Predicting lower limb muscular activity during standing and squatting using spikes of primary motor cortical neurons in monkeys

In this study, we investigated predicting lower limb muscular activities of monkeys during standing and squatting motions using neuronal spikes in primary motor cortex M1. Finite impulse response models were built for prediction. Acute electrode arrays were used to collect neuronal spikes in the lower limb representation area of M1 in the left hemisphere, and electrodes were implanted to the right leg muscles to collect EMG signals. Multiple regions of the lower limb representation area of M1 were explored. The neurons from two common regions demonstrated high predictive power on all 6 investigated right leg EMG signals. This study shows that the cortical neuronal spikes can be used to predict lower limb muscular activities with high accuracy, and identifies regions of high predictive power, where chronic electrodes can be implanted for future brain machine interface applications.