medigraphic.com
Colegio de Medicina Interna de México.

2025, Number 04

<< Back Next >>

Med Int Mex 2025; 41 (04)

Neural network model to predict 14-day dialysis in patients with acute renal failure

Guevara TA

Language: Spanish
References: 17
Page: 219-233
PDF size: 235.34 Kb.


ABSTRACT

Objectives: To develop and evaluate a multilayer perceptron neural network model to predict the need for 14-day dialysis in hospitalized patients with acute renal failure.
Materials and Methods: Analytical, longitudinal study using an international secondary database available in the Dryad scientific and medical data repository (https://datadryad.org) of patients with acute kidney failure; 42 clinical and laboratory variables were included. The database was divided into training (69.20%) and test (30.8%). An artificial neuronal network was trained with a hidden layer of 9 neurons (hyperbolic tangent) and output with Softmax function, using cross entropy as loss function. Model performance was evaluated by accuracy, area under the curve (AUC) and classification metrics.
Results: There were included 4985 patients. The model had a cross-entropy error of 64,582 (training) and 60,260 (testing), with error rates of 0.7% and 1.0%. The key neurons were H1:2, H1:3, H1:6, and H1:7, with H1:3 (previous dialysis, SOFA, creatinine) standing out. The AUC was 0.995, with accuracy of 99.30% (training) and 99% (testing). The pseudo-probability distribution showed high confidence in classification, and the gain curve identified almost 100% of positive cases in the 10% of highest risk.
Conclusions: The multilayer perceptron network showed high accuracy in dialysis prediction in acute renal failure patients, supporting clinical decision making. Validation in external cohorts is recommended.


REFERENCES

  1. Scarpioni R, Valsania T, Albertazzi V, Blanco V, et al. Acutekidney injury, a common and severe complication inhospitalized patients during the COVID-19 pandemic. JNephrol 2021; 34 (4): 1019-24. http://dx.doi.org/10.1007/s40620-021-01087-x

  2. Gameiro J, Fonseca JA, Outerelo C, Lopes JA. Acute kidneyinjury: From diagnosis to prevention and treatmentstrategies. J Clin Med 2020; 9 (6): 1704. http://dx.doi.org/10.3390/jcm9061704

  3. Loeffler J, Hassan M, Qaqish F, Dimachkie R, et al. Predictorsof renal replacement therapy requirement in cirrhoticpatients with acute kidney injury. Dig Dis Sci 2025. http://dx.doi.org/10.1007/s10620-025-08881-8

  4. Mas-Font S, Ros-Martinez J, Pérez-Calvo C, Villa-Díaz P,et al. Prevention of acute kidney injury in Intensive CareUnits. Med Intensiva (Engl Ed) 2017; 41 (2): 116-26. https://linkinghub.elsevier.com/retrieve/pii/S2173572717300395

  5. Lai T-S, Tsao H-M, Chou Y-H, Liang S-L, et al. A competingrisk predictive model for kidney failure in patientswith advanced chronic kidney disease. J Formos MedAssoc 2024; 123 (7): 751-7. http://dx.doi.org/10.1016/j.jfma.2023.11.010

  6. Bajaj T, Koyner JL. Artificial intelligence in acute kidneyinjury prediction. Adv Chronic Kidney Dis 2022; 29 (5):450-60. http://dx.doi.org/10.1053/j.ackd.2022.07.009

  7. Wilson F. Automated, medication-targeted alerts for AcuteKidney Injury – A randomized trial. Dryad 2022. http://dx.doi.org/10.5061/DRYAD.KH189327P

  8. Wilson FP, Yamamoto Y, Martin M, Coronel-Moreno C, LiF, Cheng C, et al. A randomized clinical trial assessing theeffect of automated medication-targeted alerts on acutekidney injury outcomes. Nat Commun 2023; 14 (1): 1-10.https://www.nature.com/articles/s41467-023-38532-3

  9. Hoste EAJ, Bagshaw SM, Bellomo R, Cely CM, et al. Epidemiologyof acute kidney injury in critically ill patients: the multinationalAKI-EPI study. Intensive Care Med 2015; 41 (8):1411-23. http://dx.doi.org/10.1007/s00134-015-3934-7

  10. The REporting of studies Conducted using ObservationalRoutinely-collected health Data (RECORD) Statement.https://www.equator-network.org/reporting-guidelines/record/

  11. Montesinos López OA, Montesinos López A, Crossa J. Fundamentalsof artificial neural networks and deep learning.In: Multivariate statistical machine learning methods forgenomic prediction. Cham: Springer International Publishing;2022: 379-425. http://dx.doi.org/10.1007/978-3-030-89010-0_10

  12. Isabona J, Imoize AL, Ojo S, Karunwi O, Kim Y, Lee C-C, et al.Development of a multilayer perceptron neural network foroptimal predictive modeling in urban microcellular radioenvironments. Appl Sci (Basel) 2022; 12 (11): 5713. https://www.mdpi.com/2076-3417/12/11/5713

  13. Kadnár M, Káčer P, Harničárová M, Valíček J, et al. Comparisonof linear regression and artificial neural networkmodels for the dimensional control of the welded stampedsteel arms. Machines 2023; 11 (3): 376. https://www.mdpi.com/2075-1702/11/3/376

  14. Demirjian S, Chertow GM, Zhang JH, O’Connor TZ, et al.Model to predict mortality in critically ill adults with acutekidney injury. Clin J Am Soc Nephrol 2011; 6 (9): 2114-20.http://dx.doi.org/10.2215/cjn.02900311

  15. De Rosa S, Greco M, Rauseo M, Annetta MG. The good,the bad, and the serum creatinine: Exploring the effect ofmuscle mass and nutrition. Blood Purif 2023; 52 (9-10):775-85. http://dx.doi.org/10.1159/000533173

  16. Du J, Zhang W, Niu J, Wang S. Association between bloodurea nitrogen levels and the risk of diabetes mellitus inChinese adults: secondary analysis based on a multicenter,retrospective cohort study. Front Endocrinol (Lausanne)2024; 15. http://dx.doi.org/10.3389/fendo.2024.1282015

  17. Gounden V, Bhatt H, Jialal I. Renal function tests. En: Stat-Pearls. Treasure Island (FL): StatPearls Publishing; 2025.https://www.ncbi.nlm.nih.gov/books/NBK507821/




2020     |     www.medigraphic.com