Computer Program Learns to Find Meningioma on MRI Scans
Original title: Deep Learning for Automated Detection and Segmentation of Meningioma on Multiparametric MRI.
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 was tested on past scans from real patients, but doctors do not use it in care yet.
The short version
Researchers trained a computer program to find and outline meningioma tumors on brain scans.
What was studied. Researchers looked back at MRI scans from 149 people with meningioma, the most common noncancerous tumor that starts inside the head. They fed a computer program up to 63 mixes of common scan types. Then they checked its work against outlines drawn by an experienced scan doctor.
What they found. When the program used just one scan type, the scan taken with contrast dye: A liquid put into a vein before some scans. It makes certain tissues stand out, which helps show where a tumor is. It does not always show the whole of it. See the glossary worked best. Using up to 3 scan types together always did better than using just one. The best mix was the dye scan plus two other common scan types, and it found and outlined the tumors best.
What this means, and what it doesn't
What it could mean: Someday, tools like this might help doctors measure a meningioma more quickly. They could also help plan radiation treatment. The authors say it may cut the time doctors spend outlining tumors by hand. It may also make results more alike from one doctor to the next.
What it doesn't mean: This is not a new treatment, and it is not a cure. The program was only tested on scans from the past. It is not used in everyday care yet. More work is needed before doctors could use it with patients.
Source: PubMed, September 8, 2026 · Read the original
This plain-language summary was written by AI and published automatically after passing our automatic safety checks. How we write.