# Your face could predict cancer survival, study finds

> Researchers developed an AI tool that estimates cancer patients' survival prospects from a single facial photograph.
- Title: Your face could predict cancer survival, study finds
- Summary: Researchers developed an AI tool that estimates cancer patients' survival prospects from a single facial photograph. An AI model trained on facial images can…
- Keywords: artificial intelligence, cancer research, facial recognition, medical imaging, mass general brigham, survival prediction, healthcare technology, face, predict, cancer, survival, study
- Source: Fastcompany — https://fastcompany.com/91626954/a-photo-of-your-face-could-reveal-how-likely-you-are-to-survive-cancer-new-study-finds
- Read time: 2 min
- Topics: artificial intelligence, cancer research, facial recognition, medical imaging, mass general brigham, survival prediction, healthcare technology
## AI predicts cancer survival from photos
Researchers have developed an AI tool that can estimate cancer patients' survival prospects from a single facial photograph, potentially changing how doctors make treatment decisions.
## The model behind the breakthrough
The tool, known as FAHR-FaceSurvival, was developed by the Mass General Brigham research team and trained on forty million general facial images and thirty thousand images of cancer patients.
## Study published in The Lancet
The research was published in The Lancet Digital Health. Cancer patients' facial images were linked to medical outcome details, allowing researchers to test the model's accuracy.
## What the researchers claim
Hugo Aerts, director of the Artificial Intelligence in Medicine program at Mass General Brigham, said this is the first demonstration that an AI model can directly predict cancer survival from a simple photograph.

> This is the first demonstration that an AI model can directly predict cancer survival from a simple photograph.
## The survival prediction results
A high score from FAHR-FaceSurvival was significantly associated with early mortality risk. Half of the one hundred twenty two patients in the highest ten percent of scores died within three months.
## Comparison with lowest risk group
Just two percent of patients in the group with the lowest scores died within three months, compared with fifty percent of those in the highest scoring group.
## Which facial features matter
The cancer survival model focused specifically on the part of the face below the eyes and around the bridge of the nose, although researchers do not fully understand what the AI system saw in those…
## Facial biomarkers explained
The research explores facial biomarkers, showing how a simple photo of the face can give doctors a surprisingly in-depth picture of a person's health beyond cancer survival prediction.
## Broader applications ahead
Cancer survival rates will not be the last thing scientists seek to predict using facial data. The field of facial health recognition AI is advancing to predict other health outcomes from photographs.
## Key takeaway

An AI model trained on facial images can predict cancer survival risk from a photograph, potentially changing how doctors make treatment decisions.