Modeling prostate cancer distributions

Objectives. To use computer-assisted three-dimensional (3D) reconstruction to study cancer distribution in the prostate. The distributions have been d etermined using data from 81 prostates surgically removed because of cancer . Methods. The pattern of distribution was determined by reshaping (morphing) all prostates in the same size category (small, medium, large) into the sa me shape and by overlayering the 3D distribution of all cancers into a "uni t prostate." Also, the pattern of 3D distribution was defined for small, me dium, and large volume cancers. Results. The study yielded a clear, visual "scatter diagram" representation of the statistics of prostate cancer distribution. It confirmed previous g eneral knowledge: small cancers are mainly found in the lower one third and principally in the dorsolateral part of the gland. Conclusions. This report is part of the investigations preliminary to estab lishing an optimal protocol for needle biopsy of prostate cancer. (C) 1999, Elsevier Science Inc.