Allard Marc-Antoine, Baillié Gaëlle, Castro-Benitez Carlos, Faron Matthieu, Blandin Frédérique, Cherqui Daniel, Castaing Denis, Cunha Antonio Sa, Adam René, Vibert Éric
Centre Hépato-Biliaire, Paul Brousse Hospital, AP-HP, Villejuif, F-94800, France; University of Paris-Sud, Villejuif, F-94800, France; INSERM, Unit UMRS776, Villejuif, F-94800, France.
Centre Hépato-Biliaire, Paul Brousse Hospital, AP-HP, Villejuif, F-94800, France.
HPB (Oxford). 2017 Apr;19(4):338-344. doi: 10.1016/j.hpb.2016.11.012. Epub 2016 Dec 30.
The performance of linear models predicting Total Liver Weight (TLW) remains moderate. The use of more complex models such as Artificial Neural Network (ANN) and Generalized Additive Model (GAM) or including the variable "steatosis" may improve TLW prediction. This study aimed to assess the value of ANN and GAM and the influence of steatosis for predicting TLW.
Basic clinical and morphological variables of 1560 cadaveric donors for liver transplantation were randomly split into a training (2/3) and validation set (1/3). Linear models, ANN and GAM were built by using the training cohort and evaluated with the validation cohort.
The TLW is subject to major variations among donors with similar morphological parameters. The performance of ANN and GAM were moderate and similar to that of linear models (concordance coefficient from 0.36 to 0.44). In 28-30% of cases, TLW cannot be predicted with a margin of error ≤20%. The addition of the variable "steatosis" to each model did not improve their performance.
TLW prediction based on anthropological parameters carry a significant risk of error despite the use of more complex models. Others determinants of TLW need to be identified and imaging-based volumetric measurements should be preferred when feasible.
预测全肝重量（TLW）的线性模型表现一般。使用更复杂的模型，如人工神经网络（ANN）和广义相加模型（GAM），或纳入“脂肪变性”变量，可能会改善对TLW的预测。本研究旨在评估ANN和GAM的价值以及脂肪变性对预测TLW的影响。
将1560例肝移植尸体供者的基本临床和形态学变量随机分为训练集（2/3）和验证集（1/3）。使用训练队列构建线性模型、ANN和GAM，并使用验证队列进行评估。
在形态学参数相似的供者中，TLW存在较大差异。ANN和GAM的表现一般，与线性模型相似（一致性系数为0.36至0.44）。在28% - 30%的病例中，无法在误差范围≤20%的情况下预测TLW。在每个模型中加入“脂肪变性”变量并未改善其性能。
尽管使用了更复杂的模型，但基于人类学参数预测TLW仍存在显著误差风险。需要确定TLW的其他决定因素，可行时应优先采用基于影像学的体积测量方法。
