Neural ensemble dynamics in dorsal motor cortex during speech in people with paralysis

Abstract

Speaking is a sensorimotor behavior whose neural basis is difficult to study at the resolution of single neurons due to the scarcity of human intracortical measurements and the lack of animal models. We recorded from electrode arrays in the 'hand knob' 30 area of motor cortex in people with tetraplegia. Neurons in this area, which have not previously been implicated in speech, modulated during speaking and during nonspeaking movement of the tongue, lips, and jaw. This challenges whether the conventional model of a 'motor homunculus' division by major body regions extends to the single-neuron scale. Spoken words and syllables could be decoded from single 35 trials, demonstrating the potential utility of intracortical recordings for brain-computer interfaces (BCIs) to restore speech. Two neural population dynamics features previously reported for arm movements were also present during speaking: a large initial condition-invariant signal, followed by rotatory dynamics. This suggests that common neural dynamical motifs may underlie movement of arm and speech 40 articulators.

Decoding spoken English phonemes from intracortical electrode arrays in dorsal precentral gyrus

“…3A,F ). This is consistent with our previous study decoding a small number of syllables spoken by T5 (Stavisky et al, 2018) and previous arm and hand BCI studies (Bouton et al, 2016;Pandarinath et al, 2017;Zhang et al, 2018) employing Utah arrays. The superior performance of HLFP decoding may reflect this neural feature more robustly capturing spiking activity from more neurons in the local vicinity of each electrode (Waldert et al, 2013) .…”

“…We did not spike sort these threshold crossing spikes ('TCs'), which may capture action potentials from one or more neurons near the electrode tip, since here we were interested in decoding the neural state rather than characterizing the properties of putative single neurons (Ajiboye et al, 2017;Brandman et al, 2018;Collinger et al, 2013;Even-Chen et al, 2018;Oby et al, 2016;Pandarinath et al, 2017;Trautmann et al, 2019) . The voltage time series were also bandpass-filtered using a 3rd order bandpass Butterworth causal filter from 125 to 5000 Hz to extract high-frequency local field potential (HLFP) signals, which have been previously shown to contain useful motor-related information (Pandarinath et al, 2017;Stavisky et al, 2018Stavisky et al, , 2019 .…”

“…However, this assumption is unlikely to hold under most models of what motor cortex encodes. Specifically, if neural activity reflects articulatory movements (Chartier et al, 2018;Lotte et al, 2015;Mugler et al, 2018;Stavisky et al, 2019) , then there may be across-phoneme differences between when articulatory movements start versus when the phoneme is clearly audible. Figure 4A presents empirical evidence of this problem by showing different time alignments to generate an example electrode's firing rates as T5 spoke different phonemes.…”

Real-Time Linear Prediction of Simultaneous and Independent Movements of Two Finger Groups Using an Intracortical Brain-Machine Interface

“…For example, primary motor cortex can simultaneously encode information about upper extremities, fingers, and speech, independent of body laterality (Cross et al, 2020;Diedrichsen et al, 2013;Heming et al, 2019;Jorge et al, 2020;Stavisky et al, 2019Stavisky et al, , 2020Willett et al, 2020).…”

Home Use of a Wireless Intracortical Brain-Computer Interface by Individuals With Tetraplegia

“…Motor BCIs aim to provide access to assistive devices by decoding user commands from electroencephalography (EEG), electrocorticography (ECoG), or intracortical signals. In ongoing clinical research, high-performance intracortical BCIs (iBCIs) are being developed to infer a user's movement intentions [1]- [8] or speech [9], [10] from neural activity recorded from one or more microelectrode arrays implanted in motor areas of cortex. By imagining natural hand, finger and arm movements, trial participants with paralysis have achieved reach-and grasp with robotic and prosthetic limbs [2], [3], [11] and their own reanimated limb [6], [12], and have demonstrated reliable cursor control for tablet use [8] and on-screen typing for communication [4], [7], [13].…”