PROSTATE TUMOR LOCALIZATION USING MULTIPARAMETRIC MAGNETIC RESONANCE IMAGING

This was my Master thesis that I’ve written during 2011 in collaboration with the MR Cancer Group of Norwegian University of Science and Technology in Trondheim, Norway. It was a great pleasure to work with such talented people!

Synopsis:

Prostate cancer is the leading cancer disease diagnosed in males. More than 70 percentage of the diagnosed cancers are located in the peripheral zone of the prostate. Magnetic resonance imaging (MRI) has been established as a good method for diagnosis, staging and treatment-controlling of the tumor. Different methods of MRI are used to identify tumor tissue based on radiological assessment. This paper aims to develop a more objective way to visualize tumor tissue in different MRI modalities using image processing methods. Actual medical data were used to make an effective algorithm based on new Matlab scripts to filter the MR images and visualize tumor tissue found in the peripheral zone of the prostate. The results were designed for easy use by the urologists and were tested on a patient cohort of 10 patients with different cancers.

Functional and Non-Functional Modalities were used for image processing

Different filters were set up, to validate the found tumor areas by different approaches.