Neuro-BERT: Rethinking Masked Autoencoding for Self-supervised Neurological Pretraining
Authors:
Di Wu,
Siyuan Li,
Jie Yang,
Mohamad Sawan
Abstract:
Deep learning associated with neurological signals is poised to drive major advancements in diverse fields such as medical diagnostics, neurorehabilitation, and brain-computer interfaces. The challenge in harnessing the full potential of these signals lies in the dependency on extensive, high-quality annotated data, which is often scarce and expensive to acquire, requiring specialized infrastructu…
▽ More
Deep learning associated with neurological signals is poised to drive major advancements in diverse fields such as medical diagnostics, neurorehabilitation, and brain-computer interfaces. The challenge in harnessing the full potential of these signals lies in the dependency on extensive, high-quality annotated data, which is often scarce and expensive to acquire, requiring specialized infrastructure and domain expertise. To address the appetite for data in deep learning, we present Neuro-BERT, a self-supervised pre-training framework of neurological signals based on masked autoencoding in the Fourier domain. The intuition behind our approach is simple: frequency and phase distribution of neurological signals can reveal intricate neurological activities. We propose a novel pre-training task dubbed Fourier Inversion Prediction (FIP), which randomly masks out a portion of the input signal and then predicts the missing information using the Fourier inversion theorem. Pre-trained models can be potentially used for various downstream tasks such as sleep stage classification and gesture recognition. Unlike contrastive-based methods, which strongly rely on carefully hand-crafted augmentations and siamese structure, our approach works reasonably well with a simple transformer encoder with no augmentation requirements. By evaluating our method on several benchmark datasets, we show that Neuro-BERT improves downstream neurological-related tasks by a large margin.
△ Less
Submitted 3 July, 2024; v1 submitted 20 April, 2022;
originally announced April 2022.
Spatial Resolution of Local Field Potential Signals in Macaque V4
Authors:
Armin Najarpour Foroushani,
Sujaya Neupane,
Pablo De Heredia Pastor,
Christopher C. Pack,
Mohamad Sawan
Abstract:
A main challenge for the development of cortical visual prostheses is to spatially localize individual spots of light, called phosphenes, by assigning appropriate stimulating parameters to implanted electrodes. Imitating the natural responses to phosphene-like stimuli at different positions can help in designing a systematic procedure to determine these parameters. The key characteristic of such a…
▽ More
A main challenge for the development of cortical visual prostheses is to spatially localize individual spots of light, called phosphenes, by assigning appropriate stimulating parameters to implanted electrodes. Imitating the natural responses to phosphene-like stimuli at different positions can help in designing a systematic procedure to determine these parameters. The key characteristic of such a system is the ability to discriminate between responses to different positions in the visual field. While most previous prosthetic devices have targeted the primary visual cortex, the extrastriate cortex has the advantage of covering a large part of the visual field with a smaller amount of cortical tissue, providing the possibility of a more compact implant. Here, we studied how well ensembles of Multiunit activity (MUA) and Local Field Potentials (LFPs) responses from extrastriate cortical visual area V4 of a behaving macaque monkey can discriminate between two-dimensional spatial positions. We found that despite the large receptive field sizes in V4, the combined responses from multiple sites, whether MUA or LFP, has the capability for fine and coarse discrimination of positions. We identified a selection procedure that could significantly increase the discrimination performance while reducing the required number of electrodes. Analysis of noise correlation in MUA and LFP responses showed that noise correlations in LFP responses carry more information about the spatial positions. Overall, these findings suggest that spatial positions could be localized with patterned stimulation in extrastriate area V4.
△ Less
Submitted 17 November, 2019;
originally announced November 2019.
