Congratulations to Atallah Madi, recipient of the Best Oral Presentation Award at the 9th International Conference on Advances in Artificial Intelligence (ICAAI 2025) in Manchester, UK.
His research introduces a method to assist in the diagnostic process of Hirschsprung’s disease. Hirschsprung disease (HD) is a congenital birth disorder that causes colonic nerve cell deformation and intestinal blockage. It is characterized by the lack of colon ganglion cells in the myenteric plexus regions. The disease affects 1 in 5000 newborns globally and may be deadly if untreated. In this paper, a Vision–Language Model (VLM) is trained to classify plexus regions in Hirschsprung’s disease using both digital microscope histopathology images and medical knowledge. The text descriptions of the disease are generated using a Large Language Model (LLM) that is used as a reasoning model and is limited to a set of trusted medical literature. This allows the VLM to make predictions while being aligned to how pathologists examine and interpret tissue samples.