AI selfie tool assesses cancer survival
New AI tool can predict cancer survival by a single patient selfie, study suggests
Tuesday, 20 October 2026
Photo by Tasha Kostyuk on Unsplash
Warm-up
- What health clues can a face show, if any?
- Would you trust AI to help your doctor? Why or why not?
- Have you tried an app that guesses your age?
Vocabulary
- estimate
- to guess a number or amount using the best information you have
- predict
- to say what you think will happen in the future
- survival
- continuing to live after illness or danger
- selfie
- a photo of yourself that you take with your phone
- radiotherapy
- cancer treatment that uses radiation to kill cancer cells
- algorithm
- a set of rules a computer follows to solve a problem
- risk
- the chance that something bad may happen
- complication
- a new problem that makes an illness or treatment harder
Reading
Listen to this story
British EnglishDoctors at Mass General Brigham in Massachusetts have built an artificial intelligence tool that can estimate a cancer patient’s chances of survival from a single selfie. Researchers said it can predict biological age and survival, and could guide care.
The team developed two tools from one model trained on more than 40 million facial photos. FAHR-FaceAge estimates how old a patient looks and links that to survival. FAHR-FaceSurvival predicts survival directly from the photo. The study reported that people treated with radiotherapy in the US who looked five or more years older had a 21% higher risk of dying over about four years; those who looked five or more years younger had a 17% lower risk.
Scientists added that one algorithm focused on lines near the mouth and forehead, while the other used the area below the eyes. Doctors said the tool might help plan treatment and predict complications after major surgery.
Comprehension
- Who developed the AI tool, and where is this organisation based?
- What are the names of the two tools in the system?
- How many facial photographs were used to train the model?
- What did the study find about patients who looked older or younger?
- How might the tool help doctors, according to the study?
Grammar focus
Reported speech
Use a reporting verb (say/said, report) + that + clause to tell what someone said, without quotes. With a past reporting verb, the tense often moves back. For example: "Researchers say it can predict age" becomes "Researchers said it could predict age". (Example 1 constructed.)
- The team said FAHR-FaceSurvival predicted survival from one photo.
- Doctors said the tool might help plan treatment.
Grammar exercise
Complete each sentence in reported speech: put the words in brackets one tense back.
- Researchers said it (can guide) ___ care.
- The team added that one algorithm (focuses) ___ on mouth lines.
- The team said that the other (uses) ___ the area below the eyes.
- The study said older-looking patients (have) ___ a higher risk.
- They said the model (was trained) ___ on over 40 million photos.
Discussion
- Would you share a selfie for a health test? Why or why not?
- What benefits and risks do you see with AI in hospitals?
- Should doctors rely on AI or use it only as support?
- How would you feel about age estimates based on your face?
Vocabulary, reading, and exercises with instant feedback
Spotted a mistake in this lesson? Tell us.