PSGWatch

A set of 43 parallel polysomnography (PSG) and actigraphic recordings from 43 subjects. Recordings were conducted in laboratory setting at Faculty of Physics, University of Warsaw as a part of research project aiming to construct new sleep/wake scoring algorithm operating on actigraphic data.PSG was staged by a trained expert according to AASM criteria. Only 37 out of 43 recordings were staged due to noise on some of the relevant channels in the remaining 6 recordings (majority of the channels is good quality and perfectly usable in different research).Actigraphy was collected from both hands using GeneActiv Original devices.Each PSG recording consists of two files: .raw with data from all channels and .xml with metadata describing the recording (so there are 43 pairs of .raw and .xml files). Each PSG staging is contained in .tag file (so there is 37 .tag files). To easily perform I/O operations on both PSG recordings and tags one may use python readmanager module available on PyPI. Alternatively one may use open source application SVAROG by BrainTech to browse the recordings.Actigraphic data are all packaged into actigraphy.zip.Each subject was named according to a template: PsgWatch-XX*, where XX is ID ranging from 1 to 45 (with no 37 and 38 used). All files (whether actigprahic or PSG) follow this convention.File metadata.ods contains information about subjects: handedness, age and sex.