M.Sc. Tezi Görüntüleme

      Prostate is a gland and assists the male reproduction system. Prostate disorders like prostatitis, prostate hyperplasia and prostate cancer are common in males older than fifty. Like other cancers, prostate cancer can be treated well if detected early.

      Computer-aided diagnosis is one of the most active research fields in medical imaging. There are many publications in the literature reporting improved reliability of diagnosis techniques. A program is developed in this thesis to process and analyze prostate images acquired by Transrectal Ultrasound (TRUS) imaging in order to help physician decide suspected regions to be sampled for biopsy. The program segments prostate images by using k-means clustering of the outputs of a gabor filter bank. Then, clusters are used to form Gray-Level Co-occurrence Matrices (GLCM) and finally regions are classified.

      Keywords: TRUS, Biopsy, Computer aided diagnosis, Prostate cancer