Fig. 4 | Scientific Reports

Fig. 4

From: Evaluating HbA1c-to-average glucose conversion with patient-specific kinetic models for diverse populations

Fig. 4

Data sectioning and structure. (A) Defining structure and nomenclature. Green: The training segment used for computing the personalization factors (PFs) is delimited by the start of continuous glucose monitoring (CGM) data and a contiguous HbA1c measurement. Blue: the testing segment, comprised of CGM data and a contiguous HbA1c used to compute the estimated average glucose (eAG) and personalized AG (pAGs), which are then compared to the average glucose from the CGM (CGM-AG). In some cases, there is a data gap between training and testing, defined as the gap segment (gray). (B) Data partitioning: Datasets T1.1, T1.3, and T1.4 have three HbA1c measurements every ∽3 months. We use different segments of data as the training segment.

Back to article page