AI tool uses face photos to estimate biological age and predict cancer outcomes

Researchers at Mass General Brigham have developed an AI tool called "FaceAge" that analyzes facial photos to estimate a person's biological age and predict cancer outcomes. The tool utilizes a deep learning algorithm trained on a dataset of over 58,000 photos of healthy individuals.

The study found that cancer patients, on average, exhibited a higher FaceAge than those without cancer, appearing approximately five years older than their chronological age. This innovative approach leverages readily available facial data to provide valuable insights into a patient's health status.

The study's findings, published in The Lancet Digital Health, revealed a strong correlation between FaceAge and survival outcomes in cancer patients. Older FaceAge predictions were associated with worse overall survival across multiple cancer types. Furthermore, FaceAge outperformed clinicians in predicting the short-term life expectancies of patients undergoing palliative radiotherapy.

This highlights the potential of AI-driven facial analysis to provide more objective and accurate assessments compared to subjective evaluations. The researchers emphasize the clinical significance of FaceAge, stating that a simple photograph contains valuable information that can inform clinical decision-making and care plans.

The tool's ability to estimate biological age from facial features offers a novel approach to assessing a patient's overall health and vitality. This method can help to address potential biases in age assessment and provide a more objective measure for healthcare professionals. By leveraging AI and facial recognition technology, FaceAge offers a promising tool for improving patient care.

The study's results suggest that FaceAge can be a valuable addition to the clinical toolkit, providing clinicians with a more objective and predictive measure of a patient's health and prognosis. This technology has the potential to enhance treatment planning and improve patient outcomes across various cancer types.