Inge CuC

Desarrollo de un Modelo Basado en Aprendizaje Automático para Predecir los Niveles de Glucosa en Sangre

Omar Andres Aguirre Aguirre

Universidad Nacional de Colombia

Eduardo Nicolas Guevara Povea

Universidad Nacional de Colombia

Juan Pablo Hoyos Sanchez

Universidad Nacional de Colombia

https://orcid.org/0000-0002-1844-9127

DOI: https://doi.org/10.17981/ingecuc.21.2.2025.15

Palabras clave: Aprendizaje automático, diabetes tipo 1, modelos, predicción, glucosa

---

Resumen

Un monitoreo preciso de los niveles de glucosa en sangre es fundamental para el manejo efectivo de la diabetes tipo 1, especialmente debido a la complejidad y variabilidad individual de esta enfermedad. Por tanto, para prevenir accidentes y/o alternaciones es necesario desarrollar modelos de aprendizaje automático capaces de predecir a corto plazo (1hora) los niveles de glucosa. En este artículo se propone una solución siguiendo una metodología de investigación que consta de 6 pasos: recolección y procesamiento de datos, entrenamiento y selección del modelo, métricas de evaluación, validación, visualización de las predicciones y aplicación de grid search. Los algoritmos usados fueron regresión lineal, árboles de decisión, Random Forest, XGBoost y TabNet. El dataset incluyó mediciones continuas de glucosa, administración de insulina, ingesta de carbohidratos y niveles de actividad física registrados mediante dispositivos portátiles.

Los modelos fueron evaluados mediante métricas estándar (MAE, MAPE, RMSE, R², EVR) sobre los conjuntos de entrenamiento y prueba. Los resultados mostraron que el mejor modelo predictor, Random Forest, tuvo un RMSE de solo 0.75 y un coeficiente de determinación R² de 0.94, evidenciando alta precisión y capacidad de generalización. También se encontró que el ajuste manual de hiperparámetros superó al grid search.

---

Citas

World Health Organization, Noncommunicable Diseases, Rehabilitation and Disability: Surveillance, Monitoring and Reporting (SMR). Geneva, Switzerland: WHO, 2022. https://www.who.int/publications/i/item/9789240047761

International Diabetes Federation, IDF Diabetes Atlas, 10th ed. Brussels, Belgium: IDF, 2021. https://diabetesatlas.org/

M. Romanello et al., “The 2022 report of the Lancet countdown on health and climate change,” The Lancet Diabetes & Endocrinology, vol 400, pp.1619–1654, 2022. https://doi.org/10.1016/S0140-6736(22)01540-9

Putula et al., “All-cause mortality and factors associated with it in Finnish patients with type 1 diabetes,” Journal of Diabetes and its Complications, vol 38, no. 12, 2024. https://doi.org/10.1016/j.jdiacomp.2024.108881

Mameli et al., “Lessons and gaps in the prediction and prevention of type 1 diabetes”, Pharmacological Research, vol 193, 2023. https://doi.org/10.1016/j.phrs.2023.106821

Gregory et al., “Global incidence, prevalence, and mortality of type 1 diabetes in 2021 with projection to 2040: a modelling study.” The Lancet Diabetes Endocrinology, vol 10, no. 10, pp. 741–760, 2022. https://doi.org/10.1016/S2213-8587(22)00218-2

N. Foster et al., “State of type 1 diabetes management and outcomes,” Diabetes Technology & Therapeutics, vol 21, no. 2, pp. 66–72, 2019. https://doi.org/10.1089/dia.2018.0384

S. Rich, “Type 1 Diabetes Genetics Consortium (T1DGC),” Version 7 [Dataset], NIDDK Central Repository, 2014. https://doi.org/10.58020/qrdt-eh40

TEDDY Study Group, “The Environmental Determinants of Diabetes in the Young (TEDDY) Study,” Annals of the New York Academy of Sciences, vol. 1150, pp. 1-13, 2008. https://doi.org/10.1196/annals.1447.062

A. Facchinetti, “Continuous Glucose Monitoring Sensors: Past, Present and Future Algorithmic Challenges”, Sensors, vol. 16, no. 12, 2016. https://doi.org/10.3390/s16122093

Borle, Neil C et al. “The challenge of predicting blood glucose concentration changes in patients with type I diabetes.” Health informatics journal vol. 27, no. 1, 2021. https://doi.org/10.1177/1460458220977584

M. S. Farooq et. Al., “Role of Internet of Things in diabetes healthcare: Network infrastructure, taxonomy, challenges, and security model,” Digital Health, vol. 9, 2023. https://doi.org/10.1177/20552076231179056

Sparacino et al., “Glucose concentration can be predicted ahead in time from continuous glucose monitoring sensor time-series,” IEEE Transactions on Biomedical Engineering, vol. 54, no. 5, pp. 931-937, May 2007. https://doi.org/10.1109/TBME.2006.889774

Jaloli, M., & Cescon, M. (2023). 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. https://doi.org/10.1177/19322968221092785

Contreras I, Vehi J., “Artificial Intelligence for Diabetes Management and Decision Support: Literature Review,” Journal of Medical Internet Research, vol 22, no.2, 2020. https://doi.org/10.2196/10775

Wonju S., et al. “A personalized blood glucose level prediction model with a fine-tuning strategy: A proof-of-concept study.” Computer methods and programs in biomedicine vol. 211, 2021. https://doi.org/10.1016/j.cmpb.2021.106424

Ghimire, Sarala et al. “Deep learning for blood glucose level prediction: How well do models generalize across different data sets?.” PloS one vol. 19, no. 9, pp. e0310801. 25 Sep. 2024. https://doi.org/10.1371/journal.pone.0310801

A. K. M and S. Baskar, “Internet of Things based Real-Time Hyperglycemia and Hypoglycemia Monitoring using Wearable Biosensors,” 2022 International Conference on Electronics and Renewable Systems (ICEARS), Tuticorin, India, 2022, pp. 487-493. https://doi.org/10.1109/ICEARS53579.2022.9751878

T. Zhu, K. Li, P. Herrero and P. Georgiou, “Personalized Blood Glucose Prediction for Type 1 Diabetes Using Evidential Deep Learning and Meta-Learning,” IEEE Transactions on Biomedical Engineering, vol. 70, no. 1, pp. 193-204, Jan. 2023. https://doi.org/10.1109/TBME.2022.3187703

D. Klonoff, et al. “Continuous glucose monitoring: A review of the technology and clinical use.” Diabetes research and clinical practice , vol. 133, 2017. https://doi.org/10.1016/j.diabres.2017.08.005

Khokhar, P. Bakhsh et al. “Advances in artificial intelligence for diabetes prediction: insights from a systematic literature review.” Artificial intelligence in medicineK, vol. 164, 2025. https://doi.org/10.1016/j.artmed.2025.103132

M. T. I Rimon et al. “Advancements in Insulin Pumps: A Comprehensive Exploration of Insulin Pump Systems, Technologies, and Future Directions.” Pharmaceutics, vol. 16, no. 7, Jul. 2024. https://doi.org/10.3390/pharmaceutics16070944

Dunstan, J., Villena, F., Hoyos, J. et al. “Predicting no-show appointments in a pediatric hospital in Chile using machine learning”. Health Care Manag Sci, vol. 26, pp. 313–329, 2023. https://doi.org/10.1007/s10729-022-09626-z

F. Ladislav and D. Vašata. 2024. “Predicting Blood Glucose Levels with LMU Recurrent Neural Networks: A Novel Computational Model”, In Artificial Intelligence in Medicine: 22nd International Conference, AIME 2024, Salt Lake City, UT, USA, July 9–12, 2024, Proceedings, Part I. Springer-Verlag, Berlin, Heidelberg, 117–127. https://doi.org/10.1007/978-3-031-66538-7_13

S. G. James, et. al., BrisT1D Blood Glucose Prediction Competition, 2024. Kaggle. https://kaggle.com/competitions/brist1d