We are pleased to announce that research conducted within the scientific activities surrounding ChatMED has received the Best Student Paper Award at the 7th International Conference on Deep Learning Theory and Applications (DeLTA 2026), held in Porto, Portugal, on 16–17 July 2026.
The awarded paper, entitled “Evaluating SAM3 and MedGemma for Brain Tumor MRI: Zero-Shot Segmentation, Linear Probing, and Multimodal Diagnosis,” was authored by Tamara Kostova, Ivan Kitanovski, Monika Simjanoska Misheva, and Kostadin Mishev.
The work investigates the potential of emerging foundation and multimodal AI models for brain tumor MRI analysis, examining three complementary capabilities: zero-shot image segmentation, linear probing, and multimodal diagnostic reasoning. By evaluating advanced models such as SAM3 and MedGemma in the context of neuroimaging, the research contributes to the broader ChatMED objective of exploring how state-of-the-art generative and multimodal AI can support future clinical decision-support systems in neurology.
Receiving the Best Student Paper Award is an important recognition of both the scientific quality of the work and the contribution of young researchers working at the intersection of artificial intelligence, medical imaging, and clinical decision support.
The achievement also reinforces ChatMED’s commitment to connecting cutting-edge AI research with clinically relevant challenges and to creating opportunities for young researchers to contribute to the next generation of trustworthy AI-enabled healthcare technologies.
Congratulations to the authors on this achievement!
