Curve registration for study of P-wave morphing during exercise

The P-wave shape analysis remains an interesting objective in various situations. In this paper, the aim is to investigate how the P-wave morphology changes during time-varying effort. To perform this task, a method, Self-Modeling Registration (SMR), derived from the Curve Registration (CR) theory is used. This method permits to link different shapes by mean of time warping functions. First, using the SMR method, significant P-wave shapes are extracted from two ECG records under maximal and graded exercise test where the effort is released abruptly. Then, the selected shapes are linked by warping functions estimated by SMR method. The obtained warping functions show a reversible evolution of the P-wave shape independently of the possible superimposition of the T-wave. To give a physiological explanation of this P-wave morphing, an atrial contribution evolution scenario under exercise is proposed and validated by simulation.