Acquisition and Classification of heart rate variability using time-frequency representation.

It has been shown that the heart rate varies not only in relation to the cardiac demand but is also affected by the presence of cardiac disease and diabetes. Furthermore, it has been shown that heart rate variability may be used as an early indicator of cardiac disease susceptibility and the presence of diabetes. Therefore, the heart rate variability may be used for early clinical screening of these diseases. In order to reliably assess the patient's condition, the heart rate variability information is determined from an electrocardiogram data acquisition system. Once collected, the heart rate variability signal is characterised and used as a basis for classification. This study details the development of a heart rate variability data acquisition system, method of collecting known patient data, and design of a signal-processing algorithm that characterises heart rate variability information to be used as a basis for patient classification. Specifically, six sets of 5 minute electrocardiogram signals are collected by a personal computer based data acquisition system in a clinical setting. Consecutive R-wave deflections are detected from the electrocardiogram and used to determine the individual heart beat intervals. The outlying measurements are then removed and the remaining data is interpolated. The processed data is then characterised using time-frequency analysis and specific features are determined. Lastly, these features are used as a basis in a classification system. The results are then compared to the known patient conditions and the effectiveness of the screening procedure is determined.