Real-time estimation of predictive firing rate

Abstract

Population vector analysis of multi-unit firing code in the motor cortex provides a real-time estimate of the desired voluntary movement direction and trajectory used in neuroprosthesis for assisting paralyzed patients. Accurate estimation of instantaneous firing rate in each epoch used in the population vector calculation is crucial. Yet real-time instantaneous firing rate is often unknown until the next spike has fired. We have developed statistical based estimator for predicting the waiting time for the next spike to fire. Instantaneous firing rate is computed based on the conditional probability density function (pdf), providing a better estimate of the intended movement trajectory.