Cited 49 times since 2010 (3.4 per year) source: EuropePMC World journal of surgery, Volume 34, Issue 10, 1 1 2010, Pages 2426-2433 Virtual liver resection and volumetric analysis of the future liver remnant using open source image processing software. van der Vorst JR, van Dam RM, van Stiphout RS, van den Broek MA, Hollander IH, Kessels AG, Dejong CH
Background
After extended liver resection, a remnant liver that is too small can lead to postresection liver failure. To reduce this risk, preoperative evaluation of the future liver remnant volume (FLRV) is critical. The open-source OsiriX PAC software system can be downloaded for free and used by nonradiologists to calculate liver volume using a stand-alone Apple computer. The purpose of this study was to assess the accuracy of OsiriX CT volumetry for predicting liver resection volume and FLVR in patients undergoing partial hepatectomy.
Methods
Preoperative contrast-enhanced liver CT scans of patients who underwent partial hepatectomy were analyzed by three observers. Two surgical trainees measured the total liver volume, resection volume, and tumor volume using OsiriX, and a radiologist measured these volumes using CT scanner-linked Aquarius iNtuition software. Resection volume was correlated with prospectively determined resection weight, and differences in the measured liver volumes were analyzed. Interobserver variability was assessed using Bland-Altman plots.
Results
25 patients (M/F ratio: 13/12) with a median age of 61 (range, 34-77) years were included. There were significant correlations between the weight and volume of the resected specimens (Pearson's correlation coefficient: R(2) = 0.95). There were no major differences in total liver volumes, resection volumes, or tumor volumes for observers 1, 2, and 3. Bland-Altman plots showed a small interobserver variability. The mean time to complete liver volumetry for one patient using OsiriX was 19 +/- 3 min.
Conclusions
OsiriX liver volumetry performed by surgeons is an accurate and time-efficient method for predicting resection volume and FLRV.
