AUTHOR=Zhao Xitong , Liang Pan , Yong Liuliang , Jia Yan , Gao Jianbo 

TITLE=Radiomics Study for Differentiating Focal Hepatic Lesions Based on Unenhanced CT Images

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.650797

DOI=10.3389/fonc.2022.650797

ISSN=2234-943X

ABSTRACT=Objectives：To investigate the feasibility of computer-aided discriminative diagnosis among hepatocellular carcinoma(HCC), hepatic metastasis, hepatic hemangioma, hepatic cysts, hepatic adenoma, and hepatic focal nodular hyperplasia, based on radiomics analysis of unenhanced CT images.
Methods: 452 patients were included in current study with 77 with HCC, 104 with hepatic metastases, 126 with hepatic hemangioma, 99 with hepatic cysts, 24 with FNH, 22 with HA. Radcloud Platform was used to extract radiomics features from manual delineation on unenhanced CT images. Most relevant radiomic features were selected from 1409 via LASSO (least absolute shrinkage and selection operator). The whole dataset were divided into training and validation dataset with the ratio of 8:2. Support Vector Machine (SVM) was used to establish the classifier.
Results: The computer-aided diagnosis model was established based on radiomic features of unenhanced CT images. 27 most relevant radiomic features were selected to distinguish the six different histopathological types of all lesions. The classifiers had good diagnostic performance, with area under the curve(AUC) valuse greater than 0.900 in training and validation groups. The diagnostic accuracy of the training group and the verification group about differentiating the six different histopathological types of all lesions was 0.88 and 0.76 respectively. 34 most relevant radiomic features were selected to distinguish the benign and malignant tumours. The total discriminant accuracy in training groups was 0.89, with 0.84 in validation groups.
Conclusions: The computer-aided discriminative diagnosis model based on unenhanced CT images has good clinical potential in distinguishing focal hepatic lesions with noninvasive radiomic features.