Predicting Movement from Multiunit Activity

Abstract

Previous studies have shown that intracortical activity can be used to operate prosthetic devices such as an artificial limb. Previously used neuronal signals were either the activity of tens to hundreds of spiking neurons, which are difficult to record for long periods of time, or local field potentials, which are highly correlated with each other. Here, we show that by estimating multiunit activity (MUA), the superimposed activity of many neurons around a microelectrode, and using a small number of electrodes, an accurate prediction of the upcoming movement is obtained. Compared with single-unit spikes, single MUA recordings are obtained more easily and the recordings are more stable over time. Compared with local field potentials, pairs of MUA recordings are considerably less redundant. Compared with any other intracortical signal, single MUA recordings are more informative. MUA is informative even in the absence of spikes. By combining information from multielectrode recordings from the motor cortices of monkeys that performed either discrete prehension or continuous tracing movements, we demonstrate that predictions based on multichannel MUA are superior to those based on either spikes or local field potentials. These results demonstrate that considerable information is retained in the superimposed activity of multiple neurons, and therefore suggest that neurons within the same locality process similar information. They also illustrate that complex movements can be predicted using relatively simple signal processing without the detection of spikes and, thus, hold the potential to greatly expedite the development of motor-cortical prosthetic devices.

Grasp Movement Decoding from Premotor and Parietal Cortex

“…Several studies have shown that the tuning properties of multiunits are similar to single units (Desimone and Gross, 1979;Super and Roelfsema, 2005;Zeitler et al, 2006;Stark and Abeles, 2007). Our data confirm this view for grasping in AIP and F5.…”

“…Waveforms were thus pooled into a larger "multiunit" that comprised recordings from several individual neurons simultaneously. However, care should be taken to distinguish this point-process signal from continuous "multiunit activity" data generated from envelope functions applied to the low-pass-filtered voltage trace (Super and Roelfsema, 2005;Stark and Abeles, 2007;Choi et al, 2010). Finally, the predominance of multiunit recordings in our data set was in part due to the fixed (nonmovable) nature of the electrodes, which did not allow optimization of unit isolation during recordings.…”

“…One plausible explanation is that the summation over individual neurons weakens or perturbs the orientation tuning measured in the multiunit data (Ventura, 2008;Fraser et al, 2009). However, multiunit activity may encode additional information not present at the singleunit level (Zeitler et al, 2006;Stark and Abeles, 2007), and therefore may represent a viable decoding source in its own right. Our finding that classification performance was only modestly improved by additional off-line sorting (Figs.…”

“…The use of multiunit activity can therefore be regarded as a trade-off between ease of signal acquisition and tuning fidelity, with the hope that the latter could be improved by advanced decoding paradigms and statistical models (Stark and Abeles, 2007;Ventura, 2008;Fraser et al, 2009). …”

Predicting voluntary movements from motor cortical activity with neuromorphic hardware

“…It can be generated simply by thresholding extracellular voltage signals, and thus avoids computationally costly spike sorting algorithms [40]. It has been shown that in the motor cortex multi-unit activity of a single channel may carry a similar amount of information about a movement as the single units extracted from the same recording channel [3], [41].…”