Network and learn at our imaging events

The recent emergence of diffusion models has driven significant advancements in solving inverse problems by leveraging these models as powerful generative priors. However, challenges persist due to the ill-posed nature of such problems, including extending solutions to 3D and temporal domains and addressing inherent ambiguities in measurements. In this talk, we present strategies developed by our lab at KAIST to tackle these challenges. First, we explore the manifold geometry of diffusion models, which has become a foundational concept for designing constrained diffusion models. Building on this, we introduce the Diffusion Posterior Sampling (DPS) algorithm, which enables manifold-constrained measurement guidance during the reverse sampling process. Additionally, we present its accelerated implementation, the Decomposed Diffusion Sampling (DDS) method, tailored for high-dimensional imaging problems in biomedical applications. Finally, we discuss several extensions, including text-driven reconstruction, CFG++, and applications to video and 3D domains.

‍

Bio:‍

Jong Chul Ye is a Professor and the Chung Moon Soul Mirae Chair at the Kim Jaechul Graduate School of Artificial Intelligence (AI) at the Korea Advanced Institute of Science and Technology (KAIST), Korea. He received his B.Sc. and M.Sc. degrees from Seoul National University, Korea, and his Ph.D. from Purdue University, USA. Prior to joining KAIST, he worked at Philips Research and GE Global Research in New York. Professor Ye has held several editorial roles, including Associate Editor for IEEE Transactions on Image Processing, IEEE Computational Imaging, and IEEE Transactions on Medical Imaging, as well as Senior Editor for IEEE Signal Processing and editorial board member for Magnetic Resonance in Medicine. He is an IEEE Fellow and has served as Chair of the IEEE SPS Computational Imaging Technical Committee and as an IEEE EMBS Distinguished Lecturer. He is also a Fellow of the Korean Academy of Science and Technology and currently serves as the President of the Korean Society for Artificial Intelligence in Medicine (2023–2024). Among his numerous accolades are two prestigious awards for mathematicians in Korea: the Choi Suk-Jung Award and the Kum-Kok Award, as well as the Career Achievement Award from the Korean Society for Magnetic Resonance in Medicine. Professor Ye’s research interests focus on machine learning for biomedical imaging and computer vision.

This lecture will explore the innovative application of label-free flow cyto-tomography in the study of suspended cells. Traditional methods of cellular analysis often rely on labeling techniques that can alter or obscure native structures, limiting the accuracy of observations. Flow cyto-tomography, however, provides a powerful, non-invasive alternative for visualizing and quantifying the internal architecture of cells in suspension. By combining the principles of flow cytometry with high-resolution tomographic imaging, this technique offers unprecedented insights into cellular morphology, organelle organization, and quantification of subcellular structures. The lecture will cover the underlying technology, its applications in biomedical research, and its potential to advance our understanding of cellular function in health and disease. Looking ahead, this approach could pave the way for novel diagnostic tools and therapeutic strategies, opening new frontiers in personalized medicine and cellular engineering.