Brain scan patterns alone did not predict survival well in cancer that spread to the brain
Original title: MRI Radiomics for Survival Prediction in Brain Metastases: A Machine Learning Analysis.
How far along is this research?
- Lab cells
- Animals
- Review
- Tested in people
This was tested in people. That is the most reliable kind of research we share.
This looked back at past scans from people, and the tool did not work well enough to use.
The short version
A computer studied brain scans to guess how long people would live, but it did not do very well.
What was studied. Researchers looked back at MRI scans from 198 people whose cancer had spread to the brain. A computer looked for scan patterns tied to survival, then was tested on 69 other people.
What they found. The computer settled on 7 scan features. These features were not very good at telling who would live longer, and they worked a little less well in the new group. Adding a known scoring tool to the scan patterns still gave only modest results.
What this means, and what it doesn't
What it could mean: For now, scan patterns alone are not a good way to predict how long someone will live. The researchers say better predictions will need more than scans. They point to health details, treatment, lab details about the tumor, and how the cancer is doing in the rest of the body.
What it doesn't mean: This does not change anything about care today. It was a look back at old scans, not a treatment study. The scan tool is not ready for doctors to use. It is not a cure or a new treatment.
Source: PubMed, September 9, 2026 · Read the original
This plain-language summary was written by AI and published automatically after passing our automatic safety checks. How we write.