Maximilian Fehrentz

I'm a first-year PhD candidate with Nassir Navab in the Computer Aided Medical Procedures (CAMP) group at TU Munich. Previously, I was a research intern at Harvard Medical School & Brigham and Women's Hospital with Nazim Haouchine and a student researcher at LMU University Hospital Munich with Christian Heiliger. In Boston, I worked on 2D/3D registration in neurosurgery using multi-modal radiance fields. In Munich, I focused on 4D reconstruction for visceral surgery with CT-coupled 4D Gaussian Splatting. I am also part of TUM.ai, OneAIM, and CDTM.

Email  /  Scholar  /  LinkedIn  /  Github

Research

I build intelligent systems for surgery, mostly focused on brain and visceral surgery. In particular, I work on intraoperative 4D reconstruction, 4D language-aligned representations, multi-modal representations that leverage 3D preoperative imaging (CT, MRI, …), and multi-modal reasoning based on knowledge about the procedure as embedded in (M)LLMs. To complement fine-grained 4D spatiotemporal understanding, I am also interested in long-form video understanding to reason over several hours of surgery. My work primarily involves 4D radiance fields, MLLMs, and all sorts of (vision) foundation models. As we work on the intersection of medical interventions and AI, I work closely with surgeons and spend time in the operating room.

Selected Publications

Coupling 3D preoperative imaging (CT) and intraoperative 4D Gaussian Splatting.

Intraoperative registration of the preoperative MRI to the brain using a cross-modal NeRF.

This website is adapted from Jon Barron's template.