by
Gus Iversen, Editor in Chief | July 01, 2025
An AI tool developed by the Netherlands Cancer Institute and Robovision Healthcare has demonstrated high accuracy in identifying even the smallest brain metastases on MR exams, according to a peer-reviewed study published in Radiology.
The algorithm, BrainMets.ai, was tested on more than 1,500 MR scans and validated on an external data set of 260 patients. It achieved a lesion-level sensitivity of 97.4%, including 93.3% accuracy for lesions smaller than 3mm—an area where detection has traditionally been challenging for radiologists. The model flagged all patients in the validation set who had brain metastases, with no false negatives and fewer than 2% false positives.
"With meticulously curated data, accurate annotations and an optimized neural network architecture from Robovision, we were able to create an algorithm that sets a new standard in neuroradiology,” said Stephane Willaert, head of Robovision Healthcare.
Ad Statistics
Times Displayed: 656
Times Visited: 5 Fast-moving cardiac structures have a big impact on imaging. Fujifilm’s SCENARIA View premium performance CT brings solutions to address motion in Coronary CTA while delivering unique dose saving and workflow increasing benefits.
Brain metastases are the most common tumors in the central nervous system, affecting as many as 17% of adult cancer patients. Their detection is complicated by the volume of brain MR data and the small size of many lesions. Missed diagnoses can delay treatment and reduce survival odds.
The study’s methodology included voxel-level annotations by neuroradiology experts and a model trained on data from institutions in both Europe and the U.S. BrainMets.ai integrates into clinical workflows as a concurrent reading aid, offering not just detection but segmentation, quantification, and longitudinal tracking.
“This tool is a true leap forward,” said Dr. Laurens Topff, neuroradiologist at the Netherlands Cancer Institute. “It increases our confidence in detecting even the smallest lesions, supporting more timely and accurate treatment decisions.”
Dr. Michael Iv, a neuroradiologist at Stanford Medicine, added, “We were genuinely impressed by the precision and consistency of BrainMets AI. It has significant potential to help radiologists detect, quantify, and monitor brain metastases with greater confidence.”
BrainMets is currently undergoing clinical validation in Europe and is being tested in the U.S. The developers anticipate seeking CE and FDA clearances by 2026.
