The present study by Liu et al. explores a potential methodology using clinical data, radiomics data, and images produced from deep-learning to develop a highly accurate model of distant metastases post-transcatheter arterial chemoembolization (post-TACE). Further improvements to the effectiveness of TACE treatment is needed to fully treat hepatocellular carcinomas (HCC). Approximately 500 patients undergoing TACE treatment were set into three groups (training, testing, and external validation or control) and were exposed to the following treatment modalities: a clinical model with logistic regression using clinical basis, a radiomics model with an extraction sample of quantitative imaging features produced based on pre-TACE treatment scans, and a deep-learning model programmed to automatically pull features for its images. Two additional models were the combination of deep learning and radiomics to create the DLR condition to fuse the features of both, and the overarching combination by weaving DLR-procured features with clinical parameters via logistic regression. Results noted that the combined modality produced the highest rating in tissue discrimination and excellence in identifying clinical features of the carcinomas. Thus, there is potential for clinical applications of this model in identification and managing of pre- and post-TASE conditions in patients.