Variations in the accuracy of the ECG based detection of obstructive sleep apnoea (OSA) for different numbers of ECG leads and categories of OSA events.

Abstract

This study investigates the finding that there is a more pronounced change to ECG physiological predictors during apnoea events compared to hypopnoea events and therefore accurate detection of hypopnoea events is likely to be more challenging than detection of apnoea events. The relevant statistical analysis was conducted by generating logistic regression models from the two data sets: the first one containing only the apnoea events and controls and the second data set containing only the hypopnoea events and controls. The discriminating ability of the model from the apnoea data set (AUC = 0.903, CI = 0.888 - 0.920) was significantly superior compared to the model from the hypopnoea data set..