Advanced sleep spindle identification with neural networks

• DOI: 10.1038/s41598-022-11210-y
• Publication Date: 2022-05-10T19:02:43Z
• Authors: Lars Kaulen; Justus T. C. Schw...; Jules Schneider; Philipp Ritter; Stephan Bialonski
• Keywords: Signal Processing (eess.SP); FOS: Computer and information sciences; 0301 basic medicine; Computer Science - Machine Learning; Science; Q; R; Reproducibility of Results; Electroencephalography; Quantitative Biology - Quantitative Methods; Article; Machine Learning (cs.LG); 03 medical and health sciences; FOS: Biological sciences; FOS: Electrical engineering, electronic engineering, information engineering; Medicine; Humans; Neural Networks, Computer; Sleep Stages; Electrical Engineering and Systems Science - Signal Processing; Sleep; Data Curation; Quantitative Methods (q-bio.QM)

AbstractSleep spindles are neurophysiological phenomena that appear to be linked to memory formation and other functions of the central nervous system, and that can be observed in electroencephalographic recordings (EEG) during sleep. Manually identified spindle annotations in EEG recordings suffer from substantial intra- and inter-rater variability, even if raters have been highly trained, which reduces the reliability of spindle measures as a research and diagnostic tool. The Massive Online Data Annotation (MODA) project has recently addressed this problem by forming a consensus from multiple such rating experts, thus providing a corpus of spindle annotations of enhanced quality. Based on this dataset, we present a U-Net-type deep neural network model to automatically detect sleep spindles. Our model’s performance exceeds that of the state-of-the-art detector and of most experts in the MODA dataset. We observed improved detection accuracy in subjects of all ages, including older individuals whose spindles are particularly challenging to detect reliably. Our results underline the potential of automated methods to do repetitive cumbersome tasks with super-human performance.