LAUSR

BACKGROUND & AIMS Hepatocellular nodular lesions (HNLs) constitute a heterogeneous group of disorders. Differential diagnosis among these lesions, especially high-grade dysplastic nodule (HGDN) and well differentiated hepatocellular carcinoma (WD-HCC), can be challenging, let alone biopsy specimens. We aimed to develop a deep learning system to solve these puzzles, improving the histopathologic diagnosis of HNLs (WD-HCC, HGDN, low-grade DN, focal nodular hyperplasia, hepatocellular adenoma) and background tissues (nodular cirrhosis, normal liver tissue). METHODS The samples consisting of surgical specimens and biopsies were collected from six hospitals. Each specimen was reviewed by 2-3 subspecialists. Four deep neural networks (ResNet50, InceptionV3, Xception and the Ensemble) were employed. The performances were evaluated by confusion matrix, ROC curve, classification map and heatmap. Furthermore, the predictive efficiency of the optimal model was verified by comparing with that of nine pathologists. RESULTS A total of 213,280 patches from 1115 WSIs of 738 cases were obtained. An optimal model was finally chosen based on F1 score and AUC value, named hepatocellular-nodular artificial intelligence model (HnAIM), with the overall seven-category AUC of 0.935 in the independent external validation cohort. For biopsy specimens, the agreement rate with subspecialists' majority opinion was higher for HnAIM than nine pathologists on both patch-level and WSI-level. CONCLUSIONS We first developed a deep learning diagnostic model for HNLs, which performed well and contributed to enhancing the diagnosis rate of early HCC and risk stratification of HNLs patients. Furthermore, HnAIM had significant advantages in patch-level recognition, with important diagnostic implications for fragmentary or scarce biopsy specimens.