Medical Image Analysis - IN 3200

Lecture: Prof. Koen Van Leemput (Harvard/DTU; August-Wilhelm Scheer Visiting Professor at TUM), Prof. Bjoern Menze (TUM)

Organization: Cagdas Ulas

Time & Location

• BLOCK course: week of June 20, week of June 27 (starting Tue July 21)
• Morning lectures: IMETUM lecture hall and IMETUM seminar room (Check the below course schedule!)
• Afternoon exercises: IMETUM seminar room and IMETUM lecture hall (Check the below course schedule!)

Overview

This lecture will introduce participants into basic algorithmic concepts for image registration and image segmentation and the underlying theoretical concepts. Emphasis is on the transfer of these concepts into hand on knowledge by implementing algorithms in Matlab and evaluating them to real clinical data sets. The first part will deal with registration, for example, fitting image functions; landmark and intensity based registration; non-linear deformations. The second part will introduce into segmentation methods, with active contour model, voxel-based segmentation (mixture models; expectation maximization, Markov random field priors), and probabilistic population atlases

Prerequisites

• Basic knowledge in image analysis / computer vision
• Interest in algorithmic/implementation aspects
• Ability to work in small teams
• Specific knowledge in MATLAB is not required but beneficial. However, you should be somehow skilled in programming.

Regulations

• Module Spezification: Selected Topics in Computer Graphics and Vision (IN 3200)
• SWS: 2+2
  
• ECTS: 5 Credits
  
• Course Language: English
  
• Exam: oral
  

Course Schedule

Lecture Notes

Course note: Medical Image Analysis