Manual and semi-automated computed tomography volumetry significantly overestimates the right liver lobe graft weight: a single-center study with adult living liver donors.

Abstract

Preoperative evaluation of donor liver volume is indispensable in living donor liver transplantation to ensure sufficient residual liver and graft-to-recipient weight ratio. This study aims to evaluate the accuracy of two computed tomography (CT) volumetry programs, an interactive manual and a semi-automated one, in the preoperative estimation of the right lobe graft weight. One hundred and nine right liver lobe living donors between January 2008 and January 2020 were enrolled in this retrospective study. Two radiologists measured the liver graft volumes independently using manual and semi-automated CT volumetry, and the interaction time was recorded. Actual graft weight (AGW) measured intraoperatively served as the reference standard. The paired samples t-test was used to compare the estimated graft weight (EGW) and the AGW. Inter-user and inter-method agreements were assessed with Bland-Altman plots. Both manual and semi-automated CT volumetry significantly overestimated the graft weight (EGW manual: 893 ± 155 mL vs. AGW manual: 787 ± 128 g, P < 0.001, EGW semi-automated: 879 ± 143 mL vs. AGW semi-automated, P < 0.001). The junior radiologist measured higher volumes than the senior radiologist with either method (P < 0.001). The Bland-Altman analysis revealed mean difference and standard deviation for inter-method agreement of 7 ± 48 cc for the senior radiologist, and 34 ± 54 cc for the junior radiologist. The mean difference and standard deviation for inter-method agreement was 63 ± 59 cc in manual volumetry and 22 ± 38 cc in semi-automated volumetry. The mean interaction time was 27.3 ± 14.2 min for manual volumetry and 6.8 ± 1.4 min for semi-automated volumetry (P < 0.001). Both manual and semi-automated CT volumetry significantly overestimated the right liver graft weight, while semi-automated volumetry significantly reduced the interaction time.