Master of Science (MS)
First Committee Member
Evangelos A. Yfantis
Number of Pages
Our purpose is to aid medical doctors in prostate cancer detection via computer automated analysis of prostatic ultrasound imagery. Absorption of ultrasound signals is different in cancerous areas than in non-cancerous areas. The energy of the signal, the continuity of the signal, the autocorrelation function and frequency domain properties of prostatic ultrasound images are different in normal tissue than in cancerous tissue; This thesis presents an algorithm for automated cancer recognition in prostatic ultrasound imagery. Statistical and morphological based models are employed to classify regions of ultrasound imagery as either cancerous or non-cancerous. Application of our algorithm onto a limited set of cancerous and non-cancerous ultrasound images shows that our method has the ability to recognize cancer in cancerous ultrasound images. Misclassification occurs when cancerous tissue is classified as non-cancerous and noncancerous tissue is classified as cancerous. Occurrences of misclassification have been observed and investigated. (Abstract shortened by UMI.).
Cancer; Images; Prostate; Recognition; Ultrasound
Computer science; Oncology; Diagnostic imaging
University of Nevada, Las Vegas
If you are the rightful copyright holder of this dissertation or thesis and wish to have the full text removed from Digital Scholarship@UNLV, please submit a request to firstname.lastname@example.org and include clear identification of the work, preferably with URL.
Lazarakis, Theodores Dimitrios, "Prostate cancer recognition in ultrasound images" (2002). UNLV Retrospective Theses & Dissertations. 1456.
IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/