A system for rapid identification of respiratory abnormalities using a neural network

Potential victims of Sudden Infant Death Syndrome (SIDS) can usefully be monitored in the home environment. Conventional respiration movement monitors can be helpful but may not detect potentially dangerous hypoxaemic episodes. Thus oxygen monitoring is to be preferred but can be difficult to use in the home. In an attempt to overcome these difficulties this paper presents the results of an exploratory experiment into the use of a neural network to link the output of a respiration pressure monitor to the classification of breathing patterns as effective or otherwise. It has been shown that it is possible to predict changes in oxygen saturation, which could signify potentially dangerous episodes earlier than when other methods are employed.