Design of an open research platform for smartphone-based photoplethysmography acquisition with real-time signal quality control and feature extraction

Photoplethysmography (PPG) is increasingly used in cardiovascular and digital health research due to its ease of acquisition from widely available consumer devices, including smartwatches, pulse oximeters, and smartphone cameras. In addition to heart rate and oxygen saturation estimation, commercial applications have demonstrated its potential for atrial fibrillation detection, while further emerging research explores its utility in hypertension, diabetes, sleep assessment, and endothelial reactivity, as well as more novel domains such as pain monitoring, mental stress evaluation, cognitive workload tracking, and biometric authentication.

Despite growing interest, widespread adoption of smartphone-based PPG in clinical and research contexts is limited by the lack of access to raw, standardised pulse waveforms. Available tools often restrict data access, lack transparent documentation of the acquisition process, or impose usage fees. These limitations impede reproducibility and hinder the development and validation of novel signal analysis algorithms.

We developed a free and non-commercial mobile platform designed to enable high-resolution PPG acquisition using the embedded smartphone camera. The application measures pixel intensity changes across individual red, green, and blue (RGB) channels and their combined output. It supports real-time signal display, session annotations, and event marking. Users can flexibly configure signal preprocessing using selectable filters (moving average, Chebyshev II, Butterworth) and assess real-time signal quality using an integrated algorithm.

The application enables export of raw and filtered PPG signals, along with >50 automatically extracted features from morphological and temporal domains. Morphological features include pulse amplitude, systolic upstroke time, and reflection index. Temporal features reflect heart rate variability and include RR intervals, SDNN, and RMSSD. All outputs are available in CSV format, with full access to RGB channel data and preprocessing settings, supporting reproducibility and flexible downstream analysis.

This open-access, non-commercial research tool fills a critical gap in PPG research infrastructure by enabling standardised, transparent, and configurable signal acquisition via consumer smartphones. By integrating live quality assessment, filter selection, and feature extraction, it empowers researchers to accelerate innovation in signal analysis, digital biomarkers, and AI model development for remote cardiovascular monitoring.