APPLICATION OF AUTOMATIC THRESHOLDING IN IMAGE-ANALYSIS SCORING OF CELLS IN HUMAN SOLID TUMORS LABELED FOR PROLIFERATION MARKERS

We previously reported an image analysis program that uses four invest igator-defined parameters including two thresholds, i.e., gray-level t hreshold (GLT) and hue threshold (HT), to determine the number of cell s (TC) and the proliferating cell nuclear antigen (PCNA) or bromodeoxy uridine (BrdUrd) labeling indices (LI) in human solid tumors, The pres ent study investigated if the accuracy and reproducibility of image an alysis results can be improved by using computer-defined GLT and NT. T hree investigators evaluated 142 images on 3 days, using visual analys is and four image analysis routines, which used different combinations of computer and investigator-selected GLT and MT. The data show that image analysis using computer-selected GLT and HT yielded (i) LI of PC NA aad BrdUrd that were indistinguishable from visual analysis, (ii) e qual (BrdUrd LI) or better (TC and PCNA LI) inter-day reproducibility relative to visual analysis results, and (iii) results that were equal ly (TC) or more accurate (LI of PCNA and BrdUrd) with higher inter-day reproducibility (TC and LT of PCNA and BrdUrd) than image analysis ob tained using investigator-defined thresholds. We conclude that the use of computer-defined GLT and HT improved the accuracy and reproducibil ity of image analysis results. (C) 1997 Wiley-Liss, Inc.