Search

Background and Aim: Conventional endoscopy for the early detection of esophageal and esophagogastric junctional adenocarcinoma (E/J cancer) is limited because early lesions are asymptomatic, and the associated changes in the mucosa are subtle. There are no reports on artificial intelligence (AI) diagnosis for E/J cancer from Asian countries. Therefore, we aimed to develop a computerized image analysis system using deep learning for the detection of E/J cancers., Methods: A total of 1172 images from 166 pathologically proven superficial E/J cancer cases and 2271 images of normal mucosa in esophagogastric junctional from 219 cases were used as the training image data. A total of 232 images from 36 cancer cases and 43 non-cancerous cases were used as the validation test data. The same validation test data were diagnosed by 15 board-certified specialists (experts)., Results: The sensitivity, specificity, and accuracy of the AI system were 94%, 42%, and 66%, respectively, and that of the experts were 88%, 43%, and 63%, respectively. The sensitivity of the AI system was favorable, while its specificity for non-cancerous lesions was similar to that of the experts. Interobserver agreement among the experts for detecting superficial E/J was fair (Fleiss' kappa = 0.26, z = 20.4, P < 0.001)., Conclusions: Our AI system achieved high sensitivity and acceptable specificity for the detection of E/J cancers and may be a good supporting tool for the screening of E/J cancers., (© 2020 Journal of Gastroenterology and Hepatology Foundation and John Wiley & Sons Australia, Ltd.)

Background & Aims: GALAD and BALAD-2 are statistical models for estimating the likelihood of the presence of hepatocellular carcinoma (HCC) in individual patients with chronic liver disease and the survival of patients with HCC, respectively. Both models use objective measures, particularly the serum markers α-fetoprotein (AFP), AFP-L3, and des-γ-carboxyprothrombin. We aimed to validate these models in an international cohort of patients with HCC and assess their clinical performance., Methods: We collected data on cancer diagnosis and outcomes of 6834 patients (2430 with HCC and 4404 with chronic liver disease) recruited from Germany, Japan, and Hong Kong. We also collected data from 229 patients with other hepatobiliary tract cancers (cholangiocarcinoma or pancreatic adenocarcinoma) and 92 healthy individuals (controls). For reference, the original UK cohort (on which the GALAD model initially was built and BALAD-2 was validated) was included in the analysis. We assessed the effects of tumor size and etiology on GALAD model performance, and its ability to correctly discriminate HCC from other hepatobiliary cancers. We assessed the performance of BALAD-2 in patients with different stages of HCC., Results: In all cohorts, the area under the receiver operating characteristic curve (AUROC), quantifying the ability of GALAD to discriminate patients with HCC from patients with chronic liver disease, was greater than 0.90-similar to the series on which the model originally was built (AUROC, 0.97). GALAD discriminated patients with HCC from those with other hepatobiliary cancers with an AUROC value of 0.95; values were slightly lower for patients with small unifocal HCCs, ranging from 0.85 to 0.95. Etiology and treatment of chronic viral hepatitis had no effect on the performance of this model. BALAD-2 analysis assigned patients with HCC to 4 distinct prognostic groups-overall and when patients were stratified according to disease stage., Conclusions: We validated the performance of the GALAD and BALAD-2 models for the diagnosis of HCC and predicting patient survival, respectively (based on levels of the serum markers AFP, AFP-L3, and des-γ-carboxyprothrombin), in an international cohort of almost 7000 patients. These systems might be used in HCC surveillance and determination of patient prognosis., (Copyright © 2016 AGA Institute. Published by Elsevier Inc. All rights reserved.)