Utility of Longitudinal Measurement of the Liver With Ultrasound in Comparison to Computed Tomography Liver Volume in Assessing Hepatomegaly.

ABSTRACT: The purpose of the study is to retrospectively compare the longitudinal measurement of the liver on ultrasound (US) with liver volume as measured on computed tomography (CT). This institutional review board-approved retrospective study with prospective image analysis included adult patients with US examinations of the liver and CT examinations of the abdomen performed within 2 weeks of each other from January 1, 2010, to April 30, 2016. We recorded the dome-to-tip longitudinal length measurement of the right lobe of the liver on the US study. The liver volume was calculated using manual 3-dimensional segmentation of the CT data set. Additional linear liver measurements and qualitative assessments were retrospectively made on the US and CT images. Hepatomegaly was defined as a liver volume 2000 mL or greater. The sample had 302 patients including 140 males and 162 females. Liver length measured on US had a moderate positive correlation with liver volume measured on CT, with a Pearson correlation coefficient of 0.7078. The optimum cutoff of liver length (US) was found to be 17 cm by Youden index, with a sensitivity of 76.0% and specificity of 73.4% for hepatomegaly. Additional linear measurements and qualitative assessments of the liver did not significantly improve the diagnosis of hepatomegaly. In conclusion, liver length measured with US has only a moderate correlation with liver volume. Commonly used US liver length maximum values of 16 to 18 cm produce a significant number of false-negatives and false-positives for the diagnosis of hepatomegaly.