Analýza dat MR břišní oblasti pro detekci morfologické struktury pankreatu

Abstract

This bachelor thesis deals with the creation of a testing database and semi-automatic segmentation of the pancreas using a magnetic resonance imaging/scanner. In the theoretical part, I dealt with the anatomy, physiology, pathophysiolog,y and pathology of the pancreas and delineate the basic principle of magnetic resonance imaging. Part of the thesis is a state of the art for suitable more sequences when scanning the pancreas and a state of the art for an image segmentation. For data generation were I used magnetic resonance imaging using Sempra Siemens 1.5 T Magnetom, and a T1-Vibe-Dixon. A semi-autonomous threshold method called Soft-Thresholding was used to segment the image. The data were segmented and processed in Matlab. Robustness testing was performed using Rician's and Salt&Pepper noise. The method can be used in medical diagnostics in the foreseeable future.