Chapter 9:
Computer-Aided Prostate Cancer Diagnosis: Principles, Recent Advances, and Future Prospective
Editor(s): Jinshan Tang Sos S. Agaian
Published: 2013
DOI: 10.1117/3.1002311.ch9
Abstract
Over the next few years, a dramatic increase in the number of people developing different types of cancer is expected. It is estimated from global statistics that 10 million new cancer patients are diagnosed each year, and this number will double by the year 2020. Although several billions of dollars are spent on cancer research, there is still no definitive cure for this disease. Meanwhile, various techniques have been developed to be used in all phases of cancer diagnosis and management with the aim of helping doctors define an appropriate treatment plan for patients and monitor treatment efficacy. Imaging forms an essential part of cancer clinical protocols and can provide morphological, structural, metabolic, and functional information. As a result, cancer diagnosis is becoming increasingly image-based. There are three main objectives of using imaging with respect to cancer management: (1) the earliest possible detection of benign or malignant lesions or tumors, which is probably the strongest factor in reducing mortality for certain cancers; (2) correlation of imaging results with other clinical parameters; and (3) accurate staging and follow-up after treatment. In order to aid radiologists and pathologists in cancer diagnosis, various computer-based methods have been developed to detect, classify, and grade the malignancy of tumors by using certain visual criteria. Due to the importance of imaging in cancer diagnosis and treatment, computer-aided diagnosis (CAD) has become one of the major research subjects in medical imaging and diagnostic radiology and pathology. The goal of CAD is to increase the productivity of radiologists/pathologists by improving the accuracy and consistency of diagnoses, reducing image reading time, and providing computer-based tools for image visualization and annotation. The general approach for CAD development is to find the location of a lesion (detection) and also to estimate the probability of the presence of disease (differential diagnosis). Ultimately, a CAD system may become an integrated tool in all areas of medical imaging. An important benefit of using computers to diagnose disease is the reproducibility and consistency of the diagnostic methods because the performance of computers is not affected by fatigue, perceptual errors, or variability in classification criteria. The use of computers to assist clinical diagnosis is well established in radiology. In 1998, the U. S. Food and Drug Administration (FDA) approved the clinical use of the first CAD system for mammography developed for detection of breast cancer. Since then, other systems for disease detection have been used. For instance, several CAD systems have been employed in breast screening or have been prospectively analyzed in order to be used in clinical practice for breast cancer detection. Support tools for skin cancer and leukemia have been also presented. In addition, several CAD systems for prostate cancer diagnosis have been developed with the aim of providing diagnostic information obtained from quantitative image analysis. The information provided by accurate CAD systems regarding cancer localization and grade may be used by doctors as a "second opinion" in order to make final diagnosis decisions.
Online access to SPIE eBooks is limited to subscribing institutions.
CHAPTER 9
40 PAGES


SHARE
Back to Top