U-FaceBP: Uncertainty-aware Bayesian Ensemble Deep Learning for Face Video-based Blood Pressure Measurement

TL;DR

This work tackles non-contact blood pressure estimation from face videos by embracing uncertainty. It introduces U-FaceBP, a Bayesian ensemble framework that fuses three modalities—rPPG signals, PPG reconstructed from RGB videos, and face images—through an uncertainty-driven aggregator, while explicitly modeling aleatoric and epistemic uncertainties via MC dropout. The approach yields state-of-the-art accuracy across large, racially diverse datasets and provides informative uncertainty estimates for reliability, modality fusion, and subgroup analysis. The results demonstrate potential for safe, scalable, home-based BP monitoring with built-in triage to direct uncertain cases to human experts or conventional devices.

Abstract

Blood pressure (BP) measurement is crucial for daily health assessment. Remote photoplethysmography (rPPG), which extracts pulse waves from face videos captured by a camera, has the potential to enable convenient BP measurement without specialized medical devices. However, there are various uncertainties in BP estimation using rPPG, leading to limited estimation performance and reliability. In this paper, we propose U-FaceBP, an uncertainty-aware Bayesian ensemble deep learning method for face video-based BP measurement. U-FaceBP models aleatoric and epistemic uncertainties in face video-based BP estimation with a Bayesian neural network (BNN). Additionally, we design U-FaceBP as an ensemble method, estimating BP from rPPG signals, PPG signals derived from face videos, and face images using multiple BNNs. Large-scale experiments on two datasets involving 1197 subjects from diverse racial groups demonstrate that U-FaceBP outperforms state-of-the-art BP estimation methods. Furthermore, we show that the uncertainty estimates provided by U-FaceBP are informative and useful for guiding modality fusion, assessing prediction reliability, and analyzing performance across racial groups.

Figures (12)

• Figure 1: (a) PTT- and PWA-based methods for cuffless BP measurement using ECG or PPG. (b) U-FaceBP for camera-based BP measurement. We design U-FaceBP as an ensemble method consisting of rPPG, PPG signals (reconstructed from face videos), and face image-based BP estimation models. U-FaceBP outputs predicted BP and its uncertainty.
• Figure 2: Framework of U-FaceBP: (a) BP estimation from rPPG signals, (b) BP estimation from PPG signals, (c) BP estimation from face images, and (d) estimation of uncertainties and BP. Face photo by Generated Photos (https://generated.photos/).
• Figure 3: SBP and DBP distributions in FaceBP dataset.
• Figure 4: Correlation plots of SBP and DBP in U-FaceBP.
• Figure 5: Correlation plots for SBP and DBP predictions by U-FaceBP with aleatoric, epistemic, and total uncertainties on FaceBP dataset. First and second rows correspond to SBP and DBP predictions, respectively. First, second, and third columns represent aleatoric, epistemic, and total uncertainties, respectively.