AUTOMATED NUCLEAR ANALYSIS OF PROSTATE CANCER AND BENIGN PROSTATE HYPERPLASIA

摘要

Accurate measurement of the levels of androgen receptor (AR) protein expression is critical for prostate research. The Java-based automated image analysis system can: a) detect potential prostate nuclei; b) remove artifacts; c) classify prostate epithelial nuclei; and d) measure nuclear AR expression. A set of 200 immunonegative and 200 immunopositive nuclei are used to create a classification parameter set. The parameters reduce the effect of local variations in slide preparation and image collection on image analysis. The software performs local segmentation to detect prostatic epithelial nuclei and classify each nucleus as malignant or benign. AR expression is described by (a) percentage of immunopositive nuclei; (b) percentage of immunopositive nuclear area; and (c) intensity of AR expression among immunopositive nuclei or areas.