Researchers publish study on AI models that predict cancer patient survival from facial photographs
Researchers at Mass General Brigham published a study in The Lancet Digital Health describing AI models trained on over 40 million facial images that estimate biological age and mortality risk in cancer patients. In the study, 52.5% of 122 patients flagged as high-risk died within 90 days, versus about 2% of 975 in the lowest-risk group. The models assess facial regions including the temples, forehead, nasolabial folds, nose bridge and under-eye area.
6
Divergence score
3 outlets covered it, splitting into 3 framing camps across 2 bias groups.
3 camps
2 bias groups
The spectrum · how 3 outlets placed this story
LeftCenterRight
CNN
Washington Post
Newsmax
Supportive of action
Neutral
Dismissive
Critical
Alarmist
International angle
Widest split · the two outlets furthest apart, and who published first
Furthest apart on this story
CNN
How a photo of a person’s face could help predict whether they’ll survive cancer treatment
detailed feature with researcher quotes and Hippocrates angle · positive
Newsmax
AI Tool Predicts Cancer Survival From Face Photo
compact AI-tool framing · neutral
First to report
- CNNfirst
- Newsmax+20.9 h
- Washington Post+23.0 h
By each outlet's earliest publish time in this story. Feeds that index slowly can appear later than they printed.
The split, in one line
All three outlets describe the same study and the same survival statistics; the split is only in emphasis, with the Hippocrates parallel and researcher quotes carried in the detailed account while the others compress to the headline finding.
How each outlet covered it
Lightly covered so far
Too few outlets to map a left-right split. Here is each take as it stands.
Sparse coverage · 3 outlets
“How a photo of a person’s face could help predict whether they’ll survive cancer treatment”
“How a photo of a patient’s face could predict whether they’ll survive cancer”
“AI Tool Predicts Cancer Survival From Face Photo”
4 tracked claims across 3 outlets
Fact ledger
All4Claimed3Corroborated1
1/3
Claimed
The models were trained on more than 40 million facial images
Corroborated
Disputed
3 more tracked claimsSign up free to see the full ledger