Automatic polysomnographic apnea analysis in childhood

Whereas in adults the sleep apnea syndrome is a well-defined illness, in childhood and infancy the relevance of different types of apneas is discussed controversially. In paediatric sleep laboratories experts usually decide more or less intuitively based on slopes of different important parameters measured by polysomnography. During the routine analysis of a night recording generally hundreds of apneas have to be classified. In order to accelerate this time consuming process and to emphasize the relevance of different apnea events we introduce a new apnea severity factor ASF which is calculated automatically. The algorithm is based on the expertise of four well trained physicians. at implies the examination of type and duration of each single apnea as well as the changes in oxygen saturation and heart rate variation. This apnea severity profile offers a non-subjective method of decision making and enables the physician to diagnose severe events in greater detail. For validation this method was applied to 9 newborn infants with bronchopulmonary dysplasia and to 11 children of a control group in a total amount of 32 apneas. It allowed a classification of respiratory disturbances by the first glance.