Dynamic scaling of EEG fluctuations of patients with learning disorders based on artificial intelligence

This work models the dynamics of time series fluctuations of patients with learning disorders, specifically with reading-writing problems, applying fractal geometry, rough interface growth theory and Artificial Intelligence. From the EEG of children diagnosed with reading-writing problems, we obtain data of the brain activity of these children with which time series of fluctuations (standard deviations, υ (t,τ)) for each of the 19 channels distributed in different regions of the cerebral cortex. The self-affinity of the time series of fluctuations (treated as interfaces in motion) is characterized by the scaling behavior of the structure functions by one hand σ ∝ (δ_t)^ζ, with ζ as the local or roughness exponent and the other hand σ ∝ (τ)^β, with β as the fluctuation growth exponent. These findings guide us to propose the existence of a dynamic scaling behavior similar to that of Family-Vicsek for the kinetic roughening of a moving interface. In addition these findings are implemented in an Internet of Things (IoT) Network.