English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Neurophysiological time series, such as electromyographic signal andintracortical recordings, are typically composed of many individual spikingsources, the recovery of which can give fundamental insights into thebiological system of interest or provide neural information for man-machineinterfaces. For this reason, source separation algorithms have become anincreasingly important tool in neuroscience and neuroengineering. However, innoisy or highly multivariate recordings these decomposition techniques oftenmake a large number of errors, which degrades human-machine interfacingapplications and often requires costly post-hoc manual cleaning of the outputlabel set of spike timestamps. To address both the need for automated post-hoccleaning and robust separation filters we propose a methodology based on deepmetric learning, using a novel loss function which maintains intra-classvariance, creating a rich embedding space suitable for both label cleaning andthe discovery of new activations. We then validate this method with anartificially corrupted label set based on source-separated high-density surfaceelectromyography recordings, recovering the original timestamps even in extremedegrees of feature and class-dependent label noise. This approach enables aneural network to learn to accurately decode neurophysiological time seriesusing any imperfect method of labelling the signal.

Deep Metric Learning with Locality Sensitive Angular Loss for Self-Correcting Source Separation of Neural Spiking Signals