After successful completion of the course, students are able to to process data generated by medical imaging modalities and to develop algorithms for processing the data, analyzing the observed structures and quantifying disease- and treatment-relevant markers. They are able to identify and implement modern methods of machine learning that are optimal for problems concerning computer-aided diagnosis, prognosis, and the prediction of disease progression or treatment effects. After positive completion of the course, students are able to implement algorithms for segmentation, model-based detection, texture analysis, interactive segmentation, rigid and non-rigid registration, and functional imaging analysis.

Dear colleagues the lecture and exercise Medical Image Processing will be held this semester (SoSe 2020). The course will be given via Zoom. Connection information will be provided in TUWEL

We will discuss the following topics in the course of the lecture:

• Medical imaging modalities
• Segmentation (active contours, level-set)
• Model based detection and segmentation of anatomical structures (active shape models, active appearance models)
• Texture analysis
• Interactive segmentation (graph cuts)
• Rigid and non-rigid registration
• Neuroimaging and machine learning, analysis of neuroimaging data
• Applications in interoperative / interventional visualization
• Atlas building

Methods and modalities will be explained based on real world cases. For each we will discuss the mathematical bassics, and ways of solving it. For each unit we will distribute reading material, so that we can have an interesting discussion during the lecture.

In the course of the lab exercise 183.630 we will implement and test selected methods on medical imaging data.

This semester the course will be held online via TUWEL and Zoom. All relevant information including slides and links to the online lectures will be provided via TUWEL.

The course consists on the one hand of a detailed discussion of methodical approaches for image acquisition and analysis, and on the other hand, of algorithmic solutions developed in interactive discussions based on case studies. On the one hand, the basics are taught, on the other hand, the ability to combine these methods to an effective solution approach is acquired, which starts from a problem description (e.g. detection of a tumor, quantitative tracking of disease and treatment progression, examination of large groups of patients, use of algorithms in clinical practice).

Written Exam