Combining ECG Morphology and Heart Rate Parameters for the Detection of Sleep Disorder Breathing Events


M. Palaniswami;
Post Doctoral Research Fellow: Ahsan Khandoker.

Student: Kris Nilsen.

Eugene Zilberg, Compumedics Ltd;
David Burton, Compumedics Ltd.

Introduction: Sleep apnoea is usually diagnosed in sleep laboratories through sleep studies. These procedures are typically expensive and require the patient to sleep the night in the laboratory. If sleep apnoea could be diagnosed by examining the ECG signal, the diagnostic recording could be made in the patient?s home using standard ECG monitoring technology particularly taking into account large obstructive sleep apnoea prevalence among cardiac patients.
Significance: It is hypothesised that analysis of the morphology of the ECG and the combination of other analytic techniques such as heart rate analysis could lead to the detection of actual start and end of apnoea/hypopnoea events and therefore the calculation of the apnoea-hypopnoea index. Regression modelling techniques were used to investigate the relevance of ECG waveform morphology and heart rate parameters that were identified visually. Statistical analysis showed that all the physiological predictors were significant.
Applications: The multivariate analysis demonstrated that the combination of parameters outperforms any single parameter model and the models based on only heart rate parameters or only the ECG waveform morphology parameters.
Challenges: The multivariate analysis also showed encouraging results indicating that an algorithm using a combination of heart rate parameters and ECG morphology parameters could be constructed that would enable the detection of actual events and could be used to generate an apnoea-hypopnoea index.
top of page