Anesthesiology; Artificial intelligence; Machine learning; Pediatric cardiac surgery; Humans; Infant; Female; Child; Child, Preschool; Male; Cardiopulmonary Bypass/methods; Logistic Models; Support Vector Machine; Risk Assessment/methods; Area Under Curve; Machine Learning; Cardiac Surgical Procedures/methods; Cardiac Surgical Procedures/adverse effects; Postoperative Complications/epidemiology; Postoperative Complications/diagnosis; Postoperative Complications/etiology; Anesthesiology and Pain Medicine
Abstract :
[en] [en] PURPOSE: Managing children undergoing cardiac surgery with cardiopulmonary bypass (CPB) presents a significant challenge for anesthesiologists. Machine Learning (ML)-assisted tools have the potential to enhance the recognition of patients at risk of complications and predict potential issues, ultimately improving outcomes.
METHODS: We evaluated the prediction capacity of six models, ranging from logistic regression to support vector machine, using a dataset comprising 33 variables and 1364 subjects. The Area Under the Curve (AUC) and the F1 score served as the primary evaluation metrics. Our primary objectives were twofold: first, to develop an effective prediction model, and second, to create a user-friendly comprehensive model for identifying high-risk patients.
RESULTS: The logistic regression model demonstrated the highest effectiveness, achieving an AUC of 83.65%, and an F1 score of 0.7296, with balanced sensitivity and specificity of 77.94% and 76.47%, respectively. In comparison, the comprehensive three-layer decision tree model achieved an AUC of 72.84%, with sensitivity (79.41%) comparable to more complex models.
CONCLUSION: Our machine learning-assisted tools provide an additional perspective and enhance the predictive capabilities of traditional scoring methods. These tools can assist anesthesiologists in making well-informed decisions. Furthermore, we have successfully demonstrated the feasibility of creating a practical white-box model. The next steps involve conducting clinical validation and multicenter cross-validation.
TRIAL REGISTRATION: NCT05537168.
Disciplines :
Anesthesia & intensive care
Author, co-author :
Florquin, Rémi ; Department of Anesthesiology, CHU Charleroi, Chaussée de Bruxelles 140, 6042, Lodelinsart, Belgium. remi.florquin@gmail.com ; Chair of Artificial Intelligence and Digital Medicine, Mons University, 7000, Mons, Belgium. remi.florquin@gmail.com
A.H. Kwon Z.J. Marshall C.S. Nabzdyk Why anesthesiologists could and should become the next leaders in innovative medical entrepreneurism Anesth Analg 2017 124 3 998 1004 10.1213/ANE.0000000000001793 28151812
A. Lienhart Y. Auroy F. Péquignot D. Benhamou J. Warszawski M. Bovet E. Jougla Survey of anesthesia-related mortality in France Anesthesiology 2006 10.1097/00000542-200612000-00008 17122571
B. Kadry W.W. Feaster A. Macario J.M. Ehrenfeld Anesthesia information management systems: past, present, and future of anesthesia records Mt Sinai J Med 2012 79 1 154 165 10.1002/msj.21281 22238048
F. Deng J.V. Hickey Anesthesia information management systems: an underutilized tool for outcomes research AANA J 2015 83 3 189 195 26137760
G. Al Hands-on machine learning with scikit-learn, keras, and tensorflow concepts, tools, and techniques to build intelligent systems 2019 2 Sebastopoled O’Reilly Media, Inc.
G. Briganti On the use of bayesian artificial intelligence for hypothesis generation in psychiatry Psychiatr Danub 2022 34 8 201 206 36170731
P.U. Iyer K.S. Iyer Research in pediatric cardiac anesthesia and intensive care in low- and middle- income countries and low resource settings: challenges and opportunities Ann Pediatr Cardiol 2021 14 3 356 358 10.4103/apc.apc_163_21 34667408 8457288
M. Cvetkovic Challenges in pediatric cardiac anesthesia in developing countries Front Pediatr 2018 10.3389/fped.2018.00254 30420945 6215813
B. Li R. Zhang M. Zhang J. Zheng Current anesthesia practices of pediatric cardiac surgeries in tertiary maternity and children’s hospitals in China: a national survey J Cardiothorac Vasc Anesth 2023 37 7 1213 1222 10.1053/j.jvca.2023.02.047 37002117
D.A. Hashimoto E. Witkowski L. Gao O. Meireles G. Rosman Artificial intelligence in anesthesiology: current techniques, clinical applications, and limitations Anesthesiology 2020 132 2 379 394 10.1097/ALN.0000000000002960 31939856
H.E. Jeffries G. Soto-Campos A. Katch C. Gall T.B. Rice R. Wetzel Pediatric index of cardiac surgical intensive care mortality risk score for pediatric cardiac critical care* Pediatr Crit Care Med 2015 10.1097/PCC.0000000000000489 26196254
A.R. Kang J. Lee W. Jung M. Lee S.Y. Park J. Woo Development of a prediction model for hypotension after induction of anesthesia using machine learning PLoS ONE 2020 15 4 e0231172 1:CAS:528:DC%2BB3cXotlChs78%3D 10.1371/journal.pone.0231172 32298292 7162491
I. Goodfellow Y. Bengio A. Courville Deep learning 2016 Cambridge MIT press
G. Despotis M. Avidan C. Eby Prediction and management of bleeding in cardiac surgery J Thromb Haemost 2009 7 1 111 117 1:CAS:528:DC%2BD1MXpvV2ju7s%3D 10.1111/j.1538-7836.2009.03412.x 19630781
K.J. Jenkins K. Gauvreau J.W. Newburger T.L. Spray J.H. Moller L.I. Iezzoni Consensus-based method for risk adjustment for surgery for congenital heart disease J Thorac Cardiovasc Surg 2002 123 1 110 118 10.1067/mtc.2002.119064 11782764
A. Willems C. Van Lerberghe K. Gonsette A. De Villé C. Melot J.F. Hardy The indication for perioperative red blood cell transfusions is a predictive risk factor for severe postoperative morbidity and mortality in children undergoing cardiac surgery Eur J Cardiothorac Surg 2014 45 6 1050 1057 10.1093/ejcts/ezt548 24431174
A. Székely E. Sápi L. Király A. Szatmári E. Dinya Intraoperative and postoperative risk factors for prolonged mechanical ventilation after pediatric cardiac surgery Paediatr Anaesth 2006 16 11 1166 1175 10.1111/j.1460-9592.2006.01957.x 17040306
M. Zappitelli K.K. Washburn A.A. Arikan L. Loftis Q. Ma P. Devarajan C.R. Parikh S.L. Goldstein Urine neutrophil gelatinase-associated lipocalin is an early marker of acute kidney injury in critically ill children: a prospective cohort study Crit Care 2007 10.1186/cc6089 17678545 2206519
V.R. GaD L. Fred Python 3 reference manual 2009 California CreateSpace
F. Pedregosa Fabian G. Varoquaux Gael A. Gramfort Alexandre V. Michel Vincent B. Thirion Bertrand O. Grisel Olivier Scikit-learn: machine learning in python J Mach Learn Res 2011 12 85 2825 2830
P. Kubben M. Dumontier A. Dekker Fundamentals of clinical data science 2019 Berlin Springer Nature 10.1007/978-3-319-99713-1
R Development Core Team R: a language and environment for statistical computing 2023 362 Vienna R Foundation for Statistical Computing
McKinney W (2010) Data structures for statistical computing in python. In: McKinney W (ed) Proceedings of the 9th Python in Science Conference. Austin, TX, 2010
A.M. Price-Whelan B. Sipőcz H. Günther P. Lim S. Crawford S. Conseil The astropy project: building an open-science project and status of the v2.0 core package Astron J 2018 10.3847/1538-3881/aabc4f
B. Bengfort R. Bilbro Yellowbrick: visualizing the scikit-learn model selection process J Open Sour Softw 2019 4 35 1075 10.21105/joss.01075
R. Charles K. Harris M. Jarrod S.J. van der Walt G. Ralf V. Pauli C. David W. Eric T. Julian B. Sebastian J.S. Nathaniel K. Robert P. Matti H. Stephan M.H. van Kerkwijk B. Matthew H. Allan W. Mark P. Pearu P. Gérard-Marchant S. Kevin R. Tyler W. Warren A. Hameer G. Christoph E.O. Travis Array programming with NumPy Nature 2020 10.1038/s41586-020-2649-2
P.J. Phillips C.A. Hahn P.C. Fontana D.A. Broniatowski M.A. Przybocki Four principles of explainable artificial intelligence 2020 Maryland Gaithersburg
T. Saito M. Rehmsmeier The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets PLoS ONE 2015 10 3 e0118432 10.1371/journal.pone.0118432 25738806 4349800
X. Zeng J. An R. Lin C. Dong A. Zheng J. Li Prediction of complications after paediatric cardiac surgery Eur J Cardiothorac Surg 2020 57 2 350 358 31280308
