Introduction: Artificial intelligence (AI) research within medicine is growing rapidly. AI is poised to transform medical practice. AI has been studied in several areas of healthcare and medical practice, including diagnosing, treating and caring of patients. Warfarin is one of the most commonly prescribed oral anticoagulant. Among all anticoagulants, warfarin has long been listed among the top ten drugs causing adverse drug events. Due to narrow therapeutic range and significant side effects, warfarin dosage determination becomes a challenging task in clinical practice. The purpose of this study was to determine exact dose of warfarin needed for patients with artificial heart valve using artificial neural networks (ANN).
Development: To achieved the best model, some multi-layer perceptron ANNs were constructed with different structures. The dataset used included 846 patients who had been referred to the PT clinic in Tehran heart center in the second six months of the year 2013. Finally, the best structure of ANN for warfarin dose was investigated and used for prediction system developments. In this paper the implementation of ANNs and proposed system in MatLab environment are described.
Application: The effectiveness of ANNs were evaluated in terms of classification performance using 10fold cross-validation procedure and the results showed that the best model is a network that has 7 neurons in its hidden layer with an average absolute error of 0.1, disturbance rate of 0.33 and regression of 0.87.
Conclusion: The achieved results reveal that ANN-based system is a suitable tool for warfarin dose prediction in Iranian patients with an artificial heartvalve. However, no system can be guaranteed to achieve 100% accuracy, but using such methods can reduce medical errors and thereby improve health care and patient safety.
Cardiovascular disease: World Health Organization; [updated 2017 May. Available from: http://www.who.int/cardiovascular_diseases/en.
February is American Heart month: Association AH; [updated 2018 February Available from: https://newsroom.heart.org/events/february-american-heart-monthwomen-heart-disease.
Mendis S, Puska P, Norrving B, Organization WH. Global atlas on cardiovascular disease prevention and control: Geneva: World Health Organization; 2011.
Bonow R, Carabello B, de Leon A, Edmunds LH, Fedderly B, Freed M, et al. ACC/AHA guidelines for the management of patients with valvular heart disease. Journal of the American College of Cardiology. 1998;32(5):1486-582.
Rennie LJ. Using Machine Learning to Predict the Effect of Warfarin on Heart Patients: Pennsylvania State University 2004.
Silverstein A, Silverstein VB, Nunn LS. Heart Disease: Lerner Publishing Group; 2006.
Sohrabi MK, Tajik A. Multi-objective feature selection for warfarin dose prediction. Computational Biology and Chemistry. 2017;69:126-33.
Eriksson N, Wadelius M. Prediction of warfarin dose: why, when and how? Pharmacogenomics. 2012;13(4):429-40.
Tao Y, Xiang D, Zhang Y, Jiang B, editors. Swarm ANN/SVR-Based Modeling Method for Warfarin Dose Prediction in Chinese. International Conference in Swarm Intelligence; 2017: Springer.
Saffian SM, Duffull SB, Wright DFB. Warfarin dosing algorithms under‐predict dose requirements in patients requiring≥ 7mg daily: A systematic review and meta‐analysis. Clinical Pharmacology & Therapeutics. 2017.
Sharabiani A, Bress A, Douzali E, Darabi H. Revisiting warfarin dosing using machine learning techniques. Computational and mathematical methods in medicine. 2015;2015.
Sohrabi MK, Tajik A. Effective feature selection of clinical and genetic to predict warfarin dose using artificial neural network. Tehran University Medical Journal TUMS Publications. 2016;73(12):900-5.
Nielsen PB, Lundbye-Christensen S, van der Male M, Larsen TB. Using a personalized decision support algorithm for dosing in warfarin treatment: A randomised controlled trial. Clinical Trials and Regulatory Science in Cardiology. 2017;25:1-6.
Shahmoradi L, Langarizadeh M, Pourmand G, fard ZA, Borhani A. Comparing Three Data Mining Methods to Predict Kidney Transplant Survival. Acta Informatica Medica. 2016;24(5):322-7.
Khashei M, Bijari M. An artificial neural network (p,d,q) model for timeseries forecasting. Expert Systems with Applications. 2010;37(1):479-89.
Pavani A, Naushad SM, Kumar RM, Srinath M, Malempati AR, Kutala VK. Artificial neural network-based pharmacogenomic algorithm for warfarin dose optimization. Pharmacogenomics. 2016;17(2):121-31.
Orooji A, Langarizadeh M, Aghazadeh M, Kamkarhaghighi M, Ghazisaiedi M, Moghbeli F. Dosing of Warfarin in Iranian Adult Patients with An Artificial Heart Valve Using Artificial Neural Networks. Journal of Payavard Salamat. 2018;12(4):249-59.
Sharabiani A, Darabi H, Bress A, Cavallari L, Nutescu E, Drozda K, editors. Machine learning based prediction of warfarin optimal dosing for African American patients. Automation Science and Engineering (CASE), 2013 IEEE International Conference on; 2013: IEEE.
Consortium IWP. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med. 2009;2009(360):753-64.
Tatarunas V, Lesauskaite V, Veikutiene A, Grybauskas P, Jakuska P, Jankauskiene L, et al. The effect of CYP2C9, VKORC1 and CYP4F2 polymorphism and of clinical factors on warfarin dosage during initiation and long-term treatment after heart valve surgery. Journal of thrombosis and thrombolysis. 2014;37(2):177-85.
Hu Y-H, Wu F, Lo C-L, Tai C-T. Predicting warfarin dosage from clinical data: A supervised learning approach. Artificial Intelligence in Medicine. 2012;56(1):27-34.