Sleep Studies, Computer Analysis of

Abstract

Until a few years ago, the main problems for computerized sleep systems were the need for efficient data reduction and speed of analysis. The computers were too small and inefficient or too expensive for doing either waveform detection on‐line or storage of the entire raw data. Most of the systems built on the basis of general engineering tools were little used because of the poor correlation with standard visual scoring. The advent of powerful and high computing speed processor has bridged that gap. The availability of digital polysomnography has removed the technical restrictions through increased computing power, storage space, and reduced costs. However, the crude division of the sleep processes into few discrete stages and the difficulties with the translation of spatial information into appropriate algorithms remain the major issues in the development of fully automated sleep systems. The use of predetermined, fixed criteria for methods, waveforms, and states is scientifically erroneous and hazardous. Instead standards should describe the minimum requirements for the recording and analysis of the signals in terms of sampling rate, dynamic range, linearity, and documentation of the methods used. The purpose of this article is to discuss the use of computers in sleep disorders medicine and how this technology has enhanced the diagnostic and scientific capabilities of the sleep laboratory.