The usefulness of 2B-mode feature intelligent model in grading of liver fibrosis in patients with chronic hepatitis B
10.3760/cma.j.issn.1008-1372.2018.01.013
- VernacularTitle:二维超声图像特征智能模型对慢性乙型肝炎肝纤维化分级的作用
- Author:
Yongfang LUO
1
;
Changfeng DONG
;
Can HUANG
Author Information
1. 518112,广东省深圳市第三人民医院超声科
- Keywords:
Elasticity imaging techniques;
Hepatitis B,chronic/CO/US;
Liver cirrhosis/CO/US
- From:
Journal of Chinese Physician
2018;20(1):50-53,59
- CountryChina
- Language:Chinese
-
Abstract:
Objective To discuss the diagnostic accuracy of 2B-mode feature model in grading the degree of hepatic fibrosis compared to acoustic radiation force impulse (ARFI) and liver biopsy.Methods A total of 140 patients was enrolled in the study and divided into four groups (F1-F4) according to pathological grading using METAVIR scores system,and 47 healthy volunteers were enrolled as the control group (F0) at random.All of subjects underwent standard ultrasound examination and acoustic radiation force impulse (ARFI).Ultrasound raw images were obtained and analyzed with 2B-mode feature intelligent model and then compared to the value of liver stiffness (ARFI).Results The area under the receiver operating characteristic (ROC) curve of grading of liver fibrosis (F2) using 2B-mode feature intelligent model was training =0.973 2,and testing =0.751 1,which was superior to the area under the ROC curve (F2) using ARFI with training =0.840 1,and testing =0.656 4.Conclusions 2B-mode feature intelligent model could be used for grading of liver fibrosis in patients with chronic hepatitis B (CHB).There is great potential in the quantitative diagnosis of liver fibrosis stage using 2B-mode ultrasound.
