Segmentation of cardiac images in 3D MRI
Spring 2015
Master Semester Project
Project: 00301
In this project the coronary arteries want to be localized in order to facilitate the medical examination of a heart MRI image in 3D. Ideally the coronary arteries themselves are segmented, but localization of nearby structures can also be helpful. Different segmentation approaches will be explored, such as region growing or active contour models.
The student will get familiar with the literature of segmentation models and implement a Java routine / plugin.
Having followed (or being following) a course on image processing and being familiar with the Java programming language is required for this project. It takes place as a collaboration with the CVMR group of the CHUV and CIBM-Siemens.
- Supervisors
- Daniel Schmitter, daniel.schmitter@epfl.ch, 21 693 5136, BM 4.138
- Michael Unser, michael.unser@epfl.ch, 021 693 51 75, BM 4.136
- Davide Piccini (CVMR, CIBM-Siemens)