Random Fields for Image Registration: Bridging the Gap between Continuous Transformations and Discrete Optimization

In this talk, I will present an approach for image registration based on discrete Markov Random Field optimization. While discrete optimization often provides strong solutions in purely discrete settings, the task of registration usually involves the estimation of continuous transformation parameters. We tackle this problem by introducing suitable approximation schemes and iterative refinement strategies in a principled manner. This allows us to apply discrete optimization both for linear and non-linear image registration. The performance of our implementations is demonstrated on several clinical applications including atlas generation, multi-modal brain registration, automatic segmentation via atlas-matching, and whole-body MRI stitching, as well as for the non-clinical problem of optical flow estimation.

Speaker Details

Ben Glocker received the diploma degree in medical computer science from the Technische Universität München (TUM), Germany, in November 2006. His diploma thesis on non-rigid image registration was awarded with the Werver von Siemens Excellence Award. Since December 2006, he is a PhD student in the group Computer Aided Medical Procedures (CAMP) at TUM. He received the Francois Erbsmann Prize for the best oral presentation among all first time presenters at the conference Information Processing in Medical Imaging (IPMI), in 2007. His main interests are optimization algorithms for Markov Random Fields, linear and non-linear image registration, and optical flow estimation. His research is conducted in close collaboration with the Ecole Centrale Paris, where he spent several months during the last years.

Date:
Speakers:
Ben Glocker
Affiliation:
Technische Universitat Munich

    • Portrait of Ben Glocker

      Ben Glocker

    • Portrait of Jeff Running

      Jeff Running