Use of cell image analysis in the detection of cancer from specimens obtained during endoscopic retrograde cholangiopancreatography.

Abstract

The authors report a scoring system based on cell image analysis of specimens obtained during endoscopic retrograde cholangiopancreatography from either the biliary or pancreatic duct epithelium. Nine normal, 20 inflammatory, and 19 malignant epithelia were assessed using a cell-image processor that permits determination of morphometric (nuclear area), densitometric (nuclear DNA content), and textural (chromatin distribution) parameters on Feulgen-stained nuclei. The score is based on three densitometric parameters: the DNA histogram type, the nuclear DNA content, and the percent proliferating cell population. These parameters permit distinction of normal and inflammatory from malignant epithelium with 95% sensitivity and 100% specificity. Cell-image analysis of specimens obtained during endoscopic retrograde cholangiopancreatography is a useful clinical tool in the detection of malignant lesions of the pancreas and biliary epithelia.