Value of an ultrasound radiomics model in predicting CK19 protein expression in hepatocellular carcinoma

XU Meiyuan; GAO Ruizhi; YANG Hong; HE Yun

doi:10.19300/j.2025.L22071

Abstract：Objective To explore the value of an ultrasound radiomics model in preoperatively predicting CK19 protein expression in patients with hepatocellular carcinoma(HCC).Methods A total of 156 patients with pathologically confirmed HCC who had undergone CK19 immunohistochemical staining were retrospectively enrolled.Patients were randomly divided into a training set(n=110)and a validation set(n=46)in a 7∶3 ratio.Clinical data between the two groups were compared using the Chi-square test.A total of 5,936 radiomic features were extracted from grayscale ultrasound images.After Z-score normalization,feature selection and dimensionality reduction were performed in the training set using hypothesis testing,least absolute shrinkage and selection operator(LASSO)regression,and correlation analysis.A multivariate logistic regression model was then constructed.The model's predictive performance for CK19-positivecases was evaluated using the area under the receiver operating characteristic curve(AUC),sensitivity,specificity,and accuracy.Clinical net benefit was assessed using decision curve analysis.Results There were no statistically significant differences in general clinical characteristics between the training and validation sets(all P＞0.05).A total of 13 radiomic features were selected to construct the radiomics model.The model demonstrated good predictiveperformance in both the training and validation sets,with AUCs of 0.936(sensitivity 87.5%,specificity 89.4%,accuracy 90.0%)and 0.881(sensitivity 100%,specificity 81.0%,accuracy 84.8%),respectively.Decision curve analysis indicated that the model had good clinical utility.Conclusion The ultrasound radiomics model shows good predictive value for CK19 expression in HCC patients and may provide valuable information for clinical practice.