PR-CAIDAS / AIMI-lab@ MICCAI 2025 in Daejeon, South Korea

At this year’s International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2025), our PR@CAIDAS/AIMI@FAU researchers shared exciting new advances in medical imaging and AI: Maja Schlereth presented her latest work on self-supervised super-resolution, enabling the generation of high-resolution MR images with detailed anatomical structures using only two orthogonal anisotropic low-resolution scans. Jingna Qiu showcased her research on annotation-free vision–language model adaptation at the COMPAYL workshop. Sweta Banerjee also presented at the COMPAYL workshop her recent work, which uses a triple-condition setup (context patch, inpainting mask, and chromosome mask) to generate realistic mitosis images by mixing morphology and tissue backgrounds and improve atypical-vs-normal classification with this synthetic data. Xingjian presented at Agentic AI in Medicine workshop a LLM-based agent framework to assist with the interpretation and execution of CT scan protocol configuration requests given in natural language or a structured and device-independent format, aiming to improve the workflow efficiency and reduce CT technologists’ workload. Karim Elbarbary presented at the DeepBreath workshop a clinically inspired multi-view vision-language model for enhanced mammography mass detection. It was truly inspiring to connect with researchers from around the world and exchange ideas that drive innovation in medical imaging. To explore more of our research, check out our papers or reach out to us directly. A big thank you to the MICCAI 2025 organizers for an outstanding event!