Sleep apnea detection using a feed-forward neural network on ECG signal

This paper presents a suitable and efficient implementation for detecting minute based analysis of sleep apnea by Electrocardiogram (ECG) signal processing. Using the PhysioNet apnea-ECG database, a median filter was applied to the recordings in order to obtain the Heart Rate Variability (HRV) and the ECG-derived respiration (EDR). The subsequent extracted features were used for training, testing and validation of a Artificial Neural Network (ANN). Training and testing sets were obtained by randomly divide the data until it reaches a good performance using a k-fold cross validation (k=10). According to results, the ANN classification has sufficient accuracy for sleep apnea detection and diagnosis (82,120%). This promising early-stage result may leads to complementary studies including alternative features selection methods and/or other classification models.