A face photo could help doctors personalise cancer care

For people facing cancer treatment, a simple photograph could eventually help doctors judge how much strain their body can cope with.

Researchers at Mass General Brigham in Massachusetts have developed two artificial intelligence tools that look for health clues in a face. One estimates biological age, rather than years since birth; the other predicts survival risk directly.

They share a model trained on more than 40 million facial images. Combining their scores improved survival prediction in a separate group of cancer patients, according to the research team.

The researchers’ earlier full paper explains that the study analysed existing patient records. It did not test whether using the tools to choose treatment improved patients’ outcomes. Further clinical research is needed.

Radiation oncologist Raymond Mak says the aim is to “support personalized treatment decisions rather than replace physician judgment”.

The hopeful possibility is a quick, accessible extra clue to help doctors tailor care to the person in front of them.

Sources

• The Guardian — report on facial AI and cancer survival
• Mass General Brigham — AI facial health recognition tools
• Haugg and colleagues — earlier full research paper (June 2025 preprint)