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

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...

Full description

Saved in:
Bibliographic Details
Main Authors: Antonio J. Rodriguez-Almeida, Carmelo Betancort, Ana M. Wägner, Gustavo M. Callico, Himar Fabelo, on behalf of the WARIFA Consortium
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/15/4647
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849406398313529344
author Antonio J. Rodriguez-Almeida
Carmelo Betancort
Ana M. Wägner
Gustavo M. Callico
Himar Fabelo
on behalf of the WARIFA Consortium
author_facet Antonio J. Rodriguez-Almeida
Carmelo Betancort
Ana M. Wägner
Gustavo M. Callico
Himar Fabelo
on behalf of the WARIFA Consortium
author_sort Antonio J. Rodriguez-Almeida
collection DOAJ
description 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.
format Article
id doaj-art-68e3c8eee4fb4c7783d3a586b6b03126
institution Kabale University
issn 1424-8220
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-68e3c8eee4fb4c7783d3a586b6b031262025-08-20T03:36:23ZengMDPI AGSensors1424-82202025-07-012515464710.3390/s25154647Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion TransformerAntonio J. Rodriguez-Almeida0Carmelo Betancort1Ana M. Wägner2Gustavo M. Callico3Himar Fabelo4on behalf of the WARIFA ConsortiumInstitute for Applied Microelectronics, University of Las Palmas de Gran Canaria, ULPGC, 35017 Las Palmas de Gran Canaria, SpainEndocrinology and Nutrition Department, Complejo Hospitalario Universitario Insular Materno-Infantil, CHUIMI, 35016 Las Palmas de Gran Canaria, SpainEndocrinology and Nutrition Department, Complejo Hospitalario Universitario Insular Materno-Infantil, CHUIMI, 35016 Las Palmas de Gran Canaria, SpainInstitute for Applied Microelectronics, University of Las Palmas de Gran Canaria, ULPGC, 35017 Las Palmas de Gran Canaria, SpainInstitute for Applied Microelectronics, University of Las Palmas de Gran Canaria, ULPGC, 35017 Las Palmas de Gran Canaria, SpainMore 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.https://www.mdpi.com/1424-8220/25/15/4647glucose predictiontransformersartificial intelligenceexplainable AIdeep learningpersonalized medicine
spellingShingle Antonio J. Rodriguez-Almeida
Carmelo Betancort
Ana M. Wägner
Gustavo M. Callico
Himar Fabelo
on behalf of the WARIFA Consortium
Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
Sensors
glucose prediction
transformers
artificial intelligence
explainable AI
deep learning
personalized medicine
title Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
title_full Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
title_fullStr Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
title_full_unstemmed Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
title_short Incorporating Uncertainty Estimation and Interpretability in Personalized Glucose Prediction Using the Temporal Fusion Transformer
title_sort incorporating uncertainty estimation and interpretability in personalized glucose prediction using the temporal fusion transformer
topic glucose prediction
transformers
artificial intelligence
explainable AI
deep learning
personalized medicine
url https://www.mdpi.com/1424-8220/25/15/4647
work_keys_str_mv AT antoniojrodriguezalmeida incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer
AT carmelobetancort incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer
AT anamwagner incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer
AT gustavomcallico incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer
AT himarfabelo incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer
AT onbehalfofthewarifaconsortium incorporatinguncertaintyestimationandinterpretabilityinpersonalizedglucosepredictionusingthetemporalfusiontransformer