[Paper Review] GLYFE: Review and Benchmark of Personalized Glucose Predictive Models in Type-1 Diabetes

Maxime De Bois, Mehdi Ammi+1 more|arXiv (Cornell University)|Jun 29, 2020

Diabetes Management and Research114 references36 citations

TL;DR

The paper provides a review and benchmark of personalized glucose predictive models for Type-1 diabetes.

ABSTRACT

Due to the sensitive nature of diabetes-related data, preventing them from being shared between studies, progress in the field of glucose prediction is hard to assess. To address this issue, we present GLYFE (GLYcemia Forecasting Evaluation), a benchmark of machine-learning-based glucose-predictive models. To ensure the reproducibility of the results and the usability of the benchmark in the future, we provide extensive details about the data flow. Two datasets are used, the first comprising 10 in-silico adults from the UVA/Padova Type 1 Diabetes Metabolic Simulator (T1DMS) and the second being made of 6 real type-1 diabetic patients coming from the OhioT1DM dataset. The predictive models are personalized to the patient and evaluated on 3 different prediction horizons (30, 60, and 120 minutes) with metrics assessing their accuracy and clinical acceptability. The results of nine different models coming from the glucose-prediction literature are presented. First, they show that standard autoregressive linear models are outclassed by kernel-based non-linear ones and neural networks. In particular, the support vector regression model stands out, being at the same time one of the most accurate and clinically acceptable model. Finally, the relative performances of the models are the same for both datasets. This shows that, even though data simulated by T1DMS are not fully representative of real-world data, they can be used to assess the forecasting ability of the glucose-predictive models. Those results serve as a basis of comparison for future studies. In a field where data are hard to obtain, and where the comparison of results from different studies is often irrelevant, GLYFE gives the opportunity of gathering researchers around a standardized common environment.

Motivation & Objective

• Identify and categorize existing personalized glucose predictive models for Type-1 diabetes.
• Assess the modeling approaches used for glucose prediction in Type-1 diabetes.
• Evaluate performance measurement and benchmarking practices across studies.
• Highlight gaps, challenges, and opportunities for standardized benchmarking.

Proposed method

• Conduct a structured literature review of personalized glucose predictive models for Type-1 diabetes.
• Propose or apply benchmarking criteria to compare modeling approaches.
• Analyze methods and performance metrics used in cited studies.
• Discuss methodological strengths, limitations, and reproducibility considerations.

Experimental results

Research questions

• RQ1 What are the main modeling approaches used for personalized glucose prediction in Type-1 diabetes?
• RQ2 How do different models perform across datasets, settings, and evaluation metrics?
• RQ3 What are the common limitations and gaps in current benchmarking practices?
• RQ4 What recommendations can improve fair and reproducible benchmarking of glucose predictive models?

Key findings

• A variety of modeling approaches are surveyed for personalized glucose prediction in Type-1 diabetes.
• Model performance varies across datasets and evaluation metrics.
• Benchmarking practices reveal inconsistencies and gaps in reproducibility and comparability.
• Common challenges include data quality, personalization, and feature engineering considerations.
• The work identifies opportunities for standardized benchmarking frameworks and future research directions.