Key result
Parameter-efficient deep learning for PPG signals accurately estimates vital signs with ~2 bpm HR error.
Compact, fully convolutional deep learning models can accurately estimate vital signs from PPG signals across different datasets and sensors, enabling on-device deployment.
Enables accurate multi-parameter vital sign estimation from PPG; leaves open prospective clinical validation before routine use.
Photoplethysmography (PPG) enables non‐invasive estimation of vital signs across a range of sensors, from wrist wearables and clinical monitors to smartphone cameras. We present parameter‐efficient, end‐to‐end deep models for heart rate (HR), blood oxygen saturation () and respiratory rate (RR) estimation. These models operate directly on PPG signals and integrate preprocessing as a layer. Our family of fully convolutional architectures includes a baseline FCN, a residual FCN, a ConvNeXt‐inspired FCN and a compact DCT‐based variant. These models achieve strong cross‐dataset performance while using only 3.3K–53K parameters. On PPG‐DaLiA, the best performing model achieves 6.07 2.70 bpm MAE for HR. In HR, and RR tasks on the BIDMC dataset, the best‐performing method achieves error rates of 2.2 bpm, 3.14% and 1.49 breaths per minute, respectively. We also release MTHS, a smartphone‐based vital‐sign dataset comprising 62 fingertip‐PPG recordings (35 men and 27 women) with 1‐Hz HR and annotations. Additionally, we demonstrate practical feasibility via on‐device deployment using TFLite on smartphones. These results show that compact, fully convolutional and DCT‐based designs provide accurate, generalizable, and deployment‐ready PPG vital‐sign estimators across sensors and datasets.
No takes yet. Share an insight, caveat, or question.
Samavati et al. (2026) studied Vital sign estimation (n=62). Parameter-efficient deep learning models was evaluated on Heart rate, blood oxygen saturation, and respiratory rate estimation error. Parameter-efficient deep learning models for PPG signals achieved accurate vital sign estimation, with error rates of 2.2 bpm for HR, 3.14% for blood oxygen saturation, and 1.49 breaths/min for RR.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: