A Radiomics Nomogram Based On MRI For Preoperative Prediction Of Early Recurrence Of Small Hepatocellular Carcinoma After Surgical Resection Or Radiofrequency Ablation

Objective: The aim of this study was to develop and validate a radiomics nomogram based on magnetic resonance images(MRI)to preoperatively predict early recurrence(ER)(≤2 years)of small hepatocellular carcinoma(HCC).Materials and Methods: The study population included 137 patients with small HCC who underwent surgical resection(SR)or radiofrequency ablation(RFA)between January 2015 and November 2018,and they were divided into a training set(n = 111)and a validation set(n = 26).Radiomic features were extracted from the entire tumor on MRI by using the Ma Zda software.The least absolute shrinkage and selection operator(Lasso)method was used for data dimensionality reduction,feature selection,and radiomics signature construction.And a radiomics score(Rad-score)was calculated.Univariate and multivariate logistic regression analyses were then used to search for potential independent risk factors,including clinical factors,radiologic features,and Rad-score,and a combined model was established.Furthermore,a radiomics nomogram based on the combined model was established to provide a tool for clinicians to quantitatively predict the probability of ER in patients with small HCC.The area under the curve(AUC)of the prediction model in the training set and the validation set was further analyzed to evaluate the prediction effectiveness of the radiomics model.Results: A total of 62(62/137,44.9%)patients had confirmed ER according to the final clinical outcomes.In the training group,univariate logistic regression analysis showed that cirrhosis and hepatitis B infection,hepatobiliary phase hypointensity,Child-Pugh score,the platelet count and Rad-score were correlated with ER.However,after multivariate logistic regression analysis,only the platelet count and Rad-score were included in the combined model as risk factors for predicting ER of small HCC.The radiomics nomogram based on the combined model showed good pracdiction in the training set(AUC,0.981;95% CI,0.957–1.000)and the validation set(AUC,0.784;95%CI,0.557-1.000).Conclusion: The MRI-based radiomics nomogram,a noninvasive preoperative prediction tool that incorporates the Rad-score and clinical risk factors,shows favorable predictive efficacy to preoperatively predict individual probability of ER of small HCC.As a non-invasive and quantitative method,radiomics nomogram can be used as an effective tool in the clinical decision-making process.