Despite the rapid progress and interest in brain-machine interfaces that restore motor function, the performance of prosthetic fingers and limbs has yet to mimic native function. The algorithm that converts brain signals to a control signal for the prosthetic device is one of the limitations in achieving rapid and realistic finger movements. To achieve more realistic finger movements, we developed a shallow feed-forward neural network to decode real-time two-degree-of-freedom finger movements in two adult male rhesus macaques. Using a two-step training method, a recalibrated feedback intention–trained (ReFIT) neural network is introduced to further improve performance. In 7 days of testing across two animals, neural network decoders, with higher-velocity and more natural appearing finger movements, achieved a 36% increase in throughput over the ReFIT Kalman filter, which represents the current standard. The neural network decoders introduced herein demonstrate real-time decoding of continuous movements at a level superior to the current state-of-the-art and could provide a starting point to using neural networks for the development of more naturalistic brain-controlled prostheses.
more » « less- Award ID(s):
- 1926576
- NSF-PAR ID:
- 10379910
- Publisher / Repository:
- Nature Publishing Group
- Date Published:
- Journal Name:
- Nature Communications
- Volume:
- 13
- Issue:
- 1
- ISSN:
- 2041-1723
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract Objective. Recent advances in neural decoding have accelerated the development of brain–computer interfaces aimed at assisting users with everyday tasks such as speaking, walking, and manipulating objects. However, current approaches for training neural decoders commonly require large quantities of labeled data, which can be laborious or infeasible to obtain in real-world settings. Alternatively, self-supervised models that share self-generated pseudo-labels between two data streams have shown exceptional performance on unlabeled audio and video data, but it remains unclear how well they extend to neural decoding.Approach. We learn neural decoders without labels by leveraging multiple simultaneously recorded data streams, including neural, kinematic, and physiological signals. Specifically, we apply cross-modal, self-supervised deep clustering to train decoders that can classify movements from brain recordings. After training, we then isolate the decoders for each input data stream and compare the accuracy of decoders trained using cross-modal deep clustering against supervised and unimodal, self-supervised models.Main results. We find that sharing pseudo-labels between two data streams during training substantially increases decoding performance compared to unimodal, self-supervised models, with accuracies approaching those of supervised decoders trained on labeled data. Next, we extend cross-modal decoder training to three or more modalities, achieving state-of-the-art neural decoding accuracy that matches or slightly exceeds the performance of supervised models.Significance. We demonstrate that cross-modal, self-supervised decoding can be applied to train neural decoders when few or no labels are available and extend the cross-modal framework to share information among three or more data streams, further improving self-supervised training. -
more » « less
Abstract Objective . Advances in neural decoding have enabled brain-computer interfaces to perform increasingly complex and clinically-relevant tasks. However, such decoders are often tailored to specific participants, days, and recording sites, limiting their practical long-term usage. Therefore, a fundamental challenge is to develop neural decoders that can robustly train on pooled, multi-participant data and generalize to new participants.Approach . We introduce a new decoder, HTNet, which uses a convolutional neural network with two innovations: (a) a Hilbert transform that computes spectral power at data-driven frequencies and (b) a layer that projects electrode-level data onto predefined brain regions. The projection layer critically enables applications with intracranial electrocorticography (ECoG), where electrode locations are not standardized and vary widely across participants. We trained HTNet to decode arm movements using pooled ECoG data from 11 of 12 participants and tested performance on unseen ECoG or electroencephalography (EEG) participants; these pretrained models were also subsequently fine-tuned to each test participant.Main results . HTNet outperformed state-of-the-art decoders when tested on unseen participants, even when a different recording modality was used. By fine-tuning these generalized HTNet decoders, we achieved performance approaching the best tailored decoders with as few as 50 ECoG or 20 EEG events. We were also able to interpret HTNet’s trained weights and demonstrate its ability to extract physiologically-relevant features.Significance . By generalizing to new participants and recording modalities, robustly handling variations in electrode placement, and allowing participant-specific fine-tuning with minimal data, HTNet is applicable across a broader range of neural decoding applications compared to current state-of-the-art decoders. -
more » « less
Abstract Objective. While brain–machine interfaces (BMIs) are promising technologies that could provide direct pathways for controlling the external world and thus regaining motor capabilities, their effectiveness is hampered by decoding errors. Previous research has demonstrated the detection and correction of BMI outcome errors, which occur at the end of trials. Here we focus on continuous detection and correction of BMI execution errors, which occur during real-time movements.Approach. Two adult male rhesus macaques were implanted with Utah arrays in the motor cortex. The monkeys performed single or two-finger group BMI tasks where a Kalman filter decoded binned spiking-band power into intended finger kinematics. Neural activity was analyzed to determine how it depends not only on the kinematics of the fingers, but also on the distance of each finger-group to its target. We developed a method to detect erroneous movements, i.e. consistent movements away from the target, from the same neural activity used by the Kalman filter. Detected errors were corrected by a simple stopping strategy, and the effect on performance was evaluated.Main results. First we show that including distance to target explains significantly more variance of the recorded neural activity. Then, for the first time, we demonstrate that neural activity in motor cortex can be used to detect execution errors during BMI controlled movements. Keeping false positive rate belowSignificance. Neural activity recorded in motor cortex for BMI control can be used to detect and correct BMI errors and thus to improve performance. Further improvements may be obtained by enhancing classification and correction strategies. -
A key factor in the clinical translation of brain-machine interfaces (BMIs) for restoring hand motor function will be their robustness to changes in a task. With functional electrical stimulation (FES) for example, the patient’s own hand will be used to produce a wide range of forces in otherwise similar movements. To investigate the impact of task changes on BMI performance, we trained two rhesus macaques to control a virtual hand with their physical hand while we added springs to each finger group (index or middle-ring-small) or altered their wrist posture. Using simultaneously recorded intracortical neural activity, finger positions, and electromyography, we found that decoders trained in one context did not generalize well to other contexts, leading to significant increases in prediction error, especially for muscle activations. However, with respect to online BMI control of the virtual hand, changing either the decoder training task context or the hand’s physical context during online control had little effect on online performance. We explain this dichotomy by showing that the structure of neural population activity remained similar in new contexts, which could allow for fast adjustment online. Additionally, we found that neural activity shifted trajectories proportional to the required muscle activation in new contexts. This shift in neural activity possibly explains biases to off-context kinematic predictions and suggests a feature that could help predict different magnitude muscle activations while producing similar kinematics.more » « less
-
more » « less
Abstract Objective. Extracting signals directly from the motor system poses challenges in obtaining both high amplitude and sustainable signals for upper-limb neuroprosthetic control. To translate neural interfaces into the clinical space, these interfaces must provide consistent signals and prosthetic performance.Approach. Previously, we have demonstrated that the Regenerative Peripheral Nerve Interface (RPNI) is a biologically stable, bioamplifier of efferent motor action potentials. Here, we assessed the signal reliability from electrodes surgically implanted in RPNIs and residual innervated muscles in humans for long-term prosthetic control.Main results. RPNI signal quality, measured as signal-to-noise ratio, remained greater than 15 for up to 276 and 1054 d in participant 1 (P1), and participant 2 (P2), respectively. Electromyography from both RPNIs and residual muscles was used to decode finger and grasp movements. Though signal amplitude varied between sessions, P2 maintained real-time prosthetic performance above 94% accuracy for 604 d without recalibration. Additionally, P2 completed a real-world multi-sequence coffee task with 99% accuracy for 611 d without recalibration.Significance. This study demonstrates the potential of RPNIs and implanted EMG electrodes as a long-term interface for enhanced prosthetic control.
