Hepatic Perfusion Disorder Associated with Focal Liver Lesions: Contrast-Enhanced US Patterns--correlation Study with Contrast-Enhanced CT.
Xiang Zhou,Yan Luo,Yu-Lan Peng,Wei Cai,Qiang Lu,Ling Lin,Xiao-Xi Sha,Yong-Zhong Li,Meng Zhu
DOI: https://doi.org/10.1148/radiol.11101454
IF: 19.7
2011-01-01
Radiology
Abstract:PURPOSE:To retrospectively compare the detection and characterization of hepatic perfusion disorder (HPD) associated with focal liver lesions (FLLs) at contrast material-enhanced ultrasonography (US) by using contrast-enhanced computed tomography (CT) as the reference standard.MATERIALS AND METHODS:The study was approved by the local institutional ethics committee, and informed consent was waived. Three hundred fifty patients (mean age, 50 years ± 11 [standard deviation]; age range, 19-82 years; 168 women, 182 men) underwent contrast-enhanced US and contrast-enhanced CT between April 2008 and July 2010. Two independent readers reviewed contrast-enhanced US images for the detection and characterization of HPD. The largest lesion or the lesion best identified at contrast-enhanced US per patient was used for statistical analysis. Contrast-enhanced CT was used as the reference standard. Contrast-enhanced US and CT interreader agreement of diagnoses was assessed by using the weighted κ coefficient, and influences of lesion size, enhancement covering rate, and liver cirrhosis were evaluated by using logistic regression analysis and the paired χ(2) test. Sensitivity, specificity, positive and negative predictive values, and accuracy of contrast-enhanced US for HPD detection were calculated.RESULTS:Contrast-enhanced US results showed HPD features similar to those of CT imaging. CT depicted 50 HPDs in 350 patients, and contrast-enhanced US depicted 55 HPDs in 350 patients. The agreement for HPD diagnosis between US and CT was good (κ = 0.749). Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of contrast-enhanced US were 84.0%, 95.7%, 76.4%, 97.3%, and 0.945, respectively. Rapid enhancement coverage (P < .001) and lesion size (P = .002) were significant predictors of the occurrence of HPD. Liver cirrhosis did not have significant influence for HPD detection (P = .087). Image zooming, limited acoustic window, lesion diameter greater than 5 cm, attenuation, and blurred images were the main reasons for the false-positive diagnosis of HPD at contrast-enhanced US.CONCLUSION:The HPD in FLLs can reliably be detected with contrast-enhanced US, which correlated well with contrast-enhanced CT images.