Abstract
Objective: Recently, several studies have documented the presence of a bimodal distribution of spike waveform widths in primary motor cortex. Although narrow and wide spiking neurons, corresponding to the two modes of the distribution, exhibit different response properties, it remains unknown if these differences give rise to differential decoding performance between these two classes of cells. Approach: We used a Gaussian mixture model to classify neurons into narrow and wide physiological classes. Using similar-size, random samples of neurons from these two physiological classes, we trained offline decoding models to predict a variety of movement features. We compared offline decoding performance between these two physiologically defined populations of cells. Main results: We found that narrow spiking neural ensembles decode motor parameters better than wide spiking neural ensembles including kinematics, kinetics, and muscle activity. Significance: These findings suggest that the utility of neural ensembles in brain machine interfaces may be predicted from their spike waveform widths.
Highlights
Neural interface systems for control have recently made a number of important advances in recording capabilities, decoding algorithms, and output devices [1, 2]
We found that a two-component model had optimal Akaike information criterion (AIC) values
We found that each muscle's activity was better predicted by narrow spiking ensembles
Summary
Neural interface systems for control have recently made a number of important advances in recording capabilities, decoding algorithms, and output devices [1, 2] These systems have seen nearly a doubling of simultaneously recorded neurons every seven years using either high density electrode arrays, or more recently, optical calcium fluorescence imaging [3]. These advances have provided an ever growing set of rich, high-dimensional signals for control [4]. Decoding ability has not increased correspondingly with the growth of input signals, but rather has plateaued This disparity arises, in part, because highdimensional input signals require larger models to relate neural activity to motor features. There is some debate about the optimal size of decoding models [8]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
