Computational support of dynamic contrast enchancement magnetic resonance imaging of head and neck tumors

OData support
Dr. Antal Péter
Department of Measurement and Information Systems

In the last 25 years the MRI is one of the most frequently used technics in tumor localization. In early years of the previous decade the oncologists noticed, that different types of tissue respond to the contrast agent in different way. This gave a new aim to MRI research – to use DCE-MRI technic to identify different type of tumors. In this work I created a tool, which can deal with motion artifacts and offer a color-coded represantation of the MRI image for easier tissue differentiation. I studied a lot of scientific literature. The first part of it contained information about image processing and motion tracking. The second part of the literature I looked through contained research overviews of MRI-based tissue differentiation.

I implemented a few image registration algorithms with different approaches. The accuracy of these approaches were measured on test data.

The algorithm with the best performance was tested on live data.

The colorcoding was performed basing on time-intensity curve of each pixel and was tested on numerous real MRI data.


Please sign in to download the files of this thesis.