Application and challenge of radiomics technique in the era of precision medicine for hepatobiliary disease

Abstract

Radiomics, as an emerging technique of omics, shows the pathophysiological information of images via extracting innumerable quantitative features from digital medical images. In recent years, it has been an exponential increase in the number of radiomics studies. The applications of radiomics in hepatobiliary diseases at present include: assessment of liver fibrosis, discrimination of malignant from benign tumors, prediction of biological behavior, assessment of therapeutic response, and prognosis. Integrating radiomics analysis with machine learning algorithms has emerged as a non-invasive method for predicting liver fibrosis stages, microvascular invasion and post-resection recurrence in liver cancers, lymph node metastasis in biliary tract cancers as well as treatment response in colorectal liver metastasis, with high performance. Although the challenges remain in the clinical transformation of this technique, radiomics will have a broad application prospect in promoting the precision diagnosis and treatment of hepatobiliary diseases, backed by multi-center study with large sample size or multi-omics study.