Liu D, Liu F, Xie X, et al. - Researchers retrospectively involved 130 hepatocellular carcinomas (HCC) people (89 for training, 41 for validation), who underwent ultrasound examination contrast-enhanced ultrasound (CEUS and B-mode) within 1 week prior the first transarterial chemoembolization (TACE) session in order to establish and validate an artificial intelligence-based radiomics strategy for prognosticating personalized responses of HCC to first TACE session by quantitatively analyzing CEUS cines. Ultrasonographic data was utilized for building and approving deep learning radiomics-based CEUS model (R-DLCEUS), machine learning radiomics-based time-intensity curve of CEUS model (R-TIC), and machine learning radiomics-based B-Mode images model (R-BMode), respectively, to prognosticate response to TACE with reference to transformed response evaluation criteria in solid tumor. In the validation cohort the area under the receiver operating characteristic curve of R-DLCEUS, R-TIC, and R-BMode were 0.93, 0.80, and 0.81, respectively. Findings suggest that the DL-based radiomics method could efficiently use CEUS cines to achieve perfect and personalized prognostication. Moreover, it was easy to operate and holds good potential for profiting TACE applicants in clinical practice.