ASSESSING THE QUALITY OF PHOTOPLETHYSMOGRAPHY SIGNALS FOR ARTERIAL PULSE WAVE SENSING

Abstract

Objective: Cardiovascular health has an impact on the arterial blood pulse wave features, such as shape and speed. Photoplethysmography (PPG) allows measuring the light absorption variation in the skin due to the blood pulse wave. This study aims at collecting good quality PPG signals for pulse wave sensing. First, the signal quality must be accurately defined and quantified. Then, it must be assessed for several sensor locations during both rest and exercise. Design and method: Based on ECG peaks, the PPG signal is split into beats, which are then fitted to a 4-long piece-wise linear function. Beats are valid only if they follow a specific slope pattern (the second slope has the opposite sign to the other three), and if they have a dichrotic notch. The signal quality is then quantified by the percentage of valid beats. The initial cohort of the clinical trial (NCT05393401, approved by Human Research Review Board (CPP SM I) and Health French Ministry) includes five healthy subjects (1F/4M, age [37-60] years). Their brachial and radial arteries are located by ultrasound. Each subject is monitored with a multi-sensing homemade platform with PPG sensors and ECG; the distance between the LED and the photodiode is 9mm. Measurements are realized with sensors directly above the artery on the elbow, the wrist and the carotid, first at rest and then during a physical exercise. Results: Based on these preliminary results, the quality of PPG signals is best at the wrist whether at rest (48%) or exercise (44%). The second best option is the forearm and the worst is the carotid. The quality during exercise is similar, with a decrease of 4% at the wrist. Conclusions: The signal quality of PPG sensors is strongly impacted by the position along the arteries. The pressure applied on sensors and the extended arm posture also improve the signal quality. Further work should be undertaken to study the feasibility of pulse wave analysis for ambulatory blood pressure monitoring.