BACKGROUND: Hepatocellular carcinoma (HCC) is often difficult to diagnose in cytologic material and small tissue biopsies since histomorphologic information is minimal or absent. The potential for misdiagnosis is greatest in attempting to discriminate well-differentiated HCC from dysplastic hepatocytes in cirrhosis. We investigated the feasibility of developing artificial intelligence classification methods based on nuclear image analysis data for use as adjuncts to the morphologic diagnosis of HCC. EXPERIMENTAL DESIGN: Ninety hematoxylin-eosin stained histologic slides including 56 with well- to poorly differentiated HCC and 34 showing a morphologic continuum from normal to markedly dysplastic benign hepatocytes were assembled from four laboratories. A relatively inexpensive PC-based image analysis system was used to measure 35 nuclear morphometric and densitometric parameters of 100 nuclei in each specimen. The data were randomized into classification training and testing sets containing equal numbers of benign and HCC samples. Objective diagnostic classification criteria for HCC based on neural networks and multivariate discriminant functions (DFs) were developed for the most discriminatory subsets of morphometric, densitometric, and combined morphometric/densitometric variables as selected by stepwise discriminant analysis of training data. RESULTS: Morphometric parameters provided the best results with the following testing data positive and negative predictive values (PV+ and PV-) for HCC classification: 86.2% PV+ and 81.3% PV- for a linear DF, 85.7% PV+ and 76.5% PV- for a quadratic DF and 100% PV+ and 85.0% PV- for a neural network. CONCLUSIONS: Our results demonstrate that nuclear image analysis-based objective classification criteria for HCC can be developed using artificial intelligence methods and that histologic material prepared at different institutions can be reliably classified. Neural networks for HCC classification were superior to linear and quadratic DFs. Morphometric data yielded the best results compared with densitometric or combined morphometric/densitometric data.
BACKGROUND:Hepatocellular carcinoma (HCC) is often difficult to diagnose in cytologic material and small tissue biopsies since histomorphologic information is minimal or absent. The potential for misdiagnosis is greatest in attempting to discriminate well-differentiated HCC from dysplastic hepatocytes in cirrhosis. We investigated the feasibility of developing artificial intelligence classification methods based on nuclear image analysis data for use as adjuncts to the morphologic diagnosis of HCC. EXPERIMENTAL DESIGN: Ninety hematoxylin-eosin stained histologic slides including 56 with well- to poorly differentiated HCC and 34 showing a morphologic continuum from normal to markedly dysplastic benign hepatocytes were assembled from four laboratories. A relatively inexpensive PC-based image analysis system was used to measure 35 nuclear morphometric and densitometric parameters of 100 nuclei in each specimen. The data were randomized into classification training and testing sets containing equal numbers of benign and HCC samples. Objective diagnostic classification criteria for HCC based on neural networks and multivariate discriminant functions (DFs) were developed for the most discriminatory subsets of morphometric, densitometric, and combined morphometric/densitometric variables as selected by stepwise discriminant analysis of training data. RESULTS: Morphometric parameters provided the best results with the following testing data positive and negative predictive values (PV+ and PV-) for HCC classification: 86.2% PV+ and 81.3% PV- for a linear DF, 85.7% PV+ and 76.5% PV- for a quadratic DF and 100% PV+ and 85.0% PV- for a neural network. CONCLUSIONS: Our results demonstrate that nuclear image analysis-based objective classification criteria for HCC can be developed using artificial intelligence methods and that histologic material prepared at different institutions can be reliably classified. Neural networks for HCC classification were superior to linear and quadratic DFs. Morphometric data yielded the best results compared with densitometric or combined morphometric/densitometric data.