Abstract:
Segmentation of tissues in medical imaging is an essential subject because it helps the radiologists to be able to identify diseases, tumors and follow the degenerative diseases. In Magnetic Resonance Imaging (MRI) one factor that causes a problem during segmentation is the inhomogeneity in the magnetic field. Mainly the RF coil inhomogeneity effect causes intensity inhomogeneity through the image. This intensity inhomogeneity may cause segmentation algorithms to fail for a specific imager system. In case an algorithm that can be used in many imagers is needed the difference in the tissue intensities and the RF coil inhomogeneity change may cause greater failures. To overcome this problem a method which uses calculated T1, T2 and proton density parameters is proposed. These parameters are calculated from MRI images using four sampling points (four sets of images of the same region with different parameters) and using Levenberg-Marquardt Method. Then maximum likelihood classification is applied to distinguish the tissues and the segmented images were constructed. Gray Matter, White Matter and Cerebrospinal Fluid were segmented in MR brain images of seven volunteers. The subject heads were scanned with three different MR imagers. Tissue segmentation was performed with the weighted T1, T2 and Proton Density images along with the computed true T1, T2 and PD. Comparisons across image slices; across imagers and across subjects indicated that significant improvement can be achieved if the computed T1, T2 and PD images are used for the segmentation of brain tissue.|Keywords: T1, T2, PD, Levenberg-Marquardt, maximum likelihood classification, RF coil inhomogeneity .
