Analýza biomedicínských snímků magnetické rezonance při roztroušené skleróze

Abstract

The goal of this bachelor’s thesis is to design and accomplish a tool for image processing using FOTOMNG program. The tool is supposed to be used for processing of magnetic resonance imaging, specifically of sclerosis multiplex. The whole module will be programmed in the Java programming language in the software development platform NetBeans and will be implemented into the system FOTOMNG. The final algorithm will select the most suitable segmentation method based on the quality metrics of the examined image. It will select from the segmentation methods which are parametric active contours and Level Set. Image pre-processing will also take place. Converting an image to grayscale, median filtering and sobel filtering will be used. At the end the success of the selected segmentation methods will be assessed. The theoretical part of this thesis will deal with sclerosis multiplex, magnetic resonance, photogrammetry, the system FOTOM, solutions of this problem in the world and the general division of the segmentation methods. The goal is also to conceive a user manual and programming documentation.