Quantitative Assessment of the Accuracy of Real-Time Virtual Sonography for Liver Surgery

Background. Real-time virtual sonography (RVS) is a navigation system for liver surgery. In this study, the degree of misalignment of intraoperative RVS images with computed tomographic (CT) images was measured. Methods. Between December 2014 and July 2015, intraoperative RVS was performed in a total of 33 patients undergoing liver surgery. Reconstructed CT images, rendered like intraoperative ultrasonographic (IOUS) images, were adjusted with the IOUS images and visualized side by side. The degree of misalignment between the reconstructed CT images and IOUS images was measured at anterior section, posterior section, and left liver in each patient. Furthermore, the time required for the adjustment was measured as the "adjustment time." Results. The degree of misalignment between the images could potentially be measured for a total of 96 points in the 33 patients. Of these, the actual measurement could not be conducted for 35 points due to poor visualization of the intrahepatic vasculature (n = 20) or to a large misalignment that hampered continuation of further adjustment (n = 15). The median degree of misalignment was 9.8 mm (range = 2.4-37.6 mm) in the right anterior section, 9.8 mm (range = 2.7-71.5 mm) in the right posterior section, and 9.5 mm (range = 0.9-37.6 mm) in the left liver. The median adjustment time was 105 seconds (range = 51-245 seconds). Conclusions. Although some misalignment occurred, it might be acceptable for selected situations. Further investigation is needed to reduce the frequency of adjustment failure.