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

Background: Wide-field calcium imaging (WFCI) with genetically encodedcalcium indicators allows for spatiotemporal recordings of neuronal activity inmice. When applied to the study of sleep, WFCI data are manually scored intothe sleep states of wakefulness, non-REM (NREM) and REM by use of adjunct EEGand EMG recordings. However, this process is time-consuming, invasive and oftensuffers from low inter- and intra-rater reliability. Therefore, an automatedsleep state classification method that operates on spatiotemporal WFCI data isdesired. New Method: A hybrid network architecture consisting of aconvolutional neural network (CNN) to extract spatial features of image framesand a bidirectional long short-term memory network (BiLSTM) with attentionmechanism to identify temporal dependencies among different time points wasproposed to classify WFCI data into states of wakefulness, NREM and REM sleep.Results: Sleep states were classified with an accuracy of 84% and Cohen's kappaof 0.64. Gradient-weighted class activation maps revealed that the frontalregion of the cortex carries more importance when classifying WFCI data intoNREM sleep while posterior area contributes most to the identification ofwakefulness. The attention scores indicated that the proposed network focuseson short- and long-range temporal dependency in a state-specific manner.Comparison with Existing Method: On a 3-hour WFCI recording, the CNN-BiLSTMachieved a kappa of 0.67, comparable to a kappa of 0.65 corresponding to thehuman EEG/EMG-based scoring. Conclusions: The CNN-BiLSTM effectively classifiessleep states from spatiotemporal WFCI data and will enable broader applicationof WFCI in sleep.

Attention-Based CNN-BiLSTM for Sleep State Classification of Spatiotemporal Wide-Field Calcium Imaging Data