Analysis of unpredictable intra-QRS potentials based on multi-step linear prediction modeling for evaluating the risk of ventricular arrhythmias

The abnormal intra-QRS potentials (AIQP) have been proposed to evaluate the risk of ventricular arrhythmias. However, the extremely low signal-to-noise ratio (AIQP vs. normal QRS) and the overlap between AIQP and normal QRS would limit the accuracy of AIQP estimation. This study proposed a multi-step linear prediction modeling to estimate all of the unpredictable intra-QRS potentials (UIQP) for evaluating the risk of ventricular arrhythmias. The simulation results were showing that the prediction error signal can reflect the variations at slope discontinuities of QRS complex, including onset, offset, sharp R wave and AIQP. The results have also showed that the significant reduction of UIQP in patients with ventricular tachycardia may be new promising evidence of ventricular arrhythmias.