Application of an interpretable neural network framework based on the LASSO-proj algorithm for warfarin dose prediction_Journal of Biomedical Engineering

Authors：

ZHONG Chenlu ¹ ,  ZHU Ye ² , GU Xiang ²

1. Department of Cardiology, Gaoyou People’s Hospital, Yangzhou, Jiangsu 225001, P. R. China;
2. Department of Cardiology, Northern Jiangsu People’s Hospital, Yangzhou, Jiangsu 225001, P. R. China;

Corresponding?author：

ZHU Ye, Email: 307971331@qq.com

Keywords：

Warfarin; Dose prediction; Deep learning; Artificial intelligence

DOI：

10.7507/1001-5515.202501064

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

Warfarin, a classic oral anticoagulant, is characterized by a narrow therapeutic window and considerable interindividual variability in dosing requirements. This makes precise dose adjustment challenging in clinical practice and increases the risk of bleeding or thrombosis. To improve dose prediction, this study developed a streamlined multilayer perceptron (MLP) model using real-world data from the International Warfarin Pharmacogenomics Consortium (IWPC) database. The LASSO-proj algorithm was applied for high-precision feature selection prior to model construction. The resulting model demonstrated strong predictive performance on the test set, achieving a coefficient of determination (R²) of 0.456, a mean absolute error (MAE) of 8.92 mg/week, and 48.522% of its predictions falling within ±20% of the actual stable therapeutic dose. Through SHAP-based interpretation using DeepExplainer, key features influencing warfarin dosing were identified, including the VKORC1 genotype, body weight, age, and ethnicity. The interpretable MLP framework incorporating LASSO-proj not only maintains high predictive accuracy, but also significantly enhances model transparency, providing a valuable tool for guiding warfarin therapy.

1.	Hart R G, Pearce L A, Aguilar M I. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med, 2007, 146(12): 857-867.
2.	Turen S, Turen S. Determination of factors affecting time in therapeutic range in patients on warfarin therapy. Biol Res Nurs, 2023, 25(1): 170-178.
3.	張進華, 劉茂柏, 蔡銘智, 等. 模型引導的華法林精準用藥: 中國專家共識(2022版). 中國臨床藥理學與治療學, 2022, 27(11): 1201-1212.
4.	Tan O, Wu J Z, Yeo K K, et al. Building a predictive model for warfarin dosing via machine learning. Eur Heart J, 2020, 41(Suppl 2): ehaa946.3491.
5.	Tao Y, Chen Y J, Fu X, et al. Evolutionary ensemble learning algorithm to modeling of warfarin dose prediction for Chinese. IEEE J Biomed Health Inform, 2019, 23(1): 395-406.
6.	International Warfarin Pharmacogenetics Consortium; Klein T E, Altman R B, et al. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med, 2009, 360(8): 753-764.
7.	Truda G, Marais P. Evaluating warfarin dosing models on multiple datasets with a novel software framework and evolutionary optimisation. J Biomed Inform, 2021, 113: 103634.
8.	Tibshirani R. Regression shrinkage and selection via the lasso. J R Stat Soc Ser B, 1996, 58(1): 267-288.
9.	Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw, 2010, 33(1): 1-22.
10.	Meinshausen N, Bühlmann P. Stability selection. J R Stat Soc Ser B, 2010, 72(4): 417-473.
11.	LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436-444.
12.	Tivnan M, Gang G J, Wang W, et al. Tunable neural networks for CT image formation. J Med Imaging (Bellingham), 2023, 10(3): 033501.
13.	Chicco D, Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics, 2020, 21(1): 6.
14.	Yan M, Liu H, Xu Q, et al. Development and validation of a prediction model for in-hospital death in patients with heart failure and atrial fibrillation. BMC Cardiovasc Disord, 2023, 23(1): 505.
15.	Lundberg S M, Lee S I. A unified approach to interpreting model predictions// Advances in Neural Information Processing Systems 30. Long Beach: Curran Associates, 2017: 4765-4774.
16.	Abusitta A, Li M Q, Fung B C M. Survey on explainable AI: techniques, challenges and open issues. Expert Syst Appl, 2024, 255(C): 124710.
17.	Jonas D E, McLeod H L. Genetic and clinical factors relating to warfarin dosing. Trends Pharmacol Sci, 2009, 30(7): 375-386.
18.	Zou H, Hastie T. Regularization and variable selection via the elastic net. J R Stat Soc Ser B, 2005, 67(2): 301-320.
19.	Liu R, Li X, Zhang W, et al. Comparison of nine statistical model based warfarin pharmacogenetic dosing algorithms using the racially diverse International Warfarin Pharmacogenetic Consortium cohort database. PLoS One, 2015, 10(8): e0135784.
20.	Wang Z, Poon J, Yang J, et al. Warfarin dose estimation on high-dimensional and incomplete data// Proceedings of the 54th Hawaii International Conference on System Sciences. Maui: University of Hawaii, 2021: 3455-3463.
21.	Jahmunah V, Chen S, Oh S L, et al. Automated warfarin dose prediction for Asian, American, and Caucasian populations using a deep neural network. Comput Biol Med, 2023, 153: 106548.

1. Hart R G, Pearce L A, Aguilar M I. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med, 2007, 146(12): 857-867.
2. Turen S, Turen S. Determination of factors affecting time in therapeutic range in patients on warfarin therapy. Biol Res Nurs, 2023, 25(1): 170-178.
3. 張進華, 劉茂柏, 蔡銘智, 等. 模型引導的華法林精準用藥: 中國專家共識(2022版). 中國臨床藥理學與治療學, 2022, 27(11): 1201-1212.
4. Tan O, Wu J Z, Yeo K K, et al. Building a predictive model for warfarin dosing via machine learning. Eur Heart J, 2020, 41(Suppl 2): ehaa946.3491.
5. Tao Y, Chen Y J, Fu X, et al. Evolutionary ensemble learning algorithm to modeling of warfarin dose prediction for Chinese. IEEE J Biomed Health Inform, 2019, 23(1): 395-406.
6. International Warfarin Pharmacogenetics Consortium; Klein T E, Altman R B, et al. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med, 2009, 360(8): 753-764.
7. Truda G, Marais P. Evaluating warfarin dosing models on multiple datasets with a novel software framework and evolutionary optimisation. J Biomed Inform, 2021, 113: 103634.
8. Tibshirani R. Regression shrinkage and selection via the lasso. J R Stat Soc Ser B, 1996, 58(1): 267-288.
9. Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw, 2010, 33(1): 1-22.
10. Meinshausen N, Bühlmann P. Stability selection. J R Stat Soc Ser B, 2010, 72(4): 417-473.
11. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436-444.
12. Tivnan M, Gang G J, Wang W, et al. Tunable neural networks for CT image formation. J Med Imaging (Bellingham), 2023, 10(3): 033501.
13. Chicco D, Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics, 2020, 21(1): 6.
14. Yan M, Liu H, Xu Q, et al. Development and validation of a prediction model for in-hospital death in patients with heart failure and atrial fibrillation. BMC Cardiovasc Disord, 2023, 23(1): 505.
15. Lundberg S M, Lee S I. A unified approach to interpreting model predictions// Advances in Neural Information Processing Systems 30. Long Beach: Curran Associates, 2017: 4765-4774.
16. Abusitta A, Li M Q, Fung B C M. Survey on explainable AI: techniques, challenges and open issues. Expert Syst Appl, 2024, 255(C): 124710.
17. Jonas D E, McLeod H L. Genetic and clinical factors relating to warfarin dosing. Trends Pharmacol Sci, 2009, 30(7): 375-386.
18. Zou H, Hastie T. Regularization and variable selection via the elastic net. J R Stat Soc Ser B, 2005, 67(2): 301-320.
19. Liu R, Li X, Zhang W, et al. Comparison of nine statistical model based warfarin pharmacogenetic dosing algorithms using the racially diverse International Warfarin Pharmacogenetic Consortium cohort database. PLoS One, 2015, 10(8): e0135784.
20. Wang Z, Poon J, Yang J, et al. Warfarin dose estimation on high-dimensional and incomplete data// Proceedings of the 54th Hawaii International Conference on System Sciences. Maui: University of Hawaii, 2021: 3455-3463.
21. Jahmunah V, Chen S, Oh S L, et al. Automated warfarin dose prediction for Asian, American, and Caucasian populations using a deep neural network. Comput Biol Med, 2023, 153: 106548.

Journal of Biomedical Engineering

Latest ArticlesApplication of an interpretable neural network framework based on the LASSO-proj algorithm for warfarin dose prediction

Abstract Full text Figures/Tables Video References Cited by

Format

Content