Abstrakt
In this paper, we describe a new statistical approach to estimate blood glucose concentration along time during endurance sports based on measurements of glucose concentration in subcutaneous interstitial tissue. The final goal is the monitoring of glucose concentration in blood to maximize performance in endurance sports.
Blood glucose concentration control during and after aerobic physical activity could also be useful to reduce the risk of hypoglycemia in type 1 diabetes mellitus subjects. By means of a low invasive technology known as "continuous glucose monitoring", glucose concentration in subcutaneous interstitial tissue can now be measured every five minutes.
However, it can be expressed as function of blood glucose concentration along time by means of a convolution integral equation. In the training phase of the proposed approach, based on measurements of glucose concentration in both artery and subcutaneous interstitial tissue during physical activity, the parameters of the convolution kernel are estimated.
Then, given a new subject performing aerobic physical activity, a deconvolution problem is solved to estimate glucose concentration in blood from continuous glucose monitoring measurements.