Search

Accurate segmentation of Glioblastoma multiforme (GBM) from MR images is important for sub-typing in diagnosis, determining tumor margins in surgical planning, and selecting appropriate therapies. However, it is a challenging and time-consuming task because GBM has a variety of imaging characteristics and often deforms nearby tissues in the brain. In this paper, we propose a support vector machine (SVM) active learning approach to address the problem of GBM segmentation from multi-modal MR images. First, a knowledge-based fuzzy clustering algorithm is performed to segment the brain tissues into six classes including white matter (WM), grey matter (GM), cerebrospinal fluid (CSF), T2-hyperintense regions, necrosis and enhanced tumor. Then, the SVM active learning approach is applied to refine the segmentation. Comparative studies with other segmentation methods indicate that the proposed algorithm can segment GBM more accurately. [ABSTRACT FROM PUBLISHER]

Glioblastoma multiforme and metastatic melanoma are frequent brain tumors in adults and presently still incurable diseases. Boron Neutron Capture Therapy (BNCT) is a promising alternative for this kind of pathologies. Accelerators have been proposed for BNCT as a way to circumvent the problem of siting reactors in hospitals and for their relative simplicity and lower cost among other advantages. Considerable effort is going into the development of accelerator-based BNCT neutron sources in Argentina. Epithermal neutron beams will be produced through appropriate proton-induced nuclear reactions and optimized beam shaping assemblies. Using these sources, computational dose distributions were evaluated in a real patient with diagnosed glioblastoma treated with BNCT. The simulated irradiation was delivered in order to optimize dose to the tumors within the normal tissue constraints. Using Monte Carlo radiation transport calculations, dose distributions were generated for brain, skin and tumor. Also, the dosimetry was studied by computing cumulative dose-volume histograms for volumes of interest. The results suggest acceptable skin average dose and a significant dose delivered to tumor with low average whole brain dose for irradiation times less than 60 minutes, indicating a good performance of an accelerator-based BNCT treatment. [ABSTRACT FROM AUTHOR]