Biology subjects

Au Yong, N.

Publications and source records attributed to Au Yong, N..

5 recordsLinked to original sources

bioRxiv2026

A generalizable speech neuroprosthesis

Intracortical brain-computer interfaces (BCIs) can restore communication to people with vocal tract paralysis by decoding cortical activity during attempted speech into text. State-of-the-art systems pairing neural-to-phoneme decoders with phoneme-to-word language models have achieved word error rates (WERs) as low as 1%, but only after collecting thousands of sentences of training data. Shortening the data collection process would facilitate scaling this new technology by reducing the time from device implant to high-accuracy communication. Here we introduce a transformer-based decoder model trained jointly across six intracortical speech BCI participants. For every participant -- regardless of sex, disease etiology, or attempted speaking strategy -- a multi-user model decoded speech more accurately (over 50% lower relative WER on average) than models trained on individual users data. Notably, the multi-user model could be finetuned on fewer than 200 sentences from a held-out user to achieve a WER below 7%. These results reveal how to pool intracortical data across people to yield more accurate, generalizable, and rapidly-deployable decoding models.

bioengineering↗

bioRxiv2026

Neural decoding of speech using deep neural ensembles

Speech brain-computer interfaces (BCIs) can restore rapid communication to people with paralysis, but decoding errors still limit performance. In recent brain-to-text decoding competitions, deep ensemble methods, which combine predictions from multiple independently trained decoders, have delivered striking accuracy improvements and account for the largest gains over baseline approaches. However, these methods have not previously been tested in real-time, require substantial computational resources, and their performance under various clinically relevant constraints remains poorly understood. Here, we present the first closed-loop test of deep ensembles in a participant with bilateral intracortical microelectrode arrays, demonstrating a reduction in word error rate from 33.7% to 26.0% on a large-vocabulary task. Using additional data from three participants, we then assess how these gains depend on baseline error rate, training dataset size, and ensemble size, including the resource-accuracy tradeoffs most relevant for real-world deployment. Finally, we introduce a computationally efficient pseudoensembling approach based on test-time augmentation that improves decoding accuracy while requiring only a single base decoder, greatly reducing the computational burden of ensembling. Together, these results show that the benefits of deep ensembling can be realized in real time and under practical resource constraints, bringing speech BCIs closer to broader clinical adoption.

neuroscience↗

bioRxiv2026

Oscillatory activity in rostral middle frontal gyrus and subthalamic nucleus encode proactive inhibition in cortico-subcortical motor control network

The current model of the action inhibition network includes the prefrontal cortex and the subthalamic nucleus (STN) connected via the prefrontal hyperdirect pathway. Proactive inhibition refers to preparatory mechanisms that facilitate action inhibition (i.e. enables a person to act with restraint), while reactive inhibition is a sudden stopping triggered by an external stimulus. Most research has focused on the reactive paradigm, with more limited investigation of proactive inhibition. We studied electrophysiologic activity in multiple cortical and STN regions in 17 patients with Parkinsons disease using high-resolution intracranial electrodes. Subjects performed a Go/NoGo task and a simpler Go task. Proactive inhibition was assessed by contrasting Go trials in the context of two tasks. In the rostral middle frontal gyrus, we found increased beta oscillations during movement preparation in Go trials of the Go/NoGo task compared to the Go task. A similar but weaker, preparatory beta modulation was observed in dorsal STN, while central STN was associated with significant modulation in theta power prior to movement. We interpret this activity as a reflection of the role of these regions in proactively restraining anticipated responses. Conversely, inferior frontal gyrus and ventral STN were primarily engaged during rapid post-cue action control. Specifically, withholding of action after the NoGo signal was accompanied by increased theta activity in these regions. Beta modulation within the STN mirrored those of sensorimotor cortex during successful inhibition and movement execution. In both regions, beta activity decreased during movement and was higher when movement was withheld. We conclude that communication within the hypothesized motor control network is frequency dependent, with key nodes promoting specific functions.

neuroscience↗

bioRxiv2024

Spinal interneuron population dynamics underlying flexible pattern generation

The mammalian spinal locomotor network is composed of diverse populations of interneurons that collectively orchestrate and execute a range of locomotor behaviors. Despite the identification of many classes of spinal interneurons constituting the locomotor network, it remains unclear how the networks collective activity computes and modifies locomotor output on a step-by-step basis. To investigate this, we analyzed lumbar interneuron population recordings and multi-muscle electromyography from spinalized cats performing air stepping and used artificial intelligence methods to uncover state space trajectories of spinal interneuron population activity on single step cycles and at millisecond timescales. Our analyses of interneuron population trajectories revealed that traversal of specific state space regions held millisecond-timescale correspondence to the timing adjustments of extensor-flexor alternation. Similarly, we found that small variations in the path of state space trajectories were tightly linked to single-step, microvolt-scale adjustments in the magnitude of muscle output. One sentence summaryFeatures of spinal interneuron state space trajectories capture variations in the timing and magnitude of muscle activations across individual step cycles, with precision on the scales of milliseconds and microvolts respectively.

neuroscience↗

bioRxiv2023

GABAergic synaptic scaling is triggered by changes in spiking activity rather than transmitter receptor activation.

Homeostatic plasticity represents a set of mechanisms that are thought to recover some aspect of neural function. One such mechanism called AMPAergic scaling was thought to be a likely candidate to homeostatically control spiking activity. However, recent findings have forced us to reconsider this idea as several studies suggest AMPAergic scaling is not directly triggered by changes in spiking. Moreover, studies examining homeostatic perturbations in vivo have suggested that GABAergic synapses may be more critical in terms of spiking homeostasis. Here we show results that GABAergic scaling can act to homeostatically control spiking levels. We found that perturbations which increased or decreased spiking in cortical cultures triggered multiplicative GABAergic upscaling and downscaling, respectively. In contrast, we found that changes in AMPAR or GABAR transmission only influence GABAergic scaling through their indirect effect on spiking. We propose that GABAergic scaling represents a stronger candidate for spike rate homeostat than AMPAergic scaling. Significance StatementThe nervous system maintains excitability in order to perform network behaviors when called upon to do so. Networks are thought to maintain spiking levels through homeostatic synaptic scaling, where compensatory multiplicative changes in synaptic strength are observed following alterations in cellular spike rate. Although we demonstrated that AMPAergic synaptic scaling does not appear meet these criteria as a spike rate homeostat, we now show that GABAergic scaling could play this role. Here we present evidence that the characteristics of GABAergic scaling place it in an excellent position to be a spiking homeostat. This work highlights the importance of inhibitory circuitry in the homeostatic control of excitability. Further, it provides a point of focus into neurodevelopmental disorders where excitability is impaired.

neuroscience↗