Medical Image Processing

Relations:

It is not possible to register for the course BEAM33ZMO if the student is concurrently registered for or has already completed the course BAM33ZMO (mutually exclusive courses).

It is not possible to register for the course BEAM33ZMO if the student is concurrently registered for or has already completed the course A6M33ZMO (mutually exclusive courses).

In order to register for the course BEAM33ZMO, the student must have registered for the required number of courses in the group BEZBM no later than in the same semester.

It is not possible to register for the course BEAM33ZMO if the student is concurrently registered for or has previously completed the course BAM33ZMO (mutually exclusive courses).

Course guarantor:

Jan Kybic

Lecturer:

Jan Kybic, Oleksandr Shekhovtsov

Tutor:

Denis Baručić, Jan Kybic

Supervisor:

Department of Cybernetics

Synopsis:

This subject describes algorithms for digital image processing of 2D and 3D images, with emphasis on biomedical applications. We shall therefore concentrate on the most often used techniques in medical image processing: segmentation, registration, and classification. The methods will be illustrated by a range of examples on medical data. The students will implement some of the algorithms during the practice sessions.

Because of the very large overlap between courses A6M33ZMO and A4M33ZMO, the courses will be taught together this year.

Requirements:

The knowledge of basic signal processing methods including a Fourrier transform, and the knowledge of the basic principles of medical imaging methods.

Syllabus of lectures:

1. Segmentation - active contours, level sets

2. Segmentation - shape models,

3. Segmentation - superpixels, random walker, GraphCuts, graph search, normalized cuts

4. Segmentation - texture, texture descriptors, textons

5. Segmentation - CNN, U-net

6. Detection of cells and nuclei

7. Detection of vessels and fibers

8. Detection of nodules and mammographic lesions

9. Localization of organs and structures

10. Registration - ICP, coherent point drift, B-splines, rigid registration, multiresolution

11. Registration - rigid, elastic, daemons

12. Registration by CNN

Syllabus of tutorials:

Individual works will consist of independent practical work in a computer laboratory involving the use of algorithms covered by the course for analysis of specific medical data. Some algorithms will be implemented from scratch and some using existing freely available libraries and toolkits. Apart from a general overview, the students will gain a deeper understanding of some of the methods and will learn to apply them to practical problems.

Study Objective:

Learn the principles and usage of basic algorithms for medical image processing, such as registration, segmentation and classification. The students will learn to implement some of the algorithms.

Study materials:

[1] Sonka M., Fitzpatrick J. M.: Handbook of Medical Imaging, vol.2. SPIE Press, 2000.

[2] Bankman, I. Handbook of Medical Imaging, Processing and Analysis, vol.1. Academic Press, 2000.

Note:

Further information:

https://cw.fel.cvut.cz/wiki/courses/zmo

Time-table for winter semester 2024/2025:

	06:00–08:0008:00–10:0010:00–12:0012:00–14:0014:00–16:0016:00–18:0018:00–20:0020:00–22:0022:00–24:00
Mon
Tue
Wed	roomKN:E-127 Kybic J. Shekhovtsov O. 09:15–10:45 (lecture parallel1) Karlovo nám. roomKN:E-230 Baručić D. 11:00–12:30 (lecture parallel1 parallel nr.101) Karlovo nám.
Thu
Fri

Time-table for summer semester 2024/2025:

Time-table is not available yet

The course is a part of the following study plans:

• Medical Electronics and Bioinformatics - Specialization Image Processing (PS)
• Medical Electronics and Bioinformatics - Specialization Signal Processing (compulsory elective course)
• Medical Electronics and Bioinformatics - Specialization Bioinformatics (compulsory elective course)
• Medical Electronics and Bioinformatics - Specialization Medical Instrumentation (compulsory elective course)