Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain-machine interface.
ABSTRACT Patients with severe neurological injury, such as quadriplegics, might benefit greatly from a brain-machine interface that uses neuronal activity from motor centers to control a neuroprosthetic device. Here, we report an implementation of this strategy in the human intraoperative setting to assess the feasibility of using neurons in subcortical motor areas to drive a human brain-machine interface.
Acute ensemble recordings from subthalamic nucleus and thalamic motor areas (ventralis oralis posterior [VOP]/ventralis intermediate nucleus [VIM]) were obtained in 11 awake patients during deep brain stimulator surgery by use of a 32-microwire array. During extracellular neuronal recordings, patients simultaneously performed a visual feedback hand-gripping force task. Offline analysis was then used to explore the relationship between neuronal modulation and gripping force.
Individual neurons (n = 28 VOP/VIM, n = 119 subthalamic nucleus) demonstrated a variety of modulation responses both before and after onset of changes in gripping force of the contralateral hand. Overall, 61% of subthalamic nucleus neurons and 81% of VOP/VIM neurons modulated with gripping force. Remarkably, ensembles of 3 to 55 simultaneously recorded neurons were sufficiently information-rich to predict gripping force during 30-second test periods with considerable accuracy (up to R = 0.82, R(2) = 0.68) after short training periods. Longer training periods and larger neuronal ensembles were associated with improved predictive accuracy.
This initial feasibility study bridges the gap between the nonhuman primate laboratory and the human intraoperative setting to suggest that neuronal ensembles from human subcortical motor regions may be able to provide informative control signals to a future brain-machine interface.
SourceAvailable from: conservancy.umn.edu
[Show abstract] [Hide abstract]
ABSTRACT: Objective. For intracortical brain-machine interfaces (BMIs), action potential voltage waveforms are often sorted to separate out individual neurons. If these neurons contain independent tuning information, this process could increase BMI performance. However, the sorting of action potentials ('spikes') requires high sampling rates and is computationally expensive. To explicitly define the difference between spike sorting and alternative methods, we quantified BMI decoder performance when using threshold-crossing events versus sorted action potentials. Approach. We used data sets from 58 experimental sessions from two rhesus macaques implanted with Utah arrays. Data were recorded while the animals performed a center-out reaching task with seven different angles. For spike sorting, neural signals were sorted into individual units by using a mixture of Gaussians to cluster the first four principal components of the waveforms. For thresholding events, spikes that simply crossed a set threshold were retained. We decoded the data offline using both a Naïve Bayes classifier for reaching direction and a linear regression to evaluate hand position. Main results. We found the highest performance for thresholding when placing a threshold between -3 and -4.5 × Vrms. Spike sorted data outperformed thresholded data for one animal but not the other. The mean Naïve Bayes classification accuracy for sorted data was 88.5% and changed by 5% on average when data were thresholded. The mean correlation coefficient for sorted data was 0.92, and changed by 0.015 on average when thresholded. Significance. For prosthetics applications, these results imply that when thresholding is used instead of spike sorting, only a small amount of performance may be lost. The utilization of threshold-crossing events may significantly extend the lifetime of a device because these events are often still detectable once single neurons are no longer isolated.Journal of Neural Engineering 12/2014; 12(1):016009. DOI:10.1088/1741-2560/12/1/016009 · 3.42 Impact Factor