OPTIMIZATION OF FLUID VOLUME CONTROL IN HEMODIALYSIS USING FEDERATED LEARNING

Suzana Đorđević; Stefan Ćirković; Danijela Milošević; Milan Gligorijević; Vladimir Mladenović

doi:10.5937/bnsr15-55563

Suzana Đorđević Faculty of Technical Sciences in Čačak, University of Kragujevac, Čačak, Serbia
Stefan Ćirković Faculty of Technical Sciences in Čačak, University of Kragujevac, Čačak, Serbia
Danijela Milošević Faculty of Technical Sciences in Čačak, University of Kragujevac, Čačak, Serbia
Milan Gligorijević Alfa BK University, Belgrade, Serbia
Vladimir Mladenović Faculty of Technical Sciences in Čačak, University of Kragujevac, Čačak, Serbia

DOI: https://doi.org/10.5937/bnsr15-55563

Keywords: Federated learning, Artificial Intelligence, Machine Learning, Overhydration, Hemodialysis

Abstract

Overhydration (OH) represents a significant challenge for hemodialysis patients, significantly affecting the outcomes of their treatment. Accurate prediction and management of overhydration are key to optimizing therapy and improving patients' quality of life. The aim of this paper is to present a federated learning (FL)-based approach designed to predict overhydration in hemodialysis patients, using a dataset comprising different clinical and bioimpedance parameters. Federated learning enables collaborative learning from multiple data sources while preserving the privacy and security of individual patient data. Research results show that federated learning has the potential as an effective tool for predictive modeling in clinical settings. The developed models achieve high performance in overhydration estimation, with metrics confirming their accuracy and reliability. The proposed approach achieved a R² of 0.9999999, a MAE of 0.00018 and an MSE of 0.0031, demonstrating its predictive strength and practical applicability. This study highlights the advantages of federated learning in using distributed data to advance predictive capabilities in healthcare. By overcoming challenges related to privacy and data security, the approach presented in this paper opens up opportunities for more personalized and accurate prognoses, potentially improving decision-making and patient care in hemodialysis.

References

Brentan, B. M., Luvizotto, E. Jr., Herrera M., Izquierdo, J. & Pérez-García, R. 2017. Hybrid regression model for near real-time urban water demand forecasting. Journal of Computational and Applied Mathematics. 309, pp. 532-541. DOI:10.1016/j.cam.2016.02.009.

Buchholz, A. C., Bartok, C. & Schoeller, D. A. 2004. The validity of bioelectrical impedance models in clinical populations. Nutrition in Clinical Practice. 19(5), pp. 433-446. DOI:10.1177/0115426504019005433.

Chicco, D., Warrens, M. J. & Jurman, G. 2021. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE, and RMSE in regression analysis evaluation. PeerJ Computer Science. 7(e623):e623. DOI:10.7717/peerj-cs.623.

Dekker, M. J. E., Marcelli, D., Canaud, B. J., Carioni P., Wang, Y., Grassmann, A., Konings, C. J. A. M., Kotanko, P., Leunissen, K. M., Levin, N. W., et al. 2017. Impact of fluid status and inflammation and their interaction on survival: a study in an international hemodialysis patient cohort. Kidney International. 91(5). pp. 1214-1223. DOI:10.1016/j.kint.2016.12.008.

Djordjevic, S., Kostic, M., Milosevic, D., Cvetkovic, M., Mitrovic, K. & Mladenovic, V. 2023. Prediction of overhydration in the process of pediatric hemodialysis using artificial neural network. In: 2023 12th Mediterranean Conference on Embedded Computing (MECO). IEEE. pp. 1-5. DOI: 10.1109/MECO58584.2023.10154915.

Djordjevic, S., Kostic, M., Zanaj, B., Milosevic, D. & Mladenovic, V. In press. New classes of hybrid machine learning algorithms: elastic boosting, elastic support regressor (case study pediatric hemodialysis). Measurement Science Review.

Donini, L. M., Busetto, L., Bischoff, S. C., Cederholm, T., Ballesteros-Pomar, M. D., Batsis, J. A., Bauer, J. M., Boirie, Y., Cruz-Jentoft, A. J., Dicker, D. et al. 2022. Definition and diagnostic criteria for sarcopenic obesity: ESPEN and EASO consensus statement. Obesity Facts. 15(3), pp. 321-335. DOI:10.1159/000521241

Doshi, M., Streja, E., Rhee, C. M., Park, J., Ravel, V. A., Soohoo, M., Moradi, H., Lau, W. L., Mehrotra, R., Kuttykrishnan, S, et al. 2016. Examining the robustness of the obesity paradox in maintenance hemodialysis patients: a marginal structural model analysis. Nephrology Dialysis Transplantation. 31(8), pp. 1310-1319. DOI:10.1093/ndt/gfv379.

gRPC 2024. gRPC. Available at: https://grpc.io/ (Accessed: 29 November 2024)

Hans, C. 2011. Elastic net regression modeling with the orthant normal prior. Journal of the American Statistical Association. 106(496), pp. 1383-1393. DOI:10.1198/jasa.2011.tm09241.

He, J., Baxter, S. L., Xu, Jie, Xu Jiming, Zhou, X. & Zhang, K. 2019. The practical implementation of artificial intelligence technologies in medicine. Nature Medicine. 25(1), pp. 30-36. DOI:10.1038/s41591-018-0307-0.

Huang, C. T., Wang, T. J., Kuo, L. K., Tsai, M. J., Cia, C. T., Chiang, D. H., Chang, P. J., Chong, I. W., Tsai, Y. S., Chu, Y. C. et al. 2023. Federated machine learning for predicting acute kidney injury in critically ill patients: a multicenter study in Taiwan. Health Information Science and Systems. 11(1), pp. 48. DOI:10.1007/s13755-023-00248-5.

Huan-Sheng, C., Yeong-Chang, C., Ming-Hsing, H., Fan-Lieh, T., Chu-Cheng, L., Tsai-Kun, W., Hung-Ping, C., Sze-Hung, H., Hsien-Chang, C., Chia-Chen, L. et al. 2016. Application of bioimpedance spectroscopy in Asian dialysis patients (ABISAD-III): a randomized controlled trial for clinical outcomes. International Urology and Nephrology. 48(11), pp. 1897-1909. DOI:10.1007/s11255-016-1415-8.

Ishimura, E., Okuno, S., Nakatani, S., Mori, K., Miyawaki, J., Okazaki, H., Sugie, N., Norimine, K., Yamakawa, K., Tsujimoto, Y. et al. 2022. Significant association of diabetes with mortality of chronic hemodialysis patients, independent of the presence of obesity, sarcopenia, and sarcopenic obesity. Journal of Renal Nutrition. 32(1), pp. 94-101. DOI:10.1053/j.jrn.2021.07.003.

Jaladanki, S. K., Vaid, A., Sawant, A. S., Xu, J., Shah, K., Dellepiane, S., Paranjpe, I., Chan, L., Kovatch, P., Charney, A. W. et al. 2021. Development of a federated learning approach to predict acute kidney injury in adult hospitalized patients with COVID-19 in New York City. medRxiv. DOI:10.1101/2021.07.25.21261105.

Kondo, E., Sagayama, H., Yamada, Y., Shiose, K., Osawa T, Motonaga K, Ouchi S, Kamei A, Nakajima K., Higaki, Y. et al. 2018. Energy deficit required for rapid weight loss in elite collegiate wrestlers. Nutrients. 10(5). DOI:10.3390/nu10050536.

Konečný, J., McMahan, H. B., Yu, F. X., Richtrik, P., Suresh, A. T. & Bacon D. 2016. Federated learning: strategies for improving communication efficiency. arXiv Preprint. DOI: 10.48550/arXiv.1610.05492.

Li, T., Sahu, A. K., Talwalkar, A. & Smith V. 2019. Federated learning: Challenges, methods, and future directions. arXiv [csLG]. DOI:10.48550/ARXIV.1908.07873.

Loftus, T. J., Ruppert, M. M., Shickel, B., Ozrazgat-Baslanti, T., Balch, J. A., Efron, P. A., Upchurch, G. R. Jr., Rashidi, P., Tignanelli, C., Bian, J. et al. 2022. Federated learning for preserving data privacy in collaborative healthcare research. Digital Health. 8:20552076221134455. DOI:10.1177/20552076221134455.

London, G. M. 2011. Ultrafiltration intensification for achievement of dry weight and hypertension control is not always the therapeutic gold standard. Journal of Nephrology. 24(4), pp. 395-397. DOI:10.5301/jn.5000006.

Loutradis, C., Sarafidis, P A., Ferro, C. J. & Zoccali, C. 2021. Volume overload in hemodialysis: diagnosis, cardiovascular consequences, and management. Nephrology Dialysis Transplantation. 36(12), pp. 2182-2193. DOI:10.1093/ndt/gfaa182.

Marcelli, D., Usvyat, L. A., Kotanko, P., Bayh, I., Canaud, B., Etter, M., Gatti, E., Grassmann, A., Wang, Y., Marelli, C. et al. 2015. Body composition and survival in dialysis patients: results from an international cohort study: Results from an international cohort study. Clinical Journal of the American Society of Nephrology. 10(7), pp. 1192-1200. DOI:10.2215/CJN.08550814.

McIntyre, C. W. 2010. Recurrent circulatory stress: The dark side of dialysis: Recurrent circulatory stress in hemodialysis. Seminars in Dialysis. 23(5), pp. 449-451. DOI:10.1111/j.1525-139x.2010.00782.x.

McMahan, H. B., Moore, E., Ramage, D., Hampson, S. & Arcas, B. A. 2017. Communication-efficient learning of deep networks from decentralized data. In: Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. pp. 1273-1282. DOI: 10.48550/arXiv.1602.05629.

Mladenović, V., Kostić, M., Milošević, D., Zanaj, E. & Đorđević, S. 2024. System for prediction and balancing excess fluid in the body during hemodialysis based on artificial intelligence. Patent RS20240030A2. Available from: https://worldwide.espacenet.com/patent/search/family/090057728/publication/RS20240030A2?q=pn%3DRS20240030A2

Moissl, U., Arias-Guillén, M., Wabel, P., Fontseré, N., Carrera, M., Campistol, J. M. & Maduell, F. 2013. Bioimpedance-guided fluid management in hemodialysis patients. Clinical Journal of the American Society of Nephrology. 8(9), pp. 1575-1582. DOI:10.2215/CJN.12411212.

Natekin, A. & Knoll, A. 2013. Gradient boosting machines, a tutorial. Front Neurorobot. 7, 21. DOI:10.3389/fnbot.2013.00021.

Nguyen, T. V., Dakka, M. A., Diakiw, S. M., VerMilyea, M. D., Perugini, M., Hall, J. M. M. & Perugini, D. 2022. A novel decentralized federated learning approach to train on globally distributed, poor quality, and protected private medical data. Scientific Reports. 12(1), 8888. DOI:10.1038/s41598-022-12833-x.

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G. et al. 2012. Scikit-learn: Machine Learning in Python. arXiv [csLG]. DOI:10.48550/ARXIV.1201.0490.

Perl, J., Dember, L. M., Bargman, J. M., Browne, T., Charytan, D. M., Flythe, J. E., Hickson, L. J., Hung, A. M., Jadoul, M., Lee, T. C. et al. 2017. The use of a multidimensional measure of dialysis adequacy-moving beyond small solute kinetics. Clinical Journal of the American Society of Nephrology. 12(5), pp. 839-847. DOI:10.2215/CJN.08460816.

Rymarz, A., Gibińska, J., Zajbt, M., Piechota, W. & Niemczyk, S. 2018. Low lean tissue mass can be a predictor of one-year survival in hemodialysis patients. Renal Failure. 40(1), pp. 231-237. DOI:10.1080/0886022X.2018.1456451.

Sadilek, A., Liu, L., Nguyen, D., Kamruzzaman, M., Serghiou, S., Rader, B., Ingerman, A., Mellem, S., Kairouz, P., Nsoesie, E. O. et al. 2021. Privacy-first health research with federated learning. NPJ Digital Medicine. 4(1), 132. DOI:10.1038/s41746-021-00489-2.

Sagayama, H., Jikumaru, Y., Hirata, A., Yamada, Y., Yoshimura, E., Ichikawa, M., Hatamoto, Y., Ebine, N., Kiyonaga, A., Tanaka, H. et al. 2014. Measurement of body composition in response to a short period of overfeeding. Journal of Physiological Anthropology. 33(1), 29. DOI:10.1186/1880-6805-33-29.

Sagayama, H., Yoshimura, E., Yamada, Y. & Tanaka H. 2019. The effects of rapid weight loss and 3-h recovery on energy expenditure, carbohydrate, and fat oxidation in boxing athletes. The Journal of Sports Medicine and Physical Fitness. 59(6), pp. 1018-1025. DOI:10.23736/S0022-4707.18.08677-2.

Sheller, M. J., Edwards, B., Reina, G.A., Martin, J., Pati, S., Kotrotsou, A., Milchenko, M., Xu, W., Marcus, D., Colen, R. R. et al. 2020. Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data. Scientific Reports. 10(1), 12598. DOI:10.1038/s41598-020-69250-1.

Sheller, M. J., Reina, G. A., Edwards, B., Martin, J. & Bakas, S. 2019. Multi-institutional deep learning modeling without sharing patient data: a feasibility study on brain tumor segmentation. In: Crimi A, Bakas S, Kuijf H, Keyvan F, Reyes M, van Walsum T, editors. Brainlesion: glioma, multiple sclerosis, stroke and traumatic brain injuries. Lecture Notes in Computer Science. pp. 92-104. DOI: 10.1007/978-3-030-11723-8_9.

Shiri, I., Vafaei Sadr, A., Akhavan, A., Salimi, Y., Sanaat, A., Amini, M., Razeghi, B., Saberi, A., Arabi, H., Ferdowsi, S. et al. 2023. Decentralized collaborative multi-institutional PET attenuation and scatter correction using federated deep learning. The European Journal of Nuclear Medicine and Molecular Imaging. 50(4), pp. 1034-1050. DOI:10.1007/s00259-022-06053-8.

Sinha, A. D. & Agarwal R. 2009. Can chronic volume overload be recognized and prevented in hemodialysis patients? The pitfalls of the clinical examination in assessing volume status. Seminars in Dialysis. 22(5), pp. 480-482. DOI:10.1111/j.1525-139X.2009.00641.x.

Yang, Q., Liu, Y., Chen, T. & Tong, Y. 2019. Federated Machine Learning: Concept and Applications. arXiv [csAI]. DOI:10.48550/ARXIV.1902.04885.

Yang, Y., Zhang, H., Lan, X., Qin, X., Huang, Y., Wang, J., Luo, P., Wen, Z., Li, Y., Kong, Y. et al. 2023. Low BMI and high waist-to-hip ratio are associated with mortality risk among hemodialysis patients: a multicenter prospective cohort study. Clinical Kidney Journal. 16(1), pp. 167-175. DOI:10.1093/ckj/sfac210.

Zhang, F., Kreuter, D., Chen, Y., Dittmer, S., Tull, S., Shadbahr, T., BloodCounts! consortium, Preller, J., Rudd, J. H. F., Aston, J. A. D, et al. 2024. Recent methodological advances in federated learning for healthcare. Patterns. 5(6), 101006. DOI:10.1016/j.patter.2024.101006.

Zoccali, C., Moissl, U., Chazot, C., Mallamaci, F., Tripepi, G., Arkossy, O., Wabel, P. & Stuard, S. 2017. Chronic fluid overload and mortality in ESRD. The Journal of the American Society of Nephrology. 28(8), pp. 2491-2497. DOI:10.1681/ASN.2016121341.