SleepEDF20

SleepEDF20 is a dataset comprising a collection of polysomnographic recordings of human sleep. It was created by Ted Gleeson and is often used in research related to sleep and sleep disorders.

The dataset consists of twenty real-valued signals, each representing a variable associated with a single subject's sleep. These variables, such as EEG, EMG, and EOG, were measured by appropriate electrodes during a night of sleep. Ten of the subjects were females and ten were males. In total, the recordings include 152 nights of sleep.

The data was collected using a variety of recording equipment, including the Brain Product booth, Ag/AgCl electrodes, and the PowerLab 800 or 104 systems. Each subject's sleep was monitored and recorded over an extended period, comprising a minimum of seven nights and a maximum of nine nights.

The recordings are segmented into 30-second epochs, and each epoch is labeled as either sleep or wake. The dataset also includes timestamps for each recording, as well as subject demographics, such as age and gender. SleepEDF20 includes clinical annotations that support diagnosis, including sleep spindle, awake, and REM sleep scoring.

The dataset is available to researchers and is free for download from the University of California, Berkeley's Polysomnographic Sleep Dataset. Researchers have utilized SleepEDF20 to explore various aspects of sleep patterns and could provide a tool for others interested in studying sleep sleep disorders.