Investigating the impact of contact pressure on photoplethysmograms

Abstract

Photoplethysmography (PPG) sensors are widely used to measure a variety of physiological parameters, from heart rate to clinically important peripheral oxygen saturation (SpO₂). The ease of obtaining a PPG signal simply by placing the sensor on a body site with even remotely sufficient vasculature (typically the wrist, finger, earlobe, or temple) can easily lead to overlooking the aspect of appropriate sensor contact pressure (P_c). We sought to investigate the effects of P_c from the perspective of: (i) SpO₂, (ii) pulse arrival time (PAT), and (iii) PPG features. Consequently, we developed a finger cuff device to measure multispectral (green, red, and infrared (IR)) PPG signals at different P_c levels. The SpO₂ values were found to increase, driven by the IR component, above the theoretical maximum of 100% slightly after the level of the mean arterial pressure. The maximum variation due to P_c was approximately 1.9 percentage points. PAT values calculated using the waveform feet of the red and IR channels were the most robust. PPG features were also sensitive to varying sensor P_c levels, with each feature showing a rather unique response. However, in general, a P_c between 20–30 mmHg (a slight P_c) and the level of diastolic blood pressure is likely to produce the best results on the features. Overall, the results showed that sensor P_c is a source of inaccuracy in PPG analysis and therefore should be given more consideration in device designs.