Improved Detection of Hepatic Metastases with Contrast-enhanced Low Mechanical-index Pulse Inversion Ultrasonography During the Liver-specific Phase of Sonazoid: Observer Performance Study with JAFROC Analysis

Sugimoto, Katsutoshi; Shiraishi, Junji; Moriyasu, Fuminori; Saito, Kazuhiro; Doi, Kunio

doi:10.1016/j.acra.2008.12.025

« Previous Next »Academic Radiology
Volume 16, Issue 7 , Pages 798-809, July 2009

Improved Detection of Hepatic Metastases with Contrast-enhanced Low Mechanical-index Pulse Inversion Ultrasonography During the Liver-specific Phase of Sonazoid:

Observer Performance Study with JAFROC Analysis

Kurt Rossmann Laboratories for Radiologic Imaging Research, Department of Radiology, The University of Chicago, 5841 S. Maryland Ave., MC 2026, Chicago, IL 60637 (K. Sugimoto, J.S., K.D.); Department of Gastroenterology and Hepatology, (K. Sugimoto, F.M.), and Department of Radiology (K. Saito), Tokyo Medical University, Tokyo, Japan

Received 18 November 2008; accepted 24 December 2008. published online 27 April 2009.

Rationale and Objectives

To compare B-mode ultrasonography (US) alone with the combination of B-mode and contrast-enhanced (Sonazoid) late-phase pulse-inversion US for the detection of hepatic metastases by use of jackknife free-response receiver-operating characteristic (JAFROC) analysis.

Materials and Methods

Twenty-seven patients with 57 hepatic metastases and 6 patients without hepatic metastases underwent B-mode and contrast-enhanced US. We used the diagnoses established by contrast-enhanced computed tomography and contrast-enhanced US as the standard of reference. All ultrasonographic scanning was performed by an experienced radiologist with a routine clinical procedure. All scanning data were archived with digital cine clips. A review system, which can display pairs of cine clips for B-mode and contrast-enhanced US side by side, was developed for off-site observer study. Seven radiologists interpreted each case individually first by B-mode US only, and then by the combination with contrast-enhanced US by identifying locations of possible candidates for hepatic metastasis with their confidence ratings. The figure-of-merit (FOM) values, sensitivity, and false-positives per case were estimated for B-mode US alone, and for the combination of B-mode and contrast-enhanced US.

Results

The sensitivities of the combined ultrasonographic imaging (mean, 72.2%) were clearly improved from that of B-mode US alone (mean, 41.6%) while reducing the average number of false positives from 1.1 to 0.5 per case. In the jackknife analysis, there was a statistically significant difference between mean FOM values for the combined imaging (0.76) and for B-mode US alone (0.44, P < .00001).