Sleep apnoea analysis from neural network post-processing
Siegwart DK., Tarassenko L., Roberts SJ., Stradling JR., Partlett J.
Methods of analysis of electroencephalogram (EEG) signals using artificial neural networks are presented, as well as subsequent methods of detection of obstructive sleep apnoea (OSA). The methods of detection described are based on post-processing sleep state probabilities obtained with a 10-6-4 multi-layer perceptron. Results show periodic changes in sleep states highly correlated with changes in blood oxygen saturation.