Automatic Tumor Segmentation Using Knowledge-Based Techniques

Matthew C. Clark, Lawrence O. Hall, Dmitry B. Goldgof, Robert Velthuizen,¹, F. Reed Murtagh¹, and Martin S. Silbiger¹


Department of Computer Science and Engineering
¹ Department of Radiology
University of South Florida
Tampa, Fl. 33620
hall@csee.usf.edu

ABSTRACT

A system that automatically segments and labels glioblastoma-multiforme tumors in magnetic resonance images of the human brain is presented. The magnetic resonance images consist of T1-weighted, proton density, and T2-weighted feature images and are processed by a system which integrates knowledge-based techniques with multi-spectral analysis. Initial segmentation is performed by an unsupervised clustering algorithm. The segmented image, along with cluster centers for each class are provided to a rule-based expert system which extracts the intra-cranial region. Multi-spectral histogram analysis separates suspected tumor from the rest of the intra-cranial region, with region analysis used in performing the final tumor labeling. This system has been trained on three volume data sets and tested on thirteen unseen volume data sets acquired from a single magnetic resonance imaging system. The knowledge-based tumor segmentation was compared with supervised, radiologist-labeled ``ground truth'' tumor volumes and supervised k-nearest neighbors tumor segmentations. The results of this system generally correspond well to ground truth, both on a per slice basis and more importantly in tracking total tumor volume during treatment over time.

Keywords: Knowledge-Based, Magnetic Resonance Imaging, Tumor Segmentation, Multi spectral Analysis, Region Analysis, Clustering, Unsupervised Classification