Chaoticity as a Proxy of Glucose Dynamics

Abstract

Diabetes mellitus is a widespread disease that affects many individuals throughout the world. Through an interstitial sensor, blood glucose levels can be recorded and stored for later analysis. In this work, the acquired time series was analyzed, reconstructing the attractor of the dynamic system, stability and instability zones were also established by means of clausterization techniques, and the Lypaunov exponent was calculated, giving evidence of being a deterministic chaotic system. On the other hand, the entropy (H) and the statistical complexity (C) have been calculated with the aim of formulating indicators to characterize diabetes and its dynamic evolution. The results have shown that the dynamic behavior of normal and diabetic patients is in clearly differentiated regions of the H×C plane.