GlucoSim: Optimizing Insulin Therapy Strategies for Type 2 Diabetes with TFT-Enhanced Simulation
Vol. 28 (2026): 2026 2nd International Conference on Agricultural Sciences, Economics, Biomedical and Environmental Sciences (SEMBE 2026)
Received: 2026-07-18
Accepted: 2026-07-18
Published: 2026-07-18
Abstract
Insulin therapy remains a fundamental cornerstone for effective diabetes management and glycemic control. However, the vast majority of existing research on automated insulin dosing strategies predominantly focuses on type 1 diabetes. This occurs despite the fact that type 2 diabetes mellitus (T2DM) patients comprise the overwhelming majority of diabetes cases worldwide, representing a critical public health challenge. This significant research gap is partly due to the inherent complexity of T2DM pathophysiology and the distinct lack of suitable, highly accurate T2DM simulators required for developing, testing, and evaluating advanced decision-making algorithms. In this work, we propose GlucoSim, a novel and robust computational framework that leverages a Temporal Fusion Transformer (TFT) architecture to accurately model complex blood glucose dynamics in T2DM patients responding to both exogenous insulin injections and variable meal intakes. The trained TFT model effectively serves as a high-fidelity, patient-specific black-box simulator, thereby providing a highly realistic and safe virtual environment for personalized clinical decision support. To the best of our knowledge, this is the first comprehensive study to successfully utilize an advanced time-series prediction model as a high-fidelity simulator specifically tailored for optimizing insulin therapy in T2DM cohorts. Extensive experiments demonstrate that our proposed method can accurately forecast individualized glucose trajectories over extended periods. Furthermore, it provides highly interpretable insights into the underlying metabolic responses, offering promising potential for seamless integration into real-world clinical decision support systems to enhance T2DM management and improve patient outcomes.
Keywords
References
[1] M. Tejedor, A. Z. Woldaregay, and F. Godtliebsen, "Reinforcement learning application in diabetes blood glucose control: A systematic review," Artificial Intelligence in Medicine, vol. 104, p. 101836, 2020.
[2] M. S. Haleem, D. Katsarou, E. I. Georga, G. E. Dafoulas, A. Bargiota, L. Lopez-Perez, et al., "A multimodal deep learning architecture for predicting interstitial glucose for effective type 2 diabetes management," Scientific Reports, vol. 15, no. 1, p. 27625, 2025.
[3] Y. Deng, L. Lu, L. Aponte, A. M. Angelidi, V. Novak, G. E. Karniadakis, and C. S. Mantzoros, "Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients," NPJ Digital Medicine, vol. 4, no. 1, p. 109, 2021.
[4] K. Banumathi, L. Venkatesan, L. S. Benjamin, K. Vijayalakshmi, N. S. Satchi, and N. S. Satchi IV, "Reinforcement learning in personalized medicine: a comprehensive review of treatment optimization strategies," Cureus, vol. 17, no. 4, 2025.
[5] B. Lim, S. Ö. Arık, N. Loeff, and T. Pfister, "Temporal fusion transformers for interpretable multi-horizon time series forecasting," International Journal of Forecasting, vol. 37, no. 4, pp. 1748–1764, 2021.
[6] M. E. Elhefnawy, S. M. S. Ghadzi, and S. Noor Harun, "Predictors associated with type 2 diabetes mellitus complications over time: a literature review," Journal of Vascular Diseases, vol. 1, no. 1, pp. 13–23, 2022.
[7] American Diabetes Association, "Diagnosis and classification of diabetes mellitus," Diabetes Care, vol. 33, Supplement 1, pp. S62–S69, 2010.
[8] S. Ghimire, T. Celik, M. Gerdes, and C. W. Omlin, "Deep learning for blood glucose level prediction: How well do models generalize across different data sets?," PLOS ONE, vol. 19, no. 9, p. e0310801, 2024.
[9] I. S. Mughal, L. Patanè, and R. Caponetto, "A comprehensive review of models and nonlinear control strategies for blood glucose regulation in artificial pancreas," Annual Reviews in Control, vol. 57, p. 100937, 2024.
[10] A. Wallia, and M. E. Molitch, "Insulin therapy for type 2 diabetes mellitus," JAMA, vol. 311, no. 22, pp. 2315–2325, 2014.
[11] P. Aschner, "Insulin therapy in type 2 diabetes," American Journal of Therapeutics, vol. 27, no. 1, pp. e79–e90, 2020.
[12] P. Home, M. Riddle, W. T. Cefalu, C. J. Bailey, R. G. Bretzel, S. Del Prato, et al., "Insulin therapy in people with type 2 diabetes: opportunities and challenges?," Diabetes Care, vol. 37, no. 6, pp. 1499–1508, 2014.
[13] D. E. DeWitt, and I. B. Hirsch, "Outpatient insulin therapy in type 1 and type 2 diabetes mellitus: scientific review," JAMA, vol. 289, no. 17, pp. 2254–2264, 2003.
[14] S. Mudaliar, and S. V. Edelman, "Insulin therapy in type 2 diabetes," Endocrinology and Metabolism Clinics, vol. 30, no. 4, pp. 935–982, 2001.
[15] C. Sorli, and M. K. Heile, "Identifying and meeting the challenges of insulin therapy in type 2 diabetes," Journal of Multidisciplinary Healthcare, pp. 267–282, 2014.
[16] S. G. Swinnen, J. B. Hoekstra, and J. H. DeVries, "Insulin therapy for type 2 diabetes," Diabetes Care, vol. 32, Supplement 2, pp. S253, 2009.
Copyright and License
Published in2026-07-18 17:20:37
DOI https://doi.org/10.70088/retmzz32
Creative Commons
Copyright: © 2026 by the authors.
Submitted for possible open access
publication under the terms and
conditions of the Creative Commons
Attribution (CC BY) license
(https://creativecommons.org/license
s/by/4.0/).
Copyright
Copyright © The Author(s), 2026. Published by SEMBE 2026
Journal Information
- Vol. 28 (2026): 2026 2nd International Conference on Agricultural Sciences, Economics, Biomedical and Environmental Sciences (SEMBE 2026)
- 2026-07-18
- ISSN: (Print) 3078-770X/ (Online) 3078-7718
- Journal Homepage