One man out of 11 develops adenocarcinoma of the prostate. Accurate detection and staging of the cancer are essential for its proper treatment and for the extension of patient life expectancy.

In a review of the uses of computer technology in the diagnosis and staging of prostate cancer, Yanong Zhu of the University of East Anglia in Norwich and his colleagues at Norfolk and Norwich University Hospital and at the University of Wales in Aberystwyth, all in the UK, note that a variety of imaging tools are involved in these efforts. Furthermore, improvements in computer technologies may see software approaches integrated with more conventional detection and staging methodologies.

The large amount of data derived from clinical examinations, tumor pathology and markers, and imaging studies can be difficult to integrate to draw out the best therapeutic approach for a patient with prostate cancer. Artificial neural networks may have a role in aiding medical decision-making in this complex area. Also, computerized simulations of prostate biopsies are becoming useful for evaluating existing physical biopsy schemes and designing better biopsy strategies.

The authors discuss conventional detection and staging methods, including digital rectal exams and the use of prostate-specific antigen; the ability of artificial neural networks to detect and stage disease; computerized visualization and simulation of prostate biopsies; and computer-aided analysis of ultrasound and MRI images.

They say that the remaining challenges include the further development of statistical shape-based image segmentation techniques, the registration of intra- and intermodality data, and the inclusion of data regarding the surrounding tissue. With these challenges met, they foresee the realization of a full 3-D model of the prostate that can be integrated into clinical evaluation tools. (Medical Image Analysis, April 2006, pp. 178-199.)