparatus acquires 16 sEMG channels, measure four DoFs of force/moment at the hand-wrist [hand open-close (Opn-Cls); wrist extension-flexion (Ext-Flx), radial-ulnar deviation (Rad-Uln) and pronation-supination (Pro-Sup)]. sEMG system used 16 custom encased bipolar electrodes. A commercial LTI EMG amplifier (BE328) conditioned each EMG signal before A/D conversion on a PC. A thermoplastic hand cuff secured the wrist to a six-axis load cell (AMTI, model MC3A-100) for force/moment measurement. A separate hand-grasp one-axis load cell (Omega Engineering Inc., model LCR-150) was Velcro-secured between the fingers and the thumb, to measure power grip forces. The PC acquired and stored the sEMG and load cell data, and commanded a triangular screen target. The subject controlled a second triangular cursor. Left-right movement of the cursor was controlled by wrist Ext-Flx, up-down movement by Rad-Uln deviation, cursor rotation by Pro-Sup and cursor size by hand Opn-Cls forces.

Electromyogram waveform measurements (signal amplitude, frequency) will be related to force measurements (magnitude and moment) as related to the target tracking tasks. These signal-force data will be use to generate the models. The specific outcome for each subject will be the RMS error measured in the tracking tasks as the difference between the generated control signal from the user versus the prompted target signal, as measured by an EMG array.

Inclusion Criteria: Limb-absent subjects: unilateral trans-radial limb absence or amputation Be capable of completing the requested contractions on the affected dominant side Exclusion Criteria: Past injuries to the upper limbs that would limit their ability to complete the requested contractions Scars that would impede the use of surface electrodes