Particle.news
Download on the App Store
10 ARTICLES
·
yesterday

AI Tool FaceAge Predicts Cancer Survival Using Facial Photos

Newly published research validates FaceAge as a low-cost biomarker for biological age, outperforming clinicians in short-term survival predictions for cancer patients.

Image
Image
Image
Face ID. Future. Half Face of Young Caucasian Woman for Face Detection. Brown Female Eye Biometrical Iris Scan Reading for Person Identification. Augmented Reality. 3D Technology Concept.

Overview

  • FaceAge, a deep-learning AI, estimates biological age from facial photographs and correlates it with cancer survival outcomes.
  • The tool was trained on nearly 59,000 healthy faces and validated on over 6,000 cancer patients, revealing that cancer patients appeared, on average, five years older biologically than their actual age.
  • Higher biological ages predicted by FaceAge were strongly linked to worse survival outcomes, particularly for patients with a biological age exceeding 85.
  • FaceAge outperformed clinicians in predicting short-term survival for patients undergoing palliative radiotherapy, improving decision-making accuracy when combined with clinical data.
  • Ongoing research aims to address demographic biases, expand applications to other diseases, and explore the impact of cosmetic interventions on the tool's accuracy.