This paper presents a novel algorithm for prostate tissue characterization based on Trans-rectal Ultrasound (TRUS) images. A Gabor multi-resolution technique is designed to automatically identify the Regions of Interest (ROI) in the segmented prostate image. These ROIs are the high probable cancerous regions in the gland. Furthermore, statistical texture analysis for these regions is carried out by employing Grey Level Difference Matrix (GLDM), where a set of features is constructed. The next stage is mainly feature selection that defines the most salient subset of the constructed features using exhaustive search. The selected feature set is found to be useful for the discrimination between cancerous and non-cancerous tissues. Least Square Support Vector Machines (LS-SVM) classifier is then applied to the selected feature set for the purpose of tissue characterization. The obtained results demonstrate excellent tissue characterization. © Springer-Verlag 2004.
CITATION STYLE
Mohamed, S. S., Salama, M. M. A., Kamel, M., & Rizkalla, K. (2004). Region of Interest Based Prostate Tissue Characterization Using Least Square Support Vector Machine LS-SVM. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3212, 51–58. https://doi.org/10.1007/978-3-540-30126-4_7
Mendeley helps you to discover research relevant for your work.