Accurate and transparent blood glucose forecasting is crucial for effective diabetes management. This paper presents an interpretable deep learning (DL) framework based on multi-task learning (MTL) for forecasting glucose levels at multiple prediction horizons. In contrast to complex hybrid systems, we isolate the MTL core and integrate explainability methods such as Shapley Additive Explanations (SHAP) and permutation-based feature importance (PFI). Our approach enables a clear understanding of model behavior while achieving strong predictive performance. Evaluated on the BrisT1D dataset, the model achieves an R^2 score of 0.956, RMSE of 0.045, and MAE of 0.033, while highlighting the critical influence of historical glucose readings. This focused study provides insights into how interpretable AI can support reliable decision-making in diabetes care.  

Keywords - Blood glucose forecasting, multi-task learning, explainable AI, attention mechanism, diabetes prediction

[1] P. Aschner, S. Karuranga, S. James, D. Simmons, A. Basit, J. E. Shaw, et al., "The International Diabetes Federation’s guide for diabetes epidemiological studies," Diabetes Research and Clinical Practice, vol. 172, 2021.

[2] K. Dovc, S. Lanzinger, R. Cardona-Hernandez, M. Tauschmann, M. Marigliano, V. Cherubini, R. Preikˇsa, U. Schierloh, H. Clapin, F. AlJaser et al., “Association of achieving time in range clinical targets with treatment modality among youths with type 1 diabetes,” JAMA network open, vol. 6, no. 2, pp. e230 077–e230 077, 2023.

[3] M. Jaloli and M. Cescon, “Long-term prediction of blood glucose levels in type 1 diabetes using a cnn-lstm-based deep neural network,” Journal of diabetes science and technology, vol. 17, no. 6, pp. 1590–1601, 2023.

[4] S.-M. Lee, D.-Y. Kim, and J. Woo, “Glucose transformer: Forecasting glucose level and events of hyperglycemia and hypoglycemia,” IEEE Journal of Biomedical and Health Informatics, vol. 27, no. 3, pp. 1600 1611, 2023.

[5] A. Z. Woldaregay, E. ˚Arsand, T. Botsis, D. Albers, L. Mamykina, and G. Hartvigsen, “Data-driven blood glucose pattern classification and anomalies detection: machine-learning applications in type 1 diabetes,” Journal of medical Internet research, vol. 21, no. 5, p. e11030, 2019.

[6] H. Y. Lu, P. Lu, J. E. Hirst, L. Mackillop, and D. A. Clifton, “A stacked long short-term memory approach for predictive blood glucose monitoring in women with gestational diabetes mellitus,” Sensors, vol. 23, no. 18, p. 7990, 2023.

[7] S. Rancati, P. Bosoni, R. Schiaffini, A. Deodati, P. A. Mongini, L. Sacchi, C. Toffanin, and R. Bellazzi, “Exploration of foundational models for blood glucose forecasting in type-1 diabetes pediatric patients,” Diabetology, vol. 5, no. 6, pp. 584–599, 2024.

[8] F. Prendin, J. Pavan, G. Cappon, S. Del Favero, G. Sparacino, and A. Facchinetti, “The importance of interpreting machine learning models for blood glucose prediction in diabetes: an analysis using shap,” Scientific reports, vol. 13, no. 1, p. 16865, 2023.

[9] I. Fox, “Machine learning for physiological time series: Representing and controlling blood glucose for diabetes management,” Ph.D. dissertation, 2020.

[10] S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” Advances in neural information processing systems, vol. 30, 2017.

[11] M. T. Ribeiro, S. Singh, and C. Guestrin, “” why should i trust you?” explaining the predictions of any classifier,” in Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 2016, pp. 1135–1144.

[12] L. Pan, W. Sun, W. Wan, Q. Zeng, and J. Xu, “Research progress of diabetic disease prediction model in deep learning,” Journal of Theory and Practice of Engineering Science, vol. 3, no. 12, pp. 15–21, 2023.

[13] J. U. Mendis, “A predictive model for forecasting the severity of recall of cardiac implantable electronic devices,” Ph.D. dissertation, The George Washington University, 2025.

[14] S. J. Chu, N. Amarasiri, S. Giri, and P. Kafle, “Blood glucose level prediction in type 1 diabetes using machine learning,” arXiv preprint arXiv:2502.00065, 2025.

[15] E. Afsaneh, A. Sharifdini, H. Ghazzaghi, and M. Z. Ghobadi, “Recent applications of machine learning and deep learning models in the prediction, diagnosis, and management of diabetes: a comprehensive review,” Diabetology & Metabolic Syndrome, vol. 14, no. 1, p. 196, 2022.

[16] Z. Zhang, H. Lu, S. Ma, J. Peng, C. Lin, N. Li, and B. Dong, “A general framework for generative self-supervised learning in non-invasive estimation of physiological parameters using photoplethysmography,” Biomedical Signal Processing and Control, vol. 98, p. 106788, 2024.

[17] J. Xiao, B. Chen, L. Chen, Q. Wang, S. Tan, H. Yuan, D. Xiang, B. Zhang, X. Li, S. Huang et al., “Interpretable time- series neural turing machine for prognostic prediction of patients with type 2 diabetes in physician-pharmacist collaborative clinics,” International Journal of Medical Informatics, vol. 195, p. 105737, 2025.

[18] J. Lei, X. Sun, Y. Li, K. Li, S. Zhang, H. Zeng, A. Qin, E. Chi, and Y. Zhang, “An improved adaptive glucose control approach for type 1 diabetes with temporal dependence,” in 2024 IEEE 9th International Conference on Computational Intelligence and Applications (ICCIA). IEEE, 2024, pp. 209–214.

[19] S. Hussain, S. Haider, S. Maqsood, R. Damaševičius, R. Maskeliūnas, and M. Khan, "ETISTP: An enhanced model for brain tumor identification and survival time prediction," Diagnostics, vol. 13, no. 8, p. 1456, 2023.

[20] S. Maqsood and R. Damaševičius, "Monkeypox detection and classification using deep learning based features selection and fusion approach," in Proc. 2023 IEEE Int. Systems Conf. (SysCon), Apr. 2023, pp. 1–8.

[21] R. M. Al-Tam, F. A. Hashim, S. Maqsood, L. Abualigah, and R. M. Alwhaibi, "Enhancing Parkinson’s disease diagnosis through stacking ensemble-based machine learning approach," IEEE Access, vol. 12, pp. 79549–79567, 2024.