AUTHOR=Han Zewen , Dai Hanting , Chen Xiaolin , Gao Lanmei , Chen Xiaojie , Yan Chuan , Ye Rongping , Li Yueming 

TITLE=Delta-radiomics models based on multi-phase contrast-enhanced magnetic resonance imaging can preoperatively predict glypican-3-positive hepatocellular carcinoma

JOURNAL=Frontiers in Physiology

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2023.1138239

DOI=10.3389/fphys.2023.1138239

ISSN=1664-042X

ABSTRACT=Objectives To investigate the value of multi-phase contrast-enhanced magnetic resonance imaging (CE-MRI) based delta radiomics for identifying the glypican-3 (GPC3)-positive hepatocellular carcinoma (HCC).One hundred and twenty-six patients with pathologically-confirmed HCC (training cohort: n = 88; validation cohort: n = 38) were retrospectively recruited. Basic information was obtained from medical records. Preoperative multi-phase CE-MRI images were reviewed and 3D volume of interest (VOI) of the whole tumor was delineated on non-contrast T1 weighted imaging (T1), arterial phase (AP), portal venous phase (PVP), delayed phase (DP) and hepatobiliary phase (HBP). One hundred and seven original radiomic features were extracted from each phase and delta radiomic features were calculated. After a two-step feature selection strategy, radiomics models were built using two classification algorithms. A nomogram was constructed by combining the best radiomics model and clinical risk factors. Results Serum alpha-fetoprotein (AFP) (p = 0.013) was significantly related to GPC3-positive HCC. The optimal radiomics model composed of 8 delta radiomics features with the AUC 0.805 and 0.857 in the training and validation cohort respectively. The nomogram integrated the radiomics score and AFP performed excellently (training cohort: AUC = 0.844; validation cohort: AUC = 0.862). Calibration curve showed good agreement between the nomogram-predicted probabilities and GPC3 actual expression in both training and validation cohort. Decision curve analysis further demonstrates the clinical practicality of the nomogram. Conclusion Multi-phase CE-MRI based delta radiomics model can non-invasively predict GPC3-positive HCC and could be a useful method for individualized diagnosis and treatment.