A key factor in the clinical translation of brain-machine interfaces (BMIs) for restoring hand motor function will be their robustness to changes in a task. With functional electrical stimulation (FES) for example, the patient’s own hand will be used to produce a wide range of forces in otherwise similar movements. To investigate the impact of task changes on BMI performance, we trained two rhesus macaques to control a virtual hand with their physical hand while we added springs to each finger group (index or middle-ring-small) or altered their wrist posture. Using simultaneously recorded intracortical neural activity, finger positions, and electromyography, we found that predicting finger kinematics and finger-related muscle activations across contexts led to significant increases in prediction error, especially for muscle activations. However, with respect to online BMI control of the virtual hand, changing either training task context or the hand’s physical context during online control had little effect on online performance. We explain this dichotomy by showing that the structure of neural population activity remained similar in new contexts, which could allow for fast adjustment online. Additionally, we found that neural activity shifted trajectories proportional to the required muscle activation in new contexts, possibly explaining biased kinematic predictions and suggesting a feature that could help predict different magnitude muscle activations while producing similar kinematics.

For a description of data collection and processing see the associated preprint https://doi.org/10.1101/2022.08.26.505422

Simultaneous finger movements in two groups (index and middle-ring-small), EMG, and intracortical neural activity were recorded while two monkeys did a task moving both finger groups seperately (2-DOF) or all fingers together (1-DOF). Data was collected in runs of about 100 to 400 trials and then task context was altered by adding springs to each finger group, and/or rotating the wrist to be flexed. 3 to 7 runs of trials were done in one session and each file contains multiple sessions. 

Threshold crossings in neural data were recorded and then converted into a threshold crossing firing rate (TCFR). Neural data was also recorded at 2ks/s, filtered at 300-1000Hz and then rectified. Finger angles (0 extended to 1 flexed) were recorded with bend sensors. 16 channels of EMG were recorded from 8 muscles, odd channels are referenced to their bipolar pair (even channels). Muscles are FCR, FDPid, FDPip, FDP, FCU, ECRB, EIP, EDC. EMG was bandpass filtered at 100-500Hz. Here data is presented binned into 32ms (2-DOF) or 20ms (1-DOF) intervals. To bin EMG the mean absolute value is taken in a bin. For other features the mean value is taken in each bin.

Included files are .mat format which requires MATLAB to open.