| Related Articles |
New noninvasive ultrasound techniques: can they predict liver cirrhosis?
Ultrasound Q. 2012 Mar;28(1):5-11
Authors: Yu H, Wilson SR
Abstract
OBJECTIVE: This study aimed to determine the threshold values and accuracy of 2 noninvasive techniques, contrast-enhanced ultrasound with maximum-intensity projection (MIP) imaging and ultrasound elastography, acoustic radiation force impulse (ARFI), to differentiate a normal from a cirrhotic liver.
MATERIALS AND METHODS: One hundred thirty-two patients were predicted clinically as having a normal liver (n = 60) or cirrhosis (n = 72). All had MIP liver vessel evaluation on an Acuson Sequoia and 90 of them had ARFI on an S2000 (Siemens, Mountain View, Calif). Two readers reviewed 4 parameters on MIP data and predicted cirrhosis (n = 65) or normal (n = 67) outcome. They were considered as having cirrhosis when more than 1 MIP parameter was positive. Acoustic radiation force impulse values above 1.36 m/s suggested cirrhosis. Receiver operating characteristic curve analysis was performed to evaluate the diagnostic accuracy of the ARFI measurement and to extract the optimal cutoff value in the differentiation of a cirrhotic liver from a noncirrhotic liver.
RESULTS: Sixty-four patients (64/65, 98.5%) with abnormal and 8 (8/59, 13.6%) with normal vessels on MIP imaging were clinically cirrhosis (P < 0.001). Forty-five patients (45/49, 91.8%) with abnormal ARFI (? 1.36 m/s) and 6 (6/41, 14.6%) with normal ARFI (<1.36 m/s) had a clinical cirrhotic liver (P < 0.001). Sensitivity and specificity of MIP of vessel morphology to predict cirrhosis were 89% (64/72) and 98% (59/60), respectively. Acoustic radiation force impulse showed a sensitivity of 88% (45/51) and a specificity of 90% (35/39), and the area under the receiver operating characteristic curve was 0.932, suggesting very good accuracy. The combination of ARFI and MIP showed increased sensitivity and specificity to 95.8% and 100%.
CONCLUSIONS: Liver evaluation with MIP and ARFI contributes to a noninvasive prediction of cirrhosis.
PMID: 22357223 [PubMed - indexed for MEDLINE]