医学
回顾性队列研究
自动化方法
肝移植
可靠性(半导体)
放射科
外科
核医学
移植
人工智能
计算机科学
功率(物理)
物理
量子力学
作者
Eun Chang Choi,Seok‐Hwan Kim
标识
DOI:10.1097/js9.0000000000003112
摘要
Background: Accurate preoperative graft volume assessment is fundamental to the success of living donor liver transplantation (LDLT). Although manual and automated computed tomography (CT) volume measurement methods using various volumetric tools are widely used, their accuracy remains uncertain. This study aimed to compare various CT-based volumetric measurement methods for predicting actual graft weight (AGW) in LDLT and to identify specific dry weight correction factors for each method to improve clinical reliability. Materials and Methods: A retrospective diagnostic accuracy study was performed on 109 patients who underwent LDLT between 2011 and 2024. Right liver volume was measured using automated (Philips Healthcare software), semi-automated (AnyVol software), and manual volumetry (PetaVision for clinics) methods. The optimal dry weight correction factor was calculated for each method. Results: The optimal dry weight correction factors were 0.89 for the automated method, 0.82 for the semi-automated method, and 0.88 for the manual method. After applying these correction factors, the semi-automated method yielded the highest coefficient of determination (R 2 = 0.687, standard error = 91.939). The error ratio decreased significantly: from 11.30 ± 14.59% to −0.93 ± 12.98% for the automated method, from 20.51 ± 15.65% to −1.18 ± 12.83% for the semi-automated method, and from 11.89 ± 14.67% to −1.53 ± 12.91% for the manual method. Conclusions: Accurate prediction of AGW depends on applying optimal correction factors specific to each measurement method. All three methods showed high accuracy with the semi-automated method demonstrating the highest R 2 and lowest SE, while the automated method exhibited the lowest error ratio. These findings support the use of cost-effective, software-based volumetry with tailored correction factors to improve donor safety and graft outcomes.
科研通智能强力驱动
Strongly Powered by AbleSci AI