AI tool predicts cancer survival chances with single face photo
Trained on over 40 million images of faces, the model assigns risk scores to those with the illness
Researchers have developed an artificial intelligence tool that can predict cancer patients' survival chances using just a single photo of their face. The model, trained on over 40 million face images and 30,000 cancer patient photos, assigns a risk score to each patient. Higher scores indicate a greater likelihood of death within 90 days.
The AI tool can accurately estimate a patient's age with a margin of error of around 5 years and reliably categorize patients into higher or lower-risk groups, which subsequently impact their mortality risk. The model works across different ages, races, genders, and types of cancer.
However, Miroslaw Bober, a professor of video processing and machine learning, cautioned that the tool is not yet ready for clinical use. Bober noted that predicting an outcome is different from explaining it, and the tool has not demonstrated that these predictions add clinically useful information beyond doctors' general assessments of a patient's health and fitness. Bober also pointed out that the tool needs to be tested against doctors' own judgment and across various hospitals and patient groups.
Dr. Raymond Mak, a radiation oncologist at Mass General Brigham Cancer Institute and one of the study's authors, expressed amazement at the tool's performance in predicting survival. The algorithm primarily focused on the bridge of the nose and the area beneath the eyes to make its assessment. Mak suggested that the tool could serve as an additional piece of information for doctors, preventing overtreatment at the end of life.
Written by urgent.news from The Business Times - Companies & Markets's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
