One of the current applications of artificial intelligence methods is in computer diagnostics within neurophysiology, neuropathology, neurosurgery, psychiatry, clinical psychology, and other medical disciplines related to the study of the human central nervous system. The brain is the central component of the nervous system, regulating all other organs and systems in both humans and animals. Methods of diagnosis of the brain focus on segmentation and analysis of data regarding the brain’s structure and functional state. Brain image segmentation techniques based on artificial intelligence primarily rely on neural networks. This study reviews brain image segmentation methods and examines their characteristics. A segmentation method suitable for diagnosing brain conditions and detecting pathologies is proposed. A program implementing this method based on the object detection model YOLO has been developed for brain region segmentation. For model training, a dataset with features optimized for a balance between accuracy and processing time was created, and training parameter coefficients were selected. The program processes images, video files stored in the system, and video streams from peripheral devices, utilizing both central and graphics processors. The processing accuracy according to the mAPval 50 indicator was 0.957, according to mAPval 50–95 – 0.754, while the processing time did not exceed 162.7 ms., which demonstrates the applicability of the method to solving the problem of brain detection in real time.