According to BGR, researchers from the Netherlands Cancer Institute, in collaboration with Robovision Healthcare, have unveiled a new development — the BrainMets.ai system, which utilizes artificial intelligence (AI) for the detection of brain metastases. Research published in Radiology shows that this technology achieves a high level of accuracy, particularly in cases of microtumors that often go unnoticed during standard MRI scans.
Brain metastases are the most common type of tumors in the central nervous system, occurring in about 17% of adult cancer patients. The challenge is that these tumors typically measure less than 3 mm in diameter, making them difficult to identify even with numerous MRI images.
BrainMets.ai achieved a sensitivity of 97.4% in overall lesion detection. For tumors larger than 12 mm, accuracy reached 100%, for sizes 6–12 mm it was 98%, for 3–6 mm it was 97.9%, and for tumors smaller than 3 mm it was 93%.
“This technology is an important step forward. It increases confidence in detecting even the smallest lesions, assisting doctors in making more timely and accurate treatment decisions,” stated Dr. Lawrence Topff, a neuroradiologist at the Netherlands Cancer Institute and clinical leader of the project.
It is noted that this technology is not intended to replace medical professionals but aims to expedite the diagnostic process and reduce the risk of missed cases. According to the developers, a critical factor in accuracy was the quality of the training data used to train the AI. Thanks to carefully selected data, precise annotations, and optimized neural network architecture, such high-performance results were achieved.
