Primary liver cancer classification from routine tumour biopsy using weakly supervised deep learning

• DOI: 10.1016/j.jhepr.2024.101008
• Publication Date: 2024-01-21T09:43:07Z
• Authors: Aurélie Beaufrère; Nora Ouzir; Paul Emile Zafar; Astrid Laurent-Be...; Miguel Albuquerque; Gwladys Lubuela; Jules Grégory; Catherine Guettier; Kévin Mondet; Jean-Christophe P...; Valérie Paradis
• Keywords: [INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI]; I.2; FOS: Computer and information sciences; [SDV.MHEP.AHA] Life Sciences [q-bio]/Human health and pathology/Tissues and Organs [q-bio.TO]; I.4; I.5; Computer Science - Artificial Intelligence; Computer Vision and Pattern Recognition (cs.CV); [INFO.INFO-IM] Computer Science [cs]/Medical Imaging; Computer Science - Computer Vision and Pattern Recognition; 610; [SDV.CAN]Life Sciences [q-bio]/Cancer; RC799-869; [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]; [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG]; [SDV.CAN] Life Sciences [q-bio]/Cancer; 617; [INFO.INFO-IM]Computer Science [cs]/Medical Imaging; [SDV.MHEP.AHA]Life Sciences [q-bio]/Human health and pathology/Tissues and Organs [q-bio.TO]; biopsy; Tissues and Organs (q-bio.TO); I.5; I.4; I.2; Primary liver cancer; histological slides; [SDV.MHEP.HEG]Life Sciences [q-bio]/Human health and pathology/Hépatology and Gastroenterology; Quantitative Biology - Tissues and Organs; [INFO.INFO-LG] Computer Science [cs]/Machine Learning [cs.LG]; Diseases of the digestive system. Gastroenterology; artificial intelligence; [SDV.MHEP.HEG] Life Sciences [q-bio]/Human health and pathology/Hépatology and Gastroenterology; Artificial Intelligence (cs.AI); FOS: Biological sciences; weakly supervised learning; Research Article

The diagnosis of primary liver cancers (PLCs) can be challenging, especially on biopsies and for combined hepatocellular-cholangiocarcinoma (cHCC-CCA). We automatically classified PLCs on routine-stained biopsies using a weakly supervised learning method. Weak tumour/non-tumour annotations served as labels for training a Resnet18 neural network, and the network's last convolutional layer was used to extract new tumour tile features. Without knowledge of the precise labels of the malignancies, we then applied an unsupervised clustering algorithm. Our model identified specific features of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (iCCA). Despite no specific features of cHCC-CCA being recognized, the identification of HCC and iCCA tiles within a slide could facilitate the diagnosis of primary liver cancers, particularly cHCC-CCA. Method and results: 166 PLC biopsies were divided into training, internal and external validation sets: 90, 29 and 47 samples. Two liver pathologists reviewed each whole-slide hematein eosin saffron (HES)-stained image (WSI). After annotating the tumour/non-tumour areas, 256x256 pixel tiles were extracted from the WSIs and used to train a ResNet18. The network was used to extract new tile features. An unsupervised clustering algorithm was then applied to the new tile features. In a two-cluster model, Clusters 0 and 1 contained mainly HCC and iCCA histological features. The diagnostic agreement between the pathological diagnosis and the model predictions in the internal and external validation sets was 100% (11/11) and 96% (25/26) for HCC and 78% (7/9) and 87% (13/15) for iCCA, respectively. For cHCC-CCA, we observed a highly variable proportion of tiles from each cluster (Cluster 0: 5-97%; Cluster 1: 2-94%).<br/>https://www.sciencedirect.com/science/article/pii/S2589555924000090<br/>