No dream this business of sleep
Measuring sleep patterns requires the monitoring of several body/brain signal rhythms such as EEG and ECG in order to comprehend the variables of sleep analysis. While machines are extremely helpful in registering the brain's activity, they do experience external interruptions such as artefacts. Artefacts impair data integrity and compound the problem of measuring brain activity levels as not all artefact signals may be seen for what they are. Additionally, some brain signals are so minor, that some analysts may not easily detect their occurrences. Another problem with compiling data on sleep patterns is that many rules or signal patterns are open to human interpretation. Currently, existing standards on sleep interpretation are proving increasingly disappointing for clinicians. With this in mind, an enterprising research undertook means by which to standardise as well as regulate the architecture of human nocturnal sleep. To do so, they developed an automatic sleep analyser that integrates polysomnography readings into a reliable, reproducible computer based system, not open to the faults of human interpretation. The analyser is based on the assumption that sleep is described as a continuum of three different processes; wakefulness, slow-wave sleep and non-slow-wave sleep (including REM). Validation against a large database of healthy subjects of all adult ages and troubled sleep patients was performed, including subjective and objective measurements of sleep quality. With state-of-the-art methodologies for signal processing and data modelling, the system is able to provide information on possible artefact input. The system requires improved sensitivity to signal artefact detection and recordings on laboratory settings, but the prototype has still provided novel information about sleep quality. The developers are looking to validate and embed the software code into a commercial tool before being offered on a commercial basis.
