MEPhI has created a program for accelerated and accurate diagnosis of brain metastases. It automatically analyzes MRI (magnetic resonance imaging) scans in five minutes instead of the usual 40, detecting even the smallest tumor foci that often go unnoticed during manual diagnosis.
The system is based on a neural network (an algorithm trained to find patterns in images) that not only detects metastases, but also tracks their changes over time. This is especially important for assessing the effectiveness of treatment. The results are formatted into a report that can be connected to the patient's electronic medical record.
The development improves the accuracy of diagnosis, reduces the risk of medical errors, and saves time in establishing a diagnosis. The program has already been trained on images of 600 patients and is being prepared for implementation in clinics. It also opens up new opportunities for research in oncology, including studying the effects of various doses of radiation.
Read more on the topic:
Russian scientists used AI to improve the accuracy of diagnosis of liver tumors
Sechenov University presented a neural network to improve the diagnosis of lung cancer
Chip for detecting metastases in the blood created in Krasnoyarsk
