Clinical Physiology of Circulation

Chief Editor

Leo A. Bockeria, MD, PhD, DSc, Professor, Academician of Russian Academy of Sciences, President of Bakoulev National Medical Research Center for Cardiovascular Surgery


Neural network technologies as a basis for strategy for preventing the complications after myocardial revascularization with cardiopulmonary bypass

Authors: A.S. Semenova 1, V.V. Agapov 1, V.A. Prelatov 1, A.V. Zhukov 1, M.Yu. Shigaev 2

Company:
1 Regional Center for Cardiac Surgery, ul. Krymskaya, 15, Saratov, 410039, Russia; 2 V.I. Razumovskiy Saratov State Medical University, Ministry of Health of the Russian Federation, ul. Bol'shaya Kazach'ya, 112, Saratov, 410012, Russia

Link: Clinical Physiology of Blood Circulaiton. 2013; (): -

Full text:  

Abstract

Objective of the study – is to evaluate the possibilities of predicting the development of complications in early post-operative period after myocardial revascularization based on intra-operational parameters using artificial neural network technologies. Material and methods. The retrospective “case-control” study included data on 107 patients (mean age 58.4±8.5 years), who underwent isolated coronary artery bypass graft surgery. Development of organ dysfunctions (acute kidney, respiratory, cardiovascular, central nervous system dysfunction, myocardial infarction and paroxysmal atrial fibrillation) was considered a complication in an early postoperative period (0–6 days). Hemodynamic indicators and internal milieu parameters routinely monitored during surgical procedure with cardiopulmonary bypass were used as predictors of complications. Neural network method was applied for prediction, training and testing of artificial neural networks as well as the statistical analysis were performed using software package Statistica 6.0 StatSoft Inc. Results. 2 artificial neural networks that showed the best operating characteristics were selected after analysis of more than 3 thousand automatically constructed personal computer-based networks. After entering the values of predictors in binary format neural network gives an output on perspective outcome: presence or absence of complications. The area under the ROC curve was: for EuroSCORE – 0.611 (p=0.48), for artificial neural networks: 0.778 (p=0.027), 0.833 (p=0.034) and 0.944 (p=0.005). Conclusion. We obtained prognostic models – artificial neural networks with high predictive accuracy, suitable for the use in clinical practice.

References

1. Бокерия Л.А., Гудкова Р.Г. Сердечно-сосудистая хирургия – 2010. М.: НЦССХ им. А.Н. Бакулева РАМН; 2011. 2. Miyata H., Motomura N., Murakami A. et al. Effect of benchmarking projects on outcomes of coronary artery bypass graft surgery: Challenges and prospects regarding the quality improvement initiative. J. Thorac. Cardiovasc. Surg. 2012; 143 (6): 1364–9. 3. Mangos G.J., Brown M.A., Chan W.Y. et al. Acute renal failure following cardiac surgery: incidence, outcomes and risk factors. Aust. N. Z. J. Med. 1995; 25: 284–9. 4. Stallwood M.I., Grayson A.D., Mills K. et al. Acute renal failure in coronary artery bypass surgery: independent effect of cardiopulmonary bypass. Ann. Thorac. Surg. 2004; 77 (3): 968–72. 5. Zangrillo A., Landoni G., Sparicio D. et al. Predictors of atrial fibrillation after off-pump coronary artery bypass graft surgery. J. Cardiothorac. Vasc. Anesth. 2004; 18: 704–8. 6. Carnero-Alcazar M., Silva Guisasola J.A., Reguillo Lacruz F.J. et al. Validation of EuroSCORE II on a single-centre 3800 patient cohort. Interact. Cardiovasc. Thorac. Surg. 2013; 16 (3): 293–300. 7. Noyez L., Kievit P.C., van Swieten H.A., de Boer M.J. Cardiac operative risk evaluation: The EuroSCORE II, does it make a real difference? Neth. Heart J. 2012; 20 (12): 494–8. 8. Vanagas G., Kinduris S., Leveckyte A. Comparison of various systems for fisk stratification in heart surgery. Medicina (Kaunas). 2003; 39 (8): 739–44. 9. Бабаев М.А. Синдром полиорганной недостаточности после сердечно-сосудистых операций в условиях искусственного кровообращения: Дис. … д-ра мед. наук. М.; 2011. 10. Lette J., Colletti B.W., Cerino M. et al. Artificial intelligence versus logistic regression statistical modeling to predict cardiac complications after noncardiac surgery. Clin. Cardiol. 1994; 17 (11): 609–14. 11. Orr R.K. Use of a probabilistic neural network to estimate the risk of mortality after cardiac surgery. Med. Desic. Making. 1997; 17 (2): 178–85. 12. Tu J.V., Gurriere M.R. Use of a neural network as a predictive instrument for length of stay in the intensive care unit following cardiac surgery: Proceedings 16th Annual Symposium on Computer Applications in Medical Care. Baltimore; 1992. AMIA: 666–72.

 If you found mistakes, select text and press Alt+A