||The aim of this study is to develop a method to quantitate the Prostate-Specific Acid Phosphatase (PSAP) in double stained histologic sections of prostate tumor tissue, in which the PSAP is labeled with the peroxidase-antiperoxidase (PAP) complex technique using diaminobenzidine (DAB) as a substrate (brown PAP-DAB stain) while the nuclei are stained with hematoxylin (blue stain). After development of the method, studies of clinical prostate tissue sections of RTOG (radiation therapy of oncology group) protocols will be performed. In order to make the measurement of the brown PAP-DAB stain independent in the second stain (blue stain), a new color segmentation method, called dual wavelength technique, is developed to separate the brown stain and blue stain components in actual stained slides. Two lOnm bandpass filters, peaked at 450nm and 510nm wavelength, are used in the image acquisition. Using the wavelength dependent mass absorptivity ratios of the two stains estimated in slides stained only brown or blue, the blue stain and the brown stain are completely separated by the dual wavelength technique. The blue stain is located in the nuclei and the brown stain is located in the prostate gland cytoplasm. Using the mass absorptivity of the brown PAP-DAB stain obtained in a simulating experiment, an estimation of the brown PAP-DAB stain mass is obtained. The reliability, investigated by comparing the measured mass with the true value, is indicated by a correlation coefficient 0.99, and the linear relationship with the slope 1.097 and intercept 0.00022. A quantitative evaluation of the PAP-DAB stain intensity and extent in clinical prostatic tumor sections is taken on the stain images completely separated by this technique. The PAP-DAB stain intensity is estimated by the average stain mass per stain area (8) and the average stain mass taken up by one cell (I), while the stain extent is measured by the average stain area taken up by one cell (E). In a preliminary study, it has been demonstrated that this evaluation is precise and reproducible. In order to limit the variations resulting from section thickness and stain conditions, the degree of PAP-DAB stain intensity and extent of one slide are evaluated by comparing the features of tumor glands with those of control glands in the same slide. A batch mode process, for evaluating a large group of slides, is developed using a PC 286 based imaging system and a SUN system. Using this batch mode processing technique, the evaluation of a large group of slides can be repeated without image re-acquisition. The PC based imaging system is connected to a SUN system through ETHERNET. The sequential steps involved in this batch mode process are image acquisition on the PC based imaging system, image transfer through ETHERNET, and image evaluating on SUN workstation.