Segmentation Methods for Medical Image Analysis (in English) 2018/2019

Termin(e)
Lecture

Tuesday, 1pm-3pm; Thursday, 11am-1pm G29-335

Klaus Tönnies; Tim König

Today’s segmentation algorithms solve the problem as an optimization task. The desired segmentation is described as an optimal function defined on the image that combines data knowledge and pre-specified expectations about the segments (the so-called model terms). Within this course, segmentation paradigms such as level sets or graph cuts are presented that state the optimality condition as a functional on a continuous or discrete domain. Methods are presented to find the optimal segmentation given this condition. Furthermore, strategies, such as neural networks, are presented that are able to learn part or all of this optimality condition from training data without having it to be defined explicitly.

Studienrichtungen:
Computervisualistik (86152) Master 1 – 2 WPF
Computervisualistik (88152) Master (BK) 1 – 2 WPF
Data and Knowledge Eng. (86156) Master 2 – 3 WPF
Digital Engineering (86158) Master 3 – 3 WPF
Informatik (86150) Master 1 – 2 WPF
Informatik (88150) Master (BK) 1 – 2 WPF
Ingenieurinformatik (86157) Master 1 – 2 WPF
Ingenieurinformatik (88157) Master (BK) 1 – 2 WPF
Wirtschaftsinformatik (86159) Master 1 – 2 WPF
Wirtschaftsinformatik (88159) Master (BK) 1 – 2 WPF

Credits: 6

Sprache: Deutsch


Accompanying Project


Mandatory part of the course is a programming project, where participants decide on a solution for a segmentation problem and then implement and test it.

Documents


23.10.2018ASMI01-Introduction.pdf
23.10.2018ASMI02-Gradient-Descent.pdf
23.10.2018ASMI03-Total-energy-minimization.pdf
23.10.2018ASMI04-Active-Contours.pdf
23.10.2018ASMI05-Level-Set-Segmentation-Introduction.pdf
23.10.2018ASMI06-Variational-Level-Sets.pdf
23.10.2018ASMI07-Variational-Level-Sets.pdf
23.10.2018ASMI08-Graph-Cuts.pdf
23.10.2018ASMI09-Graph-Cuts.pdf
23.10.2018ASMI10-Trained-Segmentation.pdf
23.10.2018ASMI11-Segmentation-by-Neural-Networks.pdf
23.10.2018ASMI12-Deep-Learning-and-Segmentation.pdf
Project
25.10.2018Introduction.pdf
25.10.2018data_train.zip
06.12.2018data_test.zip