This AI uses selfies to predict cancer survival chances
What's the story
Researchers at Mass General Brigham in the US have developed a new artificial intelligence (AI) tool called 'FaceAge.'
The innovative system uses a selfie to estimate an individual's "biological age," which could provide valuable insights into their health status and how they might respond to cancer treatment.
This way, it offers an objective alternative to the traditional "eyeball test," where doctors make assessments based solely on physical appearance.
AI application
The tool analyzes facial features
The FaceAge tool employs advanced deep learning algorithms to analyze facial features from a photograph and estimate biological age.
Unlike chronological age, biological age reflects an individual's physiological condition and can provide insights into their overall health status.
The tool was trained on 58,851 images of healthy individuals and tested on over 6,000 cancer patients, demonstrating its potential in clinical settings.
Continuous monitoring
The tool can track biological age over time
Dr. Hugo Aerts, the lead author of the study, stressed that this cost-effective AI tool can be employed repeatedly over time to monitor a person's biological age.
He said, "The impact can be very large because we now have a way to actually very easily monitor a patient's health status continuously."
Application
FaceAge showed higher accuracy when used alongside clinical assessments
In clinical evaluations, FaceAge showed that patients whose estimated biological age appeared older than their actual age tended to have poorer survival rates, regardless of cancer type or gender.
This was especially true for those with a FaceAge over 85 years.
Notably, when used alongside clinician assessments, FaceAge improved the accuracy of predicting six-month survival for patients undergoing palliative radiotherapy from 61% to 80%.
Impact
Implications for personalized cancer treatment
The ability to assess biological age through a simple facial photograph offers a non-invasive method to inform treatment decisions.
For instance, a 75-year-old patient with a biological age of 65 might be considered for more aggressive treatment compared to a 60-year-old with a biological age of 70.
This approach could lead to more personalized and effective cancer care strategies.