Foundation Artificial Intelligence Models for Health Recognition Using Face Photographs (FAHR-Face)

2025-06-17

AI Summary

This study investigates facial images as non-invasive health biomarkers for two clinical tasks: biological age estimation and survival risk prediction in cancer patients. We propose the first unified health foundation model pretrained on 40 million facial images, enabling dual-task adaptation via a single architecture—without requiring additional annotations. Our method employs a two-stage age-balanced fine-tuning strategy coupled with Cox proportional hazards survival analysis to decouple tasks effectively. We further introduce a novel anatomical attribution analysis, revealing task-specific facial region contributions to aging and disease risk, and provide visual interpretability via saliency maps. The model achieves a mean absolute error of 5.1 years in biological age estimation and yields a hazard ratio of 3.22 (P < 0.001) between the highest- and lowest-risk quartiles in survival prediction. Extensive validation across diverse age groups, sexes, racial backgrounds, and cancer types confirms strong generalizability and robustness.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Background: Facial appearance offers a noninvasive window into health. We built FAHR-Face, a foundation model trained on>40 million facial images and fine-tuned it for two distinct tasks: biological age estimation (FAHR-FaceAge) and survival risk prediction (FAHR-FaceSurvival). Methods: FAHR-FaceAge underwent a two-stage, age-balanced fine-tuning on 749,935 public images; FAHR-FaceSurvival was fine-tuned on 34,389 photos of cancer patients. Model robustness (cosmetic surgery, makeup, pose, lighting) and independence (saliency mapping) was tested extensively. Both models were clinically tested in two independent cancer patient datasets with survival analyzed by multivariable Cox models and adjusted for clinical prognostic factors. Findings: For age estimation, FAHR-FaceAge had the lowest mean absolute error of 5.1 years on public datasets, outperforming benchmark models and maintaining accuracy across the full human lifespan. In cancer patients, FAHR-FaceAge outperformed a prior facial age estimation model in survival prognostication. FAHR-FaceSurvival demonstrated robust prediction of mortality, and the highest-risk quartile had more than triple the mortality of the lowest (adjusted hazard ratio 3.22; P<0.001). These findings were validated in the independent cohort and both models showed generalizability across age, sex, race and cancer subgroups. The two algorithms provided distinct, complementary prognostic information; saliency mapping revealed each model relied on distinct facial regions. The combination of FAHR-FaceAge and FAHR-FaceSurvival improved prognostic accuracy. Interpretation: A single foundation model can generate inexpensive, scalable facial biomarkers that capture both biological ageing and disease-related mortality risk. The foundation model enabled effective training using relatively small clinical datasets.

Problem

• Estimating biological age from facial images accurately
• Predicting survival risk in cancer patients using photos
• Developing robust AI models for health recognition via faces

Innovation

• Foundation model trained on 40M facial images
• Two-stage fine-tuning for age and survival prediction
• Robust testing for cosmetic and clinical variations