A model of breathing abnormalities in sleep for development of classification and diagnosis techniques

Following the recognition of the obstructive sleep apnoea (OSA) disease, snoring achieved a completely new status from being a mere social problem to an important clinical disorder. Using modern signal analysis equipment and software enables breath-by-breath investigation of snoring and opens possibilities for automatic detection of specific acoustic events. The simulation model proposed in this study provides a repeat-able test-bed for signal processing techniques that are under development to aid clinical diagnosis of underlying pathological problems and to identify periods of sleep apnea that may develop into more serious conditions. Noise and artifacts can be added to the simulated data for assessment of algorithm robustness prior to application to clinical data that is being recorded non-invasively and remotely.