Adult rat vigilance states discrimination by artificial neural networks using a single EEG channel

摘要: Abstract Two multilayer neural networks were designed to discriminate vigilance states (waking, paradoxical sleep, and non-REM sleep) in the rat using a single parieto-occipital EEG derivation. After filtering (bandwidth 3.18–25 Hz) digitization at 512 Hz, signal was segmented into eight second epochs. Five variables (three statistical, two temporal) extracted from each epoch. The first network computed an epoch by classification, while also utilized contextual information contiguous A specific postprocessing procedure developed enhance state discrimination of especially sleep estimation. classifications made (with or without procedure) for six rats compared these human experts EMG informations on 63,000 High rates agreement (> 90%) between humans obtained. In view its development possibilities applicability other signals, this method could prove value biomedical research.