Biology subjects

Chestek, C. A.

Publications and source records attributed to Chestek, C. A..

At least 19 recordsLinked to original sources

bioRxiv2025

A method for efficient, rapid, and minimally invasive implantation of individual non-functional motes with penetrating subcellular-diameter carbon fiber electrodes into rat cortex

ObjectiveDistributed arrays of wireless neural interfacing chips with 1-2 channels each, known as "neural dust", could enhance brain machine interfaces (BMIs) by removing the wired connection through the scalp and increasing biocompatibility with their submillimeter size. Although several approaches for neural dust have emerged, a procedure for implanting them in batches that builds upon the safety and performance of currently used electrodes remains to be demonstrated. ApproachHere, we demonstrate the feasibility of implanting batches of wireless motes that rest on the cortical surface with carbon fiber electrodes of subcellular diameter (6.8-8.4 {micro}m) that penetrate to a target brain depth of 1 mm without insertion aids. To simulate their implantation, we assembled more than 230 mechanically equivalent motes and affixed them to insertion tools with polyethylene glycol (PEG), a quickly dissolvable and biocompatible material. Then, we implanted mote grids of multiple configurations into rat cortex in vivo and evaluated insertion success and their arrangement on the brain surface using photos and videos captured during their implantation. Main ResultsWhen placing motes onto the insertion device, we found that they aggregated in molten PEG such that the array pitch was only 5% wider than the dimensions of the mote bases themselves (240 x 240 {micro}m). Overall, we found that motes with this arrangement could be inserted into rat cortex with a high success rate, as 171/186 (92%) motes in 4x4 (N=4) and 5x5 (N=5) square grid configurations were successfully inserted using the insertion device alone. After implantation, measurements of how much motes tilted (22{+/-}9{degrees}, [X]{+/-}S) and had been displaced relative to their original positions were smaller than those measured for devices implanted inside the brain in the literature. SignificanceCollectively, these data establish the viability of safely implementing motes with ultrasmall electrodes and epicortically-situated chips for use in future BMIs.

bioengineering↗

bioRxiv2025

Exploring Synergies in Brain-Machine Interfaces: Compression vs. Performance

Individuals with severe neurological injuries often rely on assistive technologies, but current methods have limitations in accurately decoding multi-degree-of-freedom (DoF) movements. Intracortical brain-machine interfaces (iBMIs) use neural signals to provide a more natural control method, but currently struggle with higher-DoF movements--something the brain handles effortlessly. It has been theorized that the brain simplifies high-DoF movement through muscle synergies, which link multiple muscles to function as a single unit. These synergies have been studied using dimensionality reduction techniques like principal component analysis (PCA), non-negative matrix factorization (NMF), and demixed PCA (dPCA) and successfully used to reduce noise and improve offline decoder stability in non-invasive applications. However, their effectiveness in improving decoding and generalizability for implanted recordings across varied tasks is unclear. Here, we evaluated if brain and muscle synergies can enhance iBMI performance in non-human primates performing a two-DoF finger task. Specifically, we tested if PCA, dPCA, and NMF could compress and denoise brain and muscle data and improve decoder generalization across tasks. Our results showed that while all methods effectively compressed data with minimal loss in decoding accuracy, none improved performance through denoising. Additionally, none of the methods enhanced generalization across tasks. These findings suggest that while dimensionality reduction can aid data compression, alone it may not reveal the "true" control space needed to improve decoder performance or generalizability. Further research is required to determine whether synergies are the optimal control framework or if alternative approaches are required to enhance decoder robustness in iBMI applications. Significance StatementMany researchers believe that brain and muscle synergies represent a fundamental control strategy and could enhance brain-machine interface (BMI) decoding performance. These synergies, extracted through dimensionality reduction techniques, are thought to simplify complex neural data, improving the efficiency and accuracy of BMI systems. In our study, we evaluated brain and muscle synergies in a dexterous finger task. We found that while these synergies effectively compressed high-dimensional data, they did not improve performance through denoising or generalize well across different contexts. Instead, the highest performance was achieved when using all available data, suggesting that synergies, although useful for data compression, may not provide the "true" control space needed to enhance decoder robustness or adaptability in implanted BMI systems.

neuroscience↗

bioRxiv2024

Sparse neural networks enable low-power, implantable neural interfaces

Recent advances in brain-machine interfaces (BMIs) using neural network decoders and increased channel count have improved the restoration of speech and motor function, but at the cost of higher power consumption. For wireless, implantable BMIs to be clinically viable, power consumption must be limited to prevent thermal tissue damage and enable long use without frequent charging. Here, we show how neural network "pruning" creates sparse decoders that require fewer computations and active channels for reduced power consumption. Across multiple movement decoding tasks using brain and muscle signals, recurrent neural network decoders can be compressed by over 100x while maintaining strong performance, enabling decoding on the implant with <1% power increase compared to decoding externally. Pruning also allows for deactivating up to 89% of channels, reducing BMI power by up to 5x. Counterintuitively, our findings suggest that BMIs employing a subset of a larger number of channels may achieve lower power consumption than BMIs with fewer channels, for a given performance level. These results suggest a path toward power-efficient, implantable BMIs suitable for long-term clinical use.

bioengineering↗

bioRxiv2024

Exploring the trade-off between deep-learning and explainable models for brain-machine interfaces

People with brain or spinal cord-related paralysis often need to rely on others for basic tasks, limiting their independence. A potential solution is brain-machine interfaces (BMIs), which could allow them to voluntarily control external devices (e.g., robotic arm) by decoding brain activity to movement commands. In the past decade, deep-learning decoders have achieved state-of-the-art results in most BMI applications, ranging from speech production to finger control. However, the black-box nature of deep-learning decoders could lead to unexpected behaviors, resulting in major safety concerns in real-world physical control scenarios. In these applications, explainable but lower-performing decoders, such as the Kalman filter (KF), remain the norm. In this study, we designed a BMI decoder based on KalmanNet, an extension of the KF that augments its operation with recurrent neural networks to compute the Kalman gain. This results in a varying "trust" that shifts between inputs and dynamics. We used this algorithm to predict finger movements from the brain activity of two monkeys. We compared KalmanNet results offline (pre-recorded data, n = 13 days) and online (real-time predictions, n = 5 days) with a simple KF and two recent deep-learning algorithms with state-of-the-art results: tcFNN and LSTM. KalmanNet achieved comparable or better results than other deep learning models in offline and online modes, relying on the dynamical model for stopping while depending more on neural inputs for initiating movements. We further validated this mechanism by implementing a heteroscedastic KF that used the same strategy, and it also approached state-of-the-art performance while remaining in the explainable domain of standard KFs. However, we also see two downsides to KalmanNet. KalmanNet shares the limited generalization ability of existing deep-learning decoders, and its usage of the KF as an inductive bias limits its performance in the presence of unseen noise distributions. Despite this trade-off, our analysis successfully integrates traditional controls and modern deep-learning approaches to motivate high-performing yet still explainable BMI designs.

neuroscience↗

bioRxiv2024

Single unit electrophysiology recordings and computational modeling can predict octopus arm movement

The octopus simplified nervous system holds the potential to reveal principles of motor circuits and improve brain-machine interface devices through computational modeling with machine learning and statistical analysis. Here, an array of carbon electrodes providing single-unit electrophysiology recordings were implanted into the octopus anterior nerve cord. The number of spikes and arm movements in response to stimulation at different locations along the arm were recorded. We observed that the number of spikes occurring within the first 100ms after stimulation were predictive of the resultant movement response. Computational models showed that temporal electrophysiological features could be used to predict whether an arm movement occurred with 88.64% confidence, and if it was a lateral arm movement or a grasping motion with 75.45% confidence. Both supervised and unsupervised methods were applied to gain streaming measurements of octopus arm movements and how their motor circuitry produces rich movement types in real time. Deep learning models and unsupervised dimension reduction identified a consistent set of features that could be used to distinguish different types of arm movements. These models generated predictions for how to evoke a particular, complex movement in an orchestrated sequence for an individual motor circuit.

bioengineering↗

bioRxiv2024

Few-shot Algorithms for Consistent Neural Decoding (FALCON) Benchmark

Intracortical brain-computer interfaces (iBCIs) can restore movement and communication abilities to individuals with paralysis by decoding their intended behavior from neural activity recorded with an implanted device. While this activity yields high-performance decoding over short timescales, neural data are often nonstationary, which can lead to decoder failure if not accounted for. To maintain performance, users must frequently recalibrate decoders, which requires the arduous collection of new neural and behavioral data. Aiming to reduce this burden, several approaches have been developed that either limit recalibration data requirements (few-shot approaches) or eliminate explicit recalibration entirely (zero-shot approaches). However, progress is limited by a lack of standardized datasets and comparison metrics, causing methods to be compared in an ad hoc manner. Here we introduce the FALCON benchmark suite (Few-shot Algorithms for COnsistent Neural decoding) to standardize evaluation of iBCI robustness. FALCON curates five datasets of neural and behavioral data that span movement and communication tasks to focus on behaviors of interest to modern-day iBCIs. Each dataset includes calibration data, optional few-shot recalibration data, and private evaluation data. We implement a flexible evaluation platform which only requires user-submitted code to return behavioral predictions on unseen data. We also seed the benchmark by applying baseline methods spanning several classes of possible approaches. FALCON aims to provide rigorous selection criteria for robust iBCI decoders, easing their translation to real-world devices. https://snel-repo.github.io/falcon/

neuroscience↗

bioRxiv2024

Functional Electrical Stimulation and Brain-Machine Interfaces for Simultaneous Control of Wrist and Finger Flexion

Brain-machine interface (BMI) controlled functional electrical stimulation (FES) is a promising treatment to restore hand movements to people with cervical spinal cord injury. Recent intracortical BMIs have shown unprecedented successes in decoding user intentions, however the hand movements restored by FES have largely been limited to predetermined grasps. Restoring dexterous hand movements will require continuous control of many biomechanically linked degrees-of-freedom in the hand, such as wrist and finger flexion, that would form the basis of those movements. Here we investigate the ability to restore simultaneous wrist and finger flexion, which would enable grasping with a controlled hand posture and assist in manipulating objects once grasped. We demonstrate that intramuscular FES can enable monkeys with temporarily paralyzed hands to move their fingers and wrist across a functional range of motion, spanning an average 88.6 degrees at the metacarpophalangeal joint flexion and 71.3 degrees of wrist flexion, and intramuscular FES can control both joints simultaneously in a real-time task. Additionally, we demonstrate a monkey using an intracortical BMI to control the wrist and finger flexion in a virtual hand, both before and after the hand is temporarily paralyzed, even achieving success rates and acquisition times equivalent to able-bodied control with BMI control after temporary paralysis in two sessions. Together, this outlines a method using an artificial brain-to-body interface that could restore continuous wrist and finger movements after spinal cord injury. One Sentence SummaryWe show in non-human primates that intramuscular functional electrical stimulation can flex the wrist and fingers through a functional range of movement, can be controlled precisely, and that a brain-machine interface can determine command signals for intended finger and wrist flexion.

neuroscience↗

bioRxiv2024

Artificial neural network for brain-machine interface consistently produces more naturalistic finger movements than linear methods

Brain-machine interfaces (BMI) aim to restore function to persons living with spinal cord injuries by decoding neural signals into behavior. Recently, nonlinear BMI decoders have outperformed previous state-of-the-art linear decoders, but few studies have investigated what specific improvements these nonlinear approaches provide. In this study, we compare how temporally convolved feedforward neural networks (tcFNNs) and linear approaches predict individuated finger movements in open and closed-loop settings. We show that nonlinear decoders generate more naturalistic movements, producing distributions of velocities 85.3% closer to true hand control than linear decoders. Addressing concerns that neural networks may come to inconsistent solutions, we find that regularization techniques improve the consistency of tcFNN convergence by 194.6%, along with improving average performance, and training speed. Finally, we show that tcFNN can leverage training data from multiple task variations to improve generalization. The results of this study show that nonlinear methods produce more naturalistic movements and show potential for generalizing over less constrained tasks. TeaserA neural network decoder produces consistent naturalistic movements and shows potential for real-world generalization through task variations.

neuroscience↗

bioRxiv2023

Sex Dependent Regional Differences in Dopamine Release Determined by Chronic Fast-Scan Cyclic Voltammetry in Nucleus Accumbens and Dorsal Striatum

We investigated sex differences in dopamine (DA) release in the nucleus accumbens (NAc) and dorsolateral striatum (DLS) using a chronic 16-channel carbon fiber electrode and fast-scan cyclic voltammetry (FSCV). Electrical stimulation (ES; 60Hz) induced DA release was recorded in the NAc of single or pair-housed male and female rats. When core (NAcC) and shell (NAcS) were recorded simultaneously, there was greater ES DA release in NAcC of pair-housed females compared with single females and males. Housing did not affect ES NAc DA release in males. In contrast, there was significantly more ES DA release from the DLS of female rats than male rats. This was true prior to and after treatment with methamphetamine. Furthermore, in castrated (CAST) males and ovariectomized (OVX) females, there were no sex differences in ES DA release from the DLS, demonstrating the hormone dependence of this sex difference. However, in the DLS of both intact and gonadectomized rats, DA reuptake was slower in females than in males. Finally, DA release following ES of the medial forebrain bundle at 60Hz was studied over four weeks. ES DA release increased over time for both CAST males and OVX females, demonstrating sensitization. Using this novel 16-channel chronic FSCV electrode, we found sex differences in the effects of social housing in the NAcS, sex differences in DA release from intact rats in DLS, sex differences in DA reuptake in DLS of intake and gonadectomized rats, and we report sensitization of ES-induced DA release in DLS in vivo. Significance StatementDopamine release is not uniform or fixed. In the nucleus accumbens, pair housing, compared with individual housing, is shown to differentially affect dopamine responsiveness to stimulation in a sex-dependent and region-specific way. There are also sex differences in stimulated dopamine release in the dorsolateral striatum of intact rats, which are not seen in gonadectomized rats, indicating the hormone dependence of this sex difference. However, reuptake of dopamine was slower in females than in males, independent of gonadal hormones. Importantly, the electrical stimulation-induced dopamine release in the dorsolateral striatum of gonadectomized rats demonstrated sensitization of dopamine release in vivo within animals for the first time. Thus, stimulated dopamine release exhibits sex-specific neuroplasticity that is modified in females by the housing conditions.

neuroscience↗

bioRxiv2023

Balancing Memorization and Generalization in RNNsfor High Performance Brain-Machine Interfaces

Brain-machine interfaces (BMIs) can restore motor function to people with paralysis but are currently limited by the accuracy of real-time decoding algorithms. Recurrent neural networks (RNNs) using modern training techniques have shown promise in accurately predicting movements from neural signals but have yet to be rigorously evaluated against other decoding algorithms in a closed-loop setting. Here we compared RNNs to other neural network architectures in real-time, continuous decoding of finger movements using intracortical signals from nonhuman primates. Across one and two finger online tasks, LSTMs (a type of RNN) outperformed convolutional and transformer-based neural networks, averaging 18% higher throughput than the convolution network. On simplified tasks with a reduced movement set, RNN decoders were allowed to memorize movement patterns and matched able-bodied control. Performance gradually dropped as the number of distinct movements increased but did not go below fully continuous decoder performance. Finally, in a two-finger task where one degree-of-freedom had poor input signals, we recovered functional control using RNNs trained to act both like a movement classifier and continuous decoder. Our results suggest that RNNs can enable functional real-time BMI control by learning and generating accurate movement patterns.

bioengineering↗

bioRxiv2022

Profiling neurons surrounding subcellular-scale carbon fiber electrode tracts enables modeling recorded signals

Characterizing the relationship between neuron spiking and the signals electrodes record is vital to defining the neural circuits driving brain function and informing computational modeling. However, electrode biocompatibility and precisely localizing neurons around the electrodes are critical to defining this relationship. Here, we show the ability to localize post-explant recording tips of subcellular-scale carbon fiber electrodes and surrounding neurons. Immunostaining of astrocyte, microglia, and neuron markers confirmed improved tissue health. While neurons near implants were stretched, their number and distribution were similar to control, suggesting that these minimally invasive electrodes demonstrate the potential to sample naturalistic neural populations. This motivated prediction of the spikes produced by neurons nearest to the electrodes using a model fit with recorded electrophysiology. These simulations show the first direct evidence that neuron placement in the immediate vicinity of the recording site influences how many spike clusters can be reliably identified by spike sorting.

neuroscience↗

bioRxiv2022

Utah Array Characterization and Histological Analysis of a Multi-Year Implant in Non-Human Primate Motor and Sensory Cortices

The Utah array is widely used in both clinical studies and neuroscience. It has a strong track record of safety. However, it is also known that implanted electrodes promote the formation of scar tissue in the immediate vicinity of the electrodes, which negatively impacts the ability to record neural waveforms. This scarring response has been primarily studied in rats and mice, which may have a very different response than primate brain. Here, we present a rare nonhuman primate histological dataset (n=1 rhesus macaque) obtained 848 and 590 days after implantation in two brain hemispheres. For 2 of 4 arrays that remained within the cortex, NeuN was used to stain for neuron somata at 3 different electrode depths. Images were filtered and denoised, with neurons then counted in the vicinity of the arrays as well as a nearby section of control tissue. Additionally, 3 of 4 arrays were imaged with a scanning electrode microscope (SEM) to evaluate any materials damage that might be present. Overall, we found a 63% percent reduction in the number of neurons surrounding the electrode compared to control areas. In terms of materials, the arrays remained largely intact with metal and Parylene C present, though tip breakage and cracks were observed on many electrodes. Overall, these results suggest that the tissue response in the nonhuman primate brain shows similar neuron loss to previous studies using rodents. Electrode improvements, for example using smaller or softer probes, may therefore substantially increase the neuronal recording yield in primate cortex.

bioengineering↗

bioRxiv2022

The Impact of Task Context on Predicting Finger Movements in a Brain-Machine Interface

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 patients 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 predicting finger kinematics and finger-related muscle activations across contexts led to significant increases in prediction error, especially for muscle activations. However, with respect to online BMI control of the virtual hand, changing either training task context or the hands 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, possibly explaining biased kinematic predictions and suggesting a feature that could help predict different magnitude muscle activations while producing similar kinematics.

neuroscience↗

bioRxiv2022

Automated assembly of high-density carbon fiber electrode arrays for single unit electrophysiological recordings

ObjectiveCarbon fiber (CF) is good for chronic neural recording due to the small diameter (7 {micro}m), high Youngs modulus, and low electrical resistance, but most high-density carbon fiber (HDCF) arrays are manually assembled with labor-intensive procedures and limited by the accuracy and repeatability of the operator handling. A machine to automate the assembly is desired. ApproachThe HDCF array assembly machine contains: 1) a roller-based CF extruder, 2) a motion system with three linear and one rotary stages, 3) an imaging system with two digital microscope cameras, and 4) a laser cutter. The roller-based extruder automatically feeds single CF as raw material. The motion system aligns the CF with the array backend then places it. The imaging system observes the relative position between the CF and the backend. The laser cutter cuts off the CF. Two image processing algorithms are implemented to align the CF with the support shanks and circuit connection pads. Main resultsThe machine was capable of precisely handling 6.8 m carbon fiber electrodes. Each electrode was placed into a 12 m wide trenches in a silicon support shank. Two HDCF arrays with 16 CF electrodes populated on 3 mm shanks (with 80 m pitch) were fully assembled. Impedance measurements were found to be in good agreement with manual assembled arrays. One HDCF array was implanted in the motor cortex in an anesthetized rat and was able to detect single unit activity. SignificanceThis machine can eliminate the manual labor-intensive handling, alignment and placement of single carbon fiber during assembly, providing a proof-of-concepts towards fully automated HDCF array assembly and batch production.

bioengineering↗

bioRxiv2022

Brain-Controlled Electrical Stimulation Restores Continuous Finger Function

Brain-machine interfaces have shown promise in extracting upper extremity movement intention from the thoughts of nonhuman primates and people with tetraplegia. Attempts to restore a users own hand and arm function have employed functional electrical stimulation (FES), but most work has restored discrete grasps. Little is known about how well FES can control continuous finger movements. Here, we use a low-power brain-controlled functional electrical stimulation (BCFES) system to restore continuous volitional control of finger positions to a monkey with a temporarily paralyzed hand. In a one-dimensional, continuous, finger-related target acquisition task, the monkey improved his success rate to 83% (1.5s median acquisition time) when using the BCFES system during temporary paralysis from 8.8% (9.5s median acquisition, equivalent to chance) when attempting to use his temporarily paralyzed hand. With two monkeys under general anesthesia, we found FES alone could control the monkeys fingers to rapidly reach targets in a median 1.1s but caused oscillation about the target. Finally, when attempting to perform a virtual two-finger continuous target acquisition task in brain-control mode following temporary hand paralysis, we found performance could be completely recovered by executing recalibrated feedback-intention training one time following temporary paralysis. These results suggest that BCFES can restore continuous finger function during temporary paralysis using existing low-power technologies and brain-control may not be the limiting performance factor in a BCFES neuroprosthesis.

neuroscience↗

bioRxiv2022

A low-power communication scheme for wireless, 1000 channel brain-machine interfaces

ObjectiveBrain-machine interfaces (BMIs) have the potential to restore motor function but are currently limited by electrode count and long-term recording stability. These challenges may be solved through the use of free-floating "motes" which wirelessly transmit recorded neural signals, if power consumption can be kept within safe levels when scaling to thousands of motes. Here, we evaluated a pulse-interval modulation (PIM) communication scheme for infrared (IR)-based motes that aims to reduce the wireless data rate and system power consumption. ApproachTo test PIMs ability to efficiently communicate neural information, we simulated the communication scheme in a real-time closed-loop BMI with non-human primates. Additionally, we performed circuit simulations of an IR-based 1000-mote system to calculate communication accuracy and total power consumption. Main ResultsWe found that PIM at 1kb/s per channel maintained strong correlations with true firing rate and matched online BMI performance of a traditional wired system. Closed-loop BMI tests suggest that lags as small as 30 ms can have significant performance effects. Finally, unlike other IR communication schemes, PIM is feasible in terms of power, and neural data can accurately be recovered on a receiver using 3mW for 1000 channels. SignificanceThese results suggest that PIM-based communication could significantly reduce power usage of wireless motes to enable higher channel-counts for high-performance BMIs.

bioengineering↗

bioRxiv2022

Ultra-flexible and Stretchable Intrafascicular Peripheral Nerve Recording Device with Axon-dimension, Cuff-less Microneedle Electrode Array

Peripheral nerve mapping tools with higher spatial resolution are needed to advance systems neuroscience, and potentially provide a closed-loop biomarker in neuromodulation applications. Two critical challenges of microscale neural interfaces are (i) how to apply them to small peripheral nerves, and (ii) how to minimize chronic reactivity. We developed a flexible microneedle nerve array (MINA), which is the first high-density penetrating electrode array made with axon-sized silicon microneedles embedded in low-modulus thin silicone. We present the design, fabrication, acute recording, and chronic reactivity to an implanted MINA. Distinctive units were identified in the rat peroneal nerve. We also demonstrate a long-term, cuff-free, and suture-free fixation manner using rose bengal as a light-activated adhesive for two timepoints. The tissue response at 1-week included a sham (N=5) and MINA-implanted (N=5) group, and the response at 6-week also included a sham (N=3) and MINA-implanted (N=4) group. These conditions were quantified in the left vagus nerve of rats using histomorphometry. Micro-CT was added to visualize and quantify tissue encapsulation around the implant. MINA demonstrated a reduction in encapsulation thickness over previously quantified interfascicular methods. Future challenges include techniques for precise insertion of the microneedle electrodes and demonstrating long-term recording.

neuroscience↗

bioRxiv2021

Real-Time Brain-Machine Interface Achieves High-Velocity Prosthetic Finger Movements using a Biologically-Inspired Neural Network Decoder

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, loosely inspired by the biological neural pathway, to decode real-time two-degree-of-freedom finger movements. Using a two-step training method, a recalibrated feedback intention-trained (ReFIT) neural network achieved a higher throughput with higher finger velocities and more natural appearing finger movements than the ReFIT Kalman filter, which represents the current standard. The neural network decoders introduced herein are the first to 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.

neuroscience↗