This seminar presents MultiGeoMed, a multidisciplinary research project focused on advancing multimodal geometric deep learning for medical imaging – funded under the NCN Sonata programme. The project explores how diverse geometric representations – voxel grids, surface meshes, point clouds, and implicit neural representations – can be combined to build more robust and generalizable AI models for analyzing complex anatomical structures.<\/p>\n\n\n\n
During the talk, we will introduce the scientific foundations and long\u2011term vision of MultiGeoMed, outline our planned research directions, and discuss strategies for fusing heterogeneous geometric modalities in clinical\u2011scale deep learning pipelines. We will also share preliminary experimental results, demonstrating early insights into the strengths and limitations of individual geometric formats and their potential synergies. By integrating multiple geometric perspectives, MultiGeoMed aims to improve implant modeling, support personalized treatment planning, and contribute new theoretical understanding to the rapidly evolving field of geometric deep learning in medicine.<\/p>\n\n\n\n
Marek Wodzi\u0144ski<\/strong>, Ph.D., is a biomedical scientist specializing in computer science and medical informatics. His interdisciplinary expertise bridges cutting-edge technologies with real-world healthcare applications, particularly in the fields of medical image analysis, computer vision, and deep learning. He currently serves as a Senior Researcher at Sano Centre for Computational Medicine and as an Assistant Professor at AGH University of Krak\u00f3w, where he mentors students and leads innovative research initiatives. Dr. Wodzi\u0144ski has been the principal investigator on multiple nationally funded projects supported by the National Science Centre (MultiGeoMed<\/a>) and the National Centre for Research and Development (DeepImplant). He has also contributed as an AI expert in several international collaborations when working at HES-SO Valais in Switzerland, including projects such as BigPicture or ExaMode.<\/p>\n\n\n\n His recent work explores geometric deep learning, 3-D shape completion, and digital pathology, pushing the boundaries of how artificial intelligence can support diagnostics and precision medicine. He has earned recognition in numerous scientific competitions focused on medical imaging and has received prestigious awards for his contributions to science, including: (i) The Polish Ministry of Science Scholarship for outstanding young researchers, (ii) The ABB Research Award, (iii) FNP START Fellowship, (iv) The POLITYKA Scientific Award. Driven by curiosity and a commitment to impact, Dr. Wodzi\u0144ski continues to advance the intersection of AI and healthcare, shaping the future of medical technologies.<\/p>\n\n\n\n mwodzinski.edu.pl<\/a><\/p>\n\n\n\n Dr. Wodzi\u0144ski is delivering the NCN\u2011funded project \u201cMultimodal Geometric Deep Learning in Medical Image Analysis (MultiGeoMed)<\/a>.\u201d His team includes Wojciech Szyma\u0144ski<\/a> (PhD Student) and Gniewosz Drwi\u0119ga<\/a> (Postdoctoral Researcher).<\/p>\n\n\n\n