scholarly journals Artificial Intelligence—A Tool for Risk Assessment of Delayed-Graft Function in Kidney Transplant

2021 ◽  
Vol 10 (22) ◽  
pp. 5244
Author(s):  
Andrzej Konieczny ◽  
Jakub Stojanowski ◽  
Klaudia Rydzyńska ◽  
Mariusz Kusztal ◽  
Magdalena Krajewska

Delayed-graft function (DGF) might be responsible for shorter graft survival. Therefore, a clinical tool predicting its occurrence is vital for the risk assessment of transplant outcomes. In a single-center study, we conducted data mining and machine learning experiments, resulting in DGF predictive models based on random forest classifiers (RF) and an artificial neural network called multi-layer perceptron (MLP). All designed models had four common input parameters, determining the best accuracy and discriminant ability: donor’s eGFR, recipient’s BMI, donor’s BMI, and recipient–donor weight difference. RF and MLP designs, using these parameters, achieved an accuracy of 84.38% and an area under curve (AUC) 0.84. The model additionally implementing a donor’s age, gender, and Kidney Donor Profile Index (KDPI) accomplished an accuracy of 93.75% and an AUC of 0.91. The other configuration with the estimated post-transplant survival (EPTS) and the kidney donor risk profile (KDRI) achieved an accuracy of 93.75% and an AUC of 0.92. Using machine learning, we were able to assess the risk of DGF in recipients after kidney transplant from a deceased donor. Our solution is scalable and can be improved during subsequent transplants. Based on the new data, the models can achieve better outcomes.

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Satoru Kawakita ◽  
Jennifer L. Beaumont ◽  
Vadim Jucaud ◽  
Matthew J. Everly

Abstract Machine learning (ML) has shown its potential to improve patient care over the last decade. In organ transplantation, delayed graft function (DGF) remains a major concern in deceased donor kidney transplantation (DDKT). To this end, we harnessed ML to build personalized prognostic models to predict DGF. Registry data were obtained on adult DDKT recipients for model development (n = 55,044) and validation (n = 6176). Incidence rates of DGF were 25.1% and 26.3% for the development and validation sets, respectively. Twenty-six predictors were identified via recursive feature elimination with random forest. Five widely-used ML algorithms—logistic regression (LR), elastic net, random forest, artificial neural network (ANN), and extreme gradient boosting (XGB) were trained and compared with a baseline LR model fitted with previously identified risk factors. The new ML models, particularly ANN with the area under the receiver operating characteristic curve (ROC-AUC) of 0.732 and XGB with ROC-AUC of 0.735, exhibited superior performance to the baseline model (ROC-AUC = 0.705). This study demonstrates the use of ML as a viable strategy to enable personalized risk quantification for medical applications. If successfully implemented, our models may aid in both risk quantification for DGF prevention clinical trials and personalized clinical decision making.


2019 ◽  
Vol 31 (1) ◽  
pp. 175-185 ◽  
Author(s):  
Sunjae Bae ◽  
Jacqueline M. Garonzik Wang ◽  
Allan B. Massie ◽  
Kyle R. Jackson ◽  
Mara A. McAdams-DeMarco ◽  
...  

BackgroundEarly steroid withdrawal (ESW) is associated with acceptable outcomes in kidney transplant (KT) recipients. Recipients with delayed graft function (DGF), however, often have a suboptimal allograft milieu, which may alter the risk/benefit equation for ESW. This may contribute to varying practices across transplant centers.MethodsUsing the Scientific Registry of Transplant Recipients, we studied 110,019 adult deceased-donor KT recipients between 2005 and 2017. We characterized the association of DGF with the use of ESW versus continued steroid maintenance across KT centers, and quantified the association of ESW with acute rejection, graft failure, and mortality using multivariable logistic and Cox regression with DGF-ESW interaction terms.ResultsOverall 29.2% of KT recipients underwent ESW. Recipients with DGF had lower odds of ESW (aOR=0.600.670.75). The strength of this association varied across 261 KT centers, with center-specific aOR of <0.5 at 31 (11.9%) and >1.0 at 22 (8.4%) centers. ESW was associated with benefits and harms among recipients with immediate graft function (IGF), but only with harms among recipients with DGF. ESW was associated with increased acute rejection (aOR=1.091.161.23), slightly increased graft failure (aHR=1.011.061.12), but decreased mortality (aHR=0.860.890.93) among recipients with IGF. Among recipients with DGF, ESW was associated with a similar increase in rejection (aOR=1.12; 95% CI, 1.02 to 1.23), a more pronounced increase in graft failure (aHR=1.16; 95% CI, 1.08 to 1.26), and no improvement in mortality (aHR=1.00; 95% CI, 0.94 to 1.07). DGF-ESW interaction was statistically significant for graft failure (P=0.04) and mortality (P=0.003), but not for rejection (P=0.6).ConclusionsKT centers in the United States use ESW inconsistently in recipients with DGF. Our findings suggest ESW may lead to worse KT outcomes in recipients with DGF.


2012 ◽  
Vol 26 (5) ◽  
pp. 782-791 ◽  
Author(s):  
Miklos Z. Molnar ◽  
Csaba P. Kovesdy ◽  
Laszlo Rosivall ◽  
Suphamai Bunnapradist ◽  
Junichi Hoshino ◽  
...  

2022 ◽  
Vol 16 (1) ◽  
pp. 38
Author(s):  
Amel Fayed ◽  
Amr ALKouny ◽  
MohammedK ALHarbi ◽  
AbdulrahmanR ALTheaby ◽  
Ghaleb Aboalsamh

Sign in / Sign up

Export Citation Format

Share Document