Fig 1.
Overview of data preprocessing, model development and evaluation.
Data was used from The Maastricht Study, an observational population-based cohort that comprises individuals with normal glucose metabolism (NGM), prediabetes, or type 2 diabetes (panel A). We included 851 individuals who underwent continuous glucose monitoring (CGM), most of whom simultaneously wore an accelerometer to assess physical activity (X, Y, and Z accelerations). Models developed with the long-short term memory (LSTM) architecture were trained in predicting glucose levels at 15- and 60-minute intervals with either CGM data only (1) or both CGM and accelerometer data (2) (panel B). Finally, model performance was evaluated by glucose profile analysis, performance metrics (root-mean-square error [RMSE]; Spearman’s correlation coefficient [rho]; proportions), and clinical error grids (panel C).
Table 1.
Baseline statistical and machine learning model comparison for predicting glucose values.
Table 2.
Participant characteristics of the CGM-based and CGM- and accelerometry-based glucose prediction study populations.
Table 3.
Overall performance in the main study population of CGM-based and CGM- and accelerometry-based machine learning models trained in predicting glucose values at time intervals of 15 and 60 minutes.
Fig 2.
Surveillance error grid evaluation of glucose prediction safety at time intervals of 15 and 60 minutes in the main study population.
Assessment of CGM-based glucose prediction safety in individuals with type 2 diabetes (n = 43) at 15 minutes (panel A) and 60 minutes (panel C). Assessment of CGM- and accelerometry-based glucose prediction safety in individuals with type 2 diabetes (n = 13) at 15 minutes (panel B) and 60 minutes (panel D). The risk score values translate to the following degrees of risk: 0–0.5, none; 0.5–1.0, slight (lower); 1.0–1.5, slight (higher); 1.5–2.0, moderate (lower); 2.0–2.5, moderate (higher); 2.5–3.0, great (lower); 3.0–3.5, great (higher); > 3.5 extreme [27].
Fig 3.
Surveillance error grid evaluation of glucose prediction safety at time intervals of 15, 30, and 60 minutes in individuals with type 1 diabetes.
Assessment of CGM-based glucose prediction safety in individuals with type 1 diabetes (n = 6) at 15 (panel A), 30 (panel B), and 60 minutes (panel C). The risk score values translate to the following degrees of risk: 0–0.5, none; 0.5–1.0, slight (lower); 1.0–1.5, slight (higher); 1.5–2.0, moderate (lower); 2.0–2.5, moderate (higher); 2.5–3.0, great (lower); 3.0–3.5, great (higher); > 3.5 extreme [27].