Predicting Tumor Recurrence On Baseline Mr Imaging In Patients With Early-stage Hepatocellular Carcinoma Using Deep Machine Learning

This study aimed to build, validate, and test a deep learning approach to predict the post-treatment recurrence risk in early-stage hepatocellular carcinoma (HCC) patients from pre-treatment MR imaging. A total of 120 adults from 2005 to 2018 were included in the single-center retrospective study, which was compliant with the Health Insurance Portability and Accountability Act (HIPAA). The study was approved by the institutional review board of Yale University and informed consent was waived. Pre-treatment clinical variables were collected and patient data were extracted from the Electronic Health Records. The MR images were read by three board-certified radiologists to ensure all patients were diagnosed according to current imaging diagnostic criteria.