Modern approaches of biomedical image analysis
FEKT-DBT4Acad. year: 2017/2018
The course focuses on the issue of registration ( alignment ) image data with examples of applications especially in medicine and related fusion image information. The basic methods and components of mono- and multi - modal registration with a variety of rigid and flexible registration are discussed. Students will deepen their knowledge in the field of texture analysis of image data including the issue of invariance texture features .
Learning outcomes of the course unit
Recommended optional programme components
Recommended or required reading
Hajnal, J.V., Hill, D.L.G., Hawkes, D.J.: Medical Image Registration, CRC Press, 2001
M. Petrou, Sevilla, P.G. Dealing with Texture, Wiley, 2006
Planned learning activities and teaching methods
Techning methods include seminars. Course is taking advantage of e-learning (Moodle) system. Students have to write a single project during the course.
Assesment methods and criteria linked to learning outcomes
Language of instruction
1. Principles of registration ( alignment ) medical image data.
2. Fusion of image information.
3. Basic methods and components mono- and multi - modal registration .
4. Geometric transformations.
5. Metrics for registration.
6. Interpolation techniques.
7. Optimization approaches.
8. Overview of methods of texture analysis .
9. Problems of feature invariance.
10. Markov random fields texture models.
11. Methods of local binary patterns.
12. Methods based on higher-order statistics .
The aim of the course is to provide an overview of the two current topics in the field of image processing - image registration and texture analysis of image data.