The aim of this study is to develop an automated algorithm to predict the onset of PAF. The markers of ECG trace prior to the onset of PAF will be investigated with wavelet transform. In addition to, the characteristic features will be determined from wavelet scalogram in both temporal and spectral analysis. The expected results will be useful in diagnosis of cardiac arrhythmias in particular PAF rhythm. An early hospitalization may reduce number of cardiac arrest victims. The results may be implemented as medical support device for arrhythmia treatment that is very useful for improving the anti-arrhythmia therapy.
