Three-Dimensional Texture Analysis Based on PET/CT Images to Distinguish Hepatocellular Carcinoma and Hepatic Lymphoma.

ABSTRACT: Objectives: This study compared the diagnostic ability of image-based parameters with texture parameters in the differentiation of hepatocellular carcinoma (HCC) and hepatic lymphoma (HL) by positron emission tomography-computed tomography (PET/CT). Methods: Patients with pathological diagnosis of HCC and HL were included in this study. Image-based and texture parameters were obtained by manual drawing of region of interest. Receiver operating characteristic (ROC) was used to test the diagnostic capacity of each parameter. Binary logistic regression was used to transform the most discriminative image-based parameters, texture parameters, and the combination of these parameters into three regression models. ROC was used to test the diagnostic capacity of these models. Result: Ninety-nine pa

SUBMITTER: Xu H 

PROVIDER: S-EPMC6733884 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature