Normalization of the Photoplethysmographic Signal using the Post-extrasystolic Potentiation

Roel J. H. Montree1, Elisabetta Peri2, Lukas Dekker3, Xi Long2, Reinder Haakma4, Rik Vullings2
1University of Technology Eindhoven, 2Eindhoven university of technology, 3Catharina Hospital, 4Philips


Abstract

The widespread use of wearable devices such as smartwatches has made continuous heart activity monitoring using photoplethysmography (PPG) more accessible and commonplace. However, PPG suffers from a lack of an absolute, interpretable unit for the signal amplitude, limiting its full diagnostic potential.

We present a method to normalize PPG signals by leveraging the post-extrasystolic potentiation (PESP), a physiological phenomenon in which the contractility of the heart increases following a premature ventricular contraction (PVC). PVCs occur naturally in the human heart and have a well-studied resulting potentiation which behaves independent of many otherwise confounding factors. This allows the normalization to be relative to a consistent behaviour, giving meaning to the recorded signal amplitude.

The proposed method was developed on a dataset acquired during percutaneous, electrophysiological procedures. Seven patients have been included. In these patients, a protocol was performed where PVCs are provoked by disturbing the normal paced rhythm, resulting in a total of 2328 detected pulses, in which there are 286 combinations of PVC and the succeeding pulse potentiated by the PESP, whether provoked or occurring naturally during the measurement.

The PPG signals were analyzed alongside recorded electrocardiography (ECG) and arterial blood pressure (aBP) as a reference to detect PVCs and their subsequent potentiated beats. These beats are then analyzed and, based on parameters of the enhanced beats such as amplitude and maximum slope, provide a normalization factor for the surrounding signal. Preliminary findings indicate that PESP-based normalization provides insights into PPG signal interpretation, correcting for amplitude variations caused by the measurement modality, ensuring remaining variations are reflections of physiological changes.

These findings suggest that incorporating PESP analysis could enhance PPG-based cardiovascular assessments in wearable health monitoring. While promising, the method remains local to PVC-induced disturbances in normal heart rhythm, necessitating further research for broader applicability.