Prediction error filtering for the extraction of abnormal intra-QRS potentials in signal-averaged electrocardiogram

The abnormal intra-QRS potentials (AIQPs) from a signal-averaged electrocardiogram (SAECG) have been proposed to indicate the risk of ventricular arrhythmias. This study developed a new method based on the prediction error filtering for the extraction of AIQPs. Two sequential SAECGs with the same noise level were used as desired and reference input signals to estimate the normal QRS and AIQPs. The simulation results showed that the prediction error filter can effectively decorrelate the AIQPs (simulated as an autoregressive stochastic process) from the normal QRS complex under an extremely poor signal-to-noise ratio. The AIQPs of VTpatients were significantly greater than those of normal subjects in leads X and Y (p<0.05).This work demonstrated that the AIQPs extracted by the prediction error filter were useful for the evaluation of the risk of ventricular arrhythmias.