Modeling the inter-day variability of insulin sensitivity in patients with type 1 diabetes through Markov models

This paper focuses on a new technique which uses stochastic models (Markov models) to describe the inter-day variability of insulin sensitivity in patients with type 1 diabetes. Insulin sensitivity, the ability of body cells to respond to insulin to control blood glucose level, has to be tracked in patients with type 1 diabetes, and it plays a very important role in designing artificial pancreas. A subset of 100 subjects was studied, and their insulin sensitivity was estimated from CGM data at every breakfast, lunch and dinner for several days. Then Markov models were applied to describe the variability of the forsaid parameter. The study conducted aims to find a promising technique to study the variability of insulin sensitivity of patients with type 1 diabetes from CGM data, in a minimally invasive way.