SW pro modelování volumetrických MR obrazových dat mozkových metastáz

Abstract

This diploma thesis focuses on brain cancer problematics and on the development of SW for modeling MR volumetric image data of brain metastases. The theoretical part includes the basics from anatomy, principles of magnetic resonance, division of tumor types and description of segmentation methods, which are used in the practical part of this thesis. The practical part aims at the creation of algorithm for automatic segmentation and volumetry of brain tumors. By data preprocessing is highlighted the object of interest and simplify its segmentation. After that the segmentation algorithm is created to separate the tumor from surrounding tissue, which is not at the region of interest. From acquired image is possible to calculate volume to specify the size of tumor. Based on these methods could be created an intuitive software for automatic segmentation of brain tumors on MR data for clinical purposes.