AI Brain Tumor Detection: 97% Accuracy
- Robovision Healthcare and the Netherlands Cancer Institute (NKI) have introduced BrainMets AI,an artificial intelligence solution designed to detect brain metastatic disease with exceptional accuracy.
- Brain metastases, the most common tumors in the central nervous system, affect up to 17% of adult cancer patients.
- The AI solution achieved 97.4% lesion-level sensitivity, including 93.3% sensitivity on lesions smaller than 3mm.In a study of 260 patients, the model identified every scan with brain metastases,...
BrainMets AI, an artificial intelligence solution, detects brain metastases with remarkable precision, achieving 97.4% lesion-level sensitivity. This groundbreaking technology, developed by Robovision Healthcare and the Netherlands Cancer Institute, promises earlier and more accurate detection of brain tumors, perhaps revolutionizing cancer patient outcomes.The AI model excels at identifying small lesions, some under 3mm, using a diverse dataset of over 1,500 MRI cases. the meticulously curated data and advanced algorithm set a new standard in neuroradiology. With an expected commercial launch by 2026, its integration into existing clinical workflows is poised to enhance diagnostic confidence. News Directory 3 can help you stay informed and get the latest developments in this field. Discover what’s next in AI-driven cancer detection.
AI Spots Brain Metastases with High Precision
Updated June 27,2025
Robovision Healthcare and the Netherlands Cancer Institute (NKI) have introduced BrainMets AI,an artificial intelligence solution designed to detect brain metastatic disease with exceptional accuracy. The research, featured in Radiology, suggests earlier detection could significantly improve outcomes for cancer patients.
Brain metastases, the most common tumors in the central nervous system, affect up to 17% of adult cancer patients. Detecting these small lesions,sometimes smaller than 3mm,is challenging due to the volume of MRI data. BrainMets AI aims to assist radiologists in identifying these metastases, which can grow rapidly and complicate treatment.
The AI solution achieved 97.4% lesion-level sensitivity, including 93.3% sensitivity on lesions smaller than 3mm.In a study of 260 patients, the model identified every scan with brain metastases, with fewer than 2% false positives. This level of precision marks a significant advancement in neuroradiology, possibly improving the speed and accuracy of diagnosis.
Stephane Willaert, head of Robovision healthcare, said the quality of the AI stems from the quality of data and the skill in developing the algorithm. He added that the meticulously curated data, accurate annotations, and optimized neural network architecture created an algorithm that sets a new standard in neuroradiology.
The advancement involved a diverse dataset of over 1,500 MRI cases from multiple institutions across Europe and the U.S., high-fidelity voxel-level annotations by neuroradiology specialists, and a custom model architecture developed by Robovision’s AI team.
What’s next
BrainMets AI is undergoing clinical validation trials in Europe and the U.S. The AI-driven tool, poised for commercial availability by 2026 pending regulatory clearances, integrates into existing clinical workflows, offering quantification, segmentation, and longitudinal tracking to ease the documentation burden for radiologists. This advancement in artificial intelligence promises to enhance diagnostic confidence and streamline workflows in the fight against cancer.