What Imaging Modality Is More Effective in Predicting Early Recurrence of Hepatocellular Carcinoma after Hepatectomy Using Radiomics Analysis: CT or MRI or Both?

Abstract

Background: It is of great importance to predict the early recurrence (ER) of hepatocellular carcinoma (HCC) after hepatectomy using preoperative imaging modalities. Nevertheless, no comparative studies have been conducted to determine which modality, CT or MRI with radiomics analysis, is more effective.

Methods: We retrospectively enrolled 119 HCC patients who underwent preoperative CT and MRI. A total of 3776 CT features and 4720 MRI features were extracted from the whole tumor. The minimum redundancy and maximum relevance algorithm (MRMR) and least absolute shrinkage and selection operator (LASSO) regression were applied for feature selection, then support vector machines (SVMs) were applied for model construction. Multivariable logistic regression analysis was employed to construct combined models that integrate clinical-radiological-pathological (CRP) traits and radscore. Receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) were used to compare the efficacy of CT, MRI, and CT and MRI models in the test cohort.

Results: The CT model and MRI model showed no significant difference in the prediction of ER in HCC patients (p = 0.911). Radiomics^CT&MRI demonstrated a superior predictive performance than either Radiomics^CT or Radiomics^MRI alone (p = 0.032, 0.039). The combined CT and MRI model can significantly stratify patients at high risk of ER (area under the curve (AUC) of 0.951 in the training set and 0.955 in the test set) than the CT model (AUC of 0.894 and 0.784) and the MRI model (AUC of 0.856 and 0.787). DCA demonstrated that the CT and MRI model provided a greater net benefit than the models without radiomics analysis.

Conclusions: No significant difference was found in predicting the ER of HCC between CT models and MRI models. However, the multimodal radiomics model derived from CT and MRI can significantly improve the prediction of ER in HCC patients after resection.

Keywords: early recurrence; hepatocellular carcinoma; prediction; radiomics.

Figure 1

Flowchart of patient selection.

Figure 2

The workflow to conduct radiomics analysis in our study. Tumors were contoured manually and adjusted slice by slice on five MRI sequences and four CT phases. VOI of the whole tumor was segmented, and radiomics features were separately extracted. Maximum relevance minimum redundancy (MRMR) and least absolute shrinkage and selection operator (LASSO) logistic regression were used for feature selection, and SVM was used for radiomics model construction. Receiver operating characteristic curve (ROC), calibration, and decision curve analysis (DCA) were applied for model performance assessment.

Figure 3

Discrepant MRI and CT findings in HCCs. (a) HCC patients (recurrence within 18 months after hepatectomy), 60 years, a 1.9 cm HCC; enhancement in the pseudocapsule was observed in the AP, and delayed phase on MRI while the appearance was not observed on CT. (b) HCC patients (ER group), male, 52 years old, 0.6–0.8 cm HCC lesions with wash-in in the AP (red arrow) and partial wash-out in the PVP at MRI; these lesion did not show wash-in and wash-out at CT. PVP = portal venous phase; AP = arterial phase.

Figure 4

Comparison of ROC curves for single radiomics models vs. clinical–radiological–pathological (CRP) model and for combined models vs. CRP model in the training cohort (a,b) and the test cohort (c,d).

Figure 5

CRP nomogram (a) and combined CT and MRI nomogram (d). Calibration curves for the CRP nomogram (b,c) and combined CT and MRI nomogram (e,f) for the training cohort and test cohort. On the y-axis is the actual outcome of early recurrence of HCC, and on the x-axis is the predicted outcome. The perfect performance of the ideal model is represented by the blue lines. The red lines represent the performance of the model, of which a closer fit to the blue line indicates better prediction performance.

Figure 6

Decision curve analysis (DCA) of the clinical usefulness assessment of the Radiomics^CT&MRI model, CRP model, and combined CT and MRI model in the training cohort (a) and the test cohort (b). The y-axis represents the net benefit, and the x-axis represents the threshold probability. The combined CT and MRI (blue line) achieves more net benefit than the Radiomics^CT&MRI model (green line), CRP model (red line), treat-all strategy (gray line), and treat-none strategy (horizontal black line) across the majority of the range of threshold probabilities.