The SIESTA database and the SIESTA sleep analyzer

Abstract

Sleep research and sleep medicine require the recording of biosignals during sleep and their subsequent analysis. The sleep recording is called cardiorespiratory polysomnography. Currently the analysis of the recorded signals is performed by experienced and certified sleep technicians. In addition to visual sleep scoring many attempts had been made to develop computer assisted sleep analysis. In order to develop a computer assisted sleep analysis a systematic database with sleep recording from 200 healthy subjects and of 100 subjects with selected sleep disorders of high prevalence had been compiled as part of a European Commission funded research project. This database was the start for a normative polysomnography database and for the development of a computer based sleep analysis. The computer based sleep analysis is available as an internet service and is now used by many sleep centers for sleep research questions and clinical sleep evaluation in patients with sleep disorders.

Changes of sleep-stage transitions due to ageing and sleep disorder

“…From previous studies, it is known that there is a considerable human error in the classification of sleep stages [16]. Therefore, the Siesta project [17,18] funded by the European Community was initiated to record high-quality sleep in humans and provide an accurate and robust sleep classification. In order to compare the sleep of healthy subjects to clinical findings, the Siesta project recorded the sleep of patients with sleep disorders as well.…”

Validation of ‘Somnivore’, a Machine Learning Algorithm for Automated Scoring and Analysis of Polysomnography Data

“…Automated analysis protocols have been attempted extensively throughout the last five decades and have been validated using human (Svetnik et al, 2007; Sinha, 2008; Anderer et al, 2010; Nigro et al, 2011; Penzel et al, 2011; Khalighi et al, 2012; Malhotra et al, 2013; Stepnowsky et al, 2013; Kaplan et al, 2014; Koupparis et al, 2014; Punjabi et al, 2015; Wang et al, 2015; Hassan and Bhuiyan, 2017; Koolen et al, 2017; Sun et al, 2017) or rodent data (Crisler et al, 2008; Gross et al, 2009; Stephenson et al, 2009; Rytkonen et al, 2011; McShane et al, 2013; Sunagawa et al, 2013; Bastianini et al, 2014; Kreuzer et al, 2015; Rempe et al, 2015; Gao et al, 2016), with some success, but generally limited adoption by the field. The reasons for this mixed profile are numerous and include: rigidity in the classification, which is unable to accommodate individual differences in polysomnographic data; inadequate ‘user-friendliness’ for users not proficient in software engineering; and inadequate validation, rarely using ‘non-control’ subjects, and not analyzing biological end-measures, limiting metrics to those solely used by computational engineers or statisticians.…”

The Future of Sleep Staging, Revisited

“…Attempts at such a standardised database have been implemented, eg the SIESTA project 15 , 16 and the development of the European Data Format, 17 but, however worthy these endeavours, their impact has been somewhat limited.…”