UMY develops 5 AI prototypes for detecting lymphoma to brain tumours

Universitas Muhammadiyah Yogyakarta (UMY) has developed five artificial intelligence (AI) prototypes designed to assist in detecting various diseases, from malignant lymphoma to brain tumours. UMY Professor of Applied Artificial Intelligence Prof Slamet Riyadi said the five AI-based prototypes include LymphoScan, myStressD, BrainDetect, myHeartD v3, and myDBT-Enhance. “These applications are designed as support tools and not as replacements for professional evaluation by healthcare practitioners,” Slamet said while presenting at ‘The 10th International Conference on Sustainable Innovation (ICoSI) UMY’ in Yogyakarta on Wednesday. He said LymphoScan is a web-based application that analyses microscopic images to detect types of malignant lymphoma, including Chronic Lymphocytic Leukemia (CLL), Follicular Lymphoma (FL), and Mantle Cell Lymphoma (MCL). Two other prototypes focus on non-cancer detection, namely myStressD for detecting a person’s stress condition through facial images on mobile devices, and myHeartD v3 which estimates the risk of coronary heart disease from patient questionnaire data. Slamet emphasised that all prototypes have been copyrighted and tested in the laboratory with high accuracy rates. “A more accurate model does not always mean better service. Its application also depends on speed, memory, and context of use,” he said. According to him, evaluating AI models in healthcare cannot be measured by a single accuracy metric alone, but must also consider operational and clinical impacts before being applied in daily healthcare services. The UMY research underlying the development of these five prototypes covers four main categories: chest and respiratory imaging, cardiovascular and metabolic risk, cancer imaging, and mental health and biosignals.