Skip to content
DraconiticResearch
Search Sources API
197,768 studies31 sourcesUpdated 34 min ago
Peer-reviewedCC BY 4.0PLOS Digital Health

Deep learning for blood glucose prediction: Reproducibility challenges and factors affecting differential performance

  • Baiying Lu1
  • Biratal Wagle2
  • Zhaohui Liang3
  • Yanjun Cui
  • Temiloluwa Prioleau
  1. 1Department of Computer Science, Dartmouth College, Hanover, New Hampshire, United States of America
  2. 2Department of Quantitative Biomedical Sciences, Dartmouth College, Hanover, New Hampshire, United States of America
  3. 3Department of Computer Science, Emory University, Atlanta, GeorgiaUnited States of America
Read full textPDFDOICite

Abstract

Blood glucose prediction is a critical component of next-generation diabetes technologies, such as artificial pancreas systems, where reliable performance is essential for safety and effectiveness. Although deep learning methods have achieved promising advances in this area, a critical gap remains in understanding the reproducibility and generalizability of these methods. To contextualize the gap, this study reviewed 67 recent papers that proposed a deep learning method for glucose prediction to identify key reproducibility challenges. Next, we adopted a standardized framework, encompassing technical, statistical, and conceptual reproducibility evaluations, to experimentally assess the reproducibility of eight representative deep learning methods. To achieve this, we reimplemented and evaluated these eight deep learning methods using over 1.36 million continuous glucose monitoring samples (5,061 days) from 128 individuals with type 1 diabetes across three public datasets: OhioT1DM, DiaTrend, and T1DEXI. We found that even though these models demonstrated good technical and statistical reproducibility, their conceptual reproducibility—the ability to generalize to datasets with different diabetes management patterns—was limited. Further analyses revealed that each model’s overall prediction performance was strongly influenced by individual glycemic control, with higher prediction errors observed among participants with lower time with blood glucose in the target range (70–180 mg/dL). This study identified key reproducibility challenges associated with current blood glucose prediction methods within type 1 diabetes populations, highlighting the need for increased transparency, dataset diversity, standardized evaluation practices, and code accessibility to ensure reproducible and reliable models for blood glucose prediction. Author summary: Managing type 1 diabetes requires individuals to frequently monitor their blood glucose levels and constantly consider how decisions related to food, activity, insulin use, and more affect their blood glucose levels. Accurate blood glucose prediction is critical to inform diabetes management strategies and mitigate adverse events. Consequently, many researchers have explored deep learning models for predicting future blood glucose levels, with the goal of providing earlier warnings of adverse events and personalized guidance for diabetes management. However, despite the progress, little effort exists on evaluating the reproducibility of published models for blood glucose prediction. To address this gap, we evaluated the reproducibility of eight state-of-the-art deep learning models using three publicly available datasets from distinct populations with type 1 diabetes. This evaluation allowed us to examine not only whether the reported results could be accurately replicated but also whether these models could generalize to new groups of individuals with different blood glucose management strategies and outcomes. Findings from this study revealed several barriers that limit model reproducibility. Furthermore, we found that the performance of many models was significantly affected by the heterogeneity of glucose management across individuals. Our work highlights the importance of transparent research practices and evaluation across diverse datasets to ensure that future glucose prediction models are reliable and equitable for all people living with diabetes.

Related

  1. Partial Identification of Survival Prediction Performance under Dependent Right Censoring
    Yuji Komiyama, Yasumasa Matsuda
    arXivPreprintarXiv licenceFull textStatisticsarXiv
  2. Simulation-Supervised Foundation Models for Retention Time Prediction in High-Performance Liquid Chromatography beyond Experimental Data Coverage
    Stephen Wu, Yufeng Han, Yasuhiro Mito +5
    arXivPreprintCC BY-NC-ND 4.0Full textPhysical chemistryStatisticsarXiv
  3. Prediction of the performance of industrial gas turbine based on different environmental conditions
    Fizza Ghulam Nabi, Muhammad Usman, Muhammad Ehtisham +4
    PLOS OnePeer-reviewedCC BY 4.0Full textCell biologyNeurosciencePLOS
  4. Evaluation of clinical prediction models: Decomposing performance uncertainty into model development and validation components
    Mohsen Sadatsafavi, Paul Gustafson, Richard D. Riley
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  5. Improving sport horse selection using machine learning: Early prediction of elite-level show jumping performance
    Marco Zanchi, Clara Bordin, Michela Ablondi +4
    PLOS OnePeer-reviewedCC BY 4.0Full textZoology & animal behaviourCombinatorics & discrete mathematicsPLOS

Cite

BibTeXRIS
BibTeX
@article{lu2026deep,
  title = {{Deep learning for blood glucose prediction: Reproducibility challenges and factors affecting differential performance}},
  author = {Baiying Lu and Biratal Wagle and Zhaohui Liang and Yanjun Cui and Temiloluwa Prioleau},
  journal = {PLOS Digital Health},
  year = {2026},
  month = sep,
  doi = {10.1371/journal.pdig.0001633},
  url = {https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001633},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Deep learning for blood glucose prediction: Reproducibility challenges and factors affecting differential performance
AU  - Baiying Lu
AU  - Biratal Wagle
AU  - Zhaohui Liang
AU  - Yanjun Cui
AU  - Temiloluwa Prioleau
PY  - 2026
DA  - 2026/09/03
JO  - PLOS Digital Health
DO  - 10.1371/journal.pdig.0001633
UR  - https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001633
AB  - Blood glucose prediction is a critical component of next-generation diabetes technologies, such as artificial pancreas systems, where reliable performance is essential for safety and effectiveness. Although deep learning methods have achieved promising advances in this area, a critical gap remains in understanding the reproducibility and generalizability of these methods. To contextualize the gap, this study reviewed 67 recent papers that proposed a deep learning method for glucose prediction to identify key reproducibility challenges. Next, we adopted a standardized framework, encompassing technical, statistical, and conceptual reproducibility evaluations, to experimentally assess the reproducibility of eight representative deep learning methods. To achieve this, we reimplemented and evaluated these eight deep learning methods using over 1.36 million continuous glucose monitoring samples (5,061 days) from 128 individuals with type 1 diabetes across three public datasets: OhioT1DM, DiaTrend, and T1DEXI. We found that even though these models demonstrated good technical and statistical reproducibility, their conceptual reproducibility—the ability to generalize to datasets with different diabetes management patterns—was limited. Further analyses revealed that each model’s overall prediction performance was strongly influenced by individual glycemic control, with higher prediction errors observed among participants with lower time with blood glucose in the target range (70–180 mg/dL). This study identified key reproducibility challenges associated with current blood glucose prediction methods within type 1 diabetes populations, highlighting the need for increased transparency, dataset diversity, standardized evaluation practices, and code accessibility to ensure reproducible and reliable models for blood glucose prediction. Author summary: Managing type 1 diabetes requires individuals to frequently monitor their blood glucose levels and constantly consider how decisions related to food, activity, insulin use, and more affect their blood glucose levels. Accurate blood glucose prediction is critical to inform diabetes management strategies and mitigate adverse events. Consequently, many researchers have explored deep learning models for predicting future blood glucose levels, with the goal of providing earlier warnings of adverse events and personalized guidance for diabetes management. However, despite the progress, little effort exists on evaluating the reproducibility of published models for blood glucose prediction. To address this gap, we evaluated the reproducibility of eight state-of-the-art deep learning models using three publicly available datasets from distinct populations with type 1 diabetes. This evaluation allowed us to examine not only whether the reported results could be accurately replicated but also whether these models could generalize to new groups of individuals with different blood glucose management strategies and outcomes. Findings from this study revealed several barriers that limit model reproducibility. Furthermore, we found that the performance of many models was significantly affected by the heterogeneity of glucose management across individuals. Our work highlights the importance of transparent research practices and evaluation across diverse datasets to ensure that future glucose prediction models are reliable and equitable for all people living with diabetes.
SN  - 2767-3170
ER  -