Visual images are holistic, and when verbalized there is a partial loss of semantic content. Inclusion of images in the knowledge base of intelligent systems can significantly improve their effectiveness. At the stage of formation of intermediate diagnostic hypotheses, the system will present to the user (physician) a hypothesis specific to the verbal and visual characteristics. The subsequent process will depend on the physician’s confidence in the coincidence of the image of the diagnosed patient with the image(s) in the knowledge base of the intellectual system.
Kobrinskii B. A. Images in Logical-and-Linguistic Artificial Intelligence Systems // Journal of Biomedical Engineering and Medical Imaging. 2019. Vol 6, No 1. P.1-8.