EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer

Antonio J Rodriguez-Almeida iD, Carmelo Betancort, Ana M Wägner iD, Gustavo M Callico iD, Himar Fabelo iD, On Behalf Of The Warifa Consortium

Sensors · 26.07.2025 · 25 (15) · 4647

DOI: 10.3390/s25154647 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 26.07.2025. Verfasst von Antonio J Rodriguez-Almeida und 5 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Applied Microelectronics, University of Las Palmas de Gran Canaria, ULPGC, 35017 Las Palmas de Gran Canaria, Spain. Hauptthemen laut MEDLINE: Blood Glucose, Diabetes Mellitus, Type 1, Blood Glucose Self-Monitoring. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 63 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Agencia Canaria de Investigación, Innovación y Sociedad de la Información, European Union, European Social Fund (FSE).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

More than 14% of the world’s population suffered from diabetes mellitus in 2022. This metabolic condition is defined by increased blood glucose concentrations. Among the different types of diabetes, type 1 diabetes, caused by a lack of insulin secretion, is particularly challenging to treat. In this regard, automatic glucose level estimation implements Continuous Glucose Monitoring (CGM) devices, showing positive therapeutic outcomes. AI-based glucose prediction has commonly followed a deterministic approach, usually with a lack of interpretability. Therefore, these AI-based methods do not provide enough information in critical decision-making scenarios, like in the medical field. This work intends to provide accurate, interpretable, and personalized glucose prediction using the Temporal Fusion Transformer (TFT), and also includes an uncertainty estimation. The TFT was trained using two databases, an in-house-collected dataset and the OhioT1DM dataset, commonly used for glucose forecasting benchmarking. For both datasets, the set of input features to train the model was varied to assess their impact on model interpretability and prediction performance. Models were evaluated using common prediction metrics, diabetes-specific metrics, uncertainty estimation, and interpretability of the model, including feature importance and attention. The obtained results showed that TFT outperforms existing methods in terms of RMSE by at least 13% for both datasets.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Blood GlucoseDiabetes Mellitus, Type 1Blood Glucose Self-MonitoringHumansUncertaintyAlgorithms

Stichwörter der Autor:innen

artificial intelligencedeep learningexplainable AIglucose predictionmHealthpersonalized medicinetransformers

Substanzen

Blood Glucose

Autor:innen und Institutionen

  • Institute for Applied Microelectronics, University of Las Palmas de Gran Canaria, ULPGC, 35017 Las Palmas de Gran Canaria, Spain.
  • Endocrinology and Nutrition Department, Complejo Hospitalario Universitario Insular Materno-Infantil, CHUIMI, 35016 Las Palmas de Gran Canaria, Spain.
  • Instituto de Investigaciones Biomédicas y Sanitarias, University of de Las Palmas de Gran Canaria, ULPGC, 35016 Las Palmas de Gran Canaria, Spain.
  • Fundación Canaria Instituto de Investigación Sanitaria de Canarias, FIISC, 35019 Las Palmas de Gran Canaria, Spain.
  • Research Unit, Hospital Universitario de Gran Canaria Dr. Negrin, 35010 Las Palmas de Gran Canaria, Spain.

ORCID-Profile: Antonio J Rodriguez-Almeida, Ana M Wägner, Gustavo M Callico, Himar Fabelo

Förderung

  • Agencia Canaria de Investigación, Innovación y Sociedad de la Información
  • European Union
  • European Social Fund (FSE)

Zitieren

APA

Rodriguez-Almeida, A. J., Betancort, C., Wägner, A. M., Callico, G. M., Fabelo, H., & On Behalf Of The Warifa Consortium (2025). Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer. Sensors, 25(15), 4647. https://doi.org/10.3390/s25154647

Vancouver

Rodriguez-Almeida AJ, Betancort C, Wägner AM, Callico GM, Fabelo H, On Behalf Of The Warifa Consortium. Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer. Sensors. 2025;25(15):4647. doi:10.3390/s25154647

BibTeX
@article{RodriguezAlmeida2025,
  author = {Rodriguez-Almeida, Antonio J and Betancort, Carmelo and Wägner, Ana M and Callico, Gustavo M and Fabelo, Himar and On Behalf Of The Warifa Consortium},
  title = {Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer},
  journal = {Sensors},
  year = {2025},
  volume = {25},
  number = {15},
  pages = {4647},
  doi = {10.3390/s25154647}
}
Kostenlos starten →