A multidimensional ODE-based model of Alzheimer’s disease progression

Statistical setting

The proposed statistical model comprises three key components: an evolution model of an internal state (dynamical) variable, a likelihood function linking the observed biomarkers and cognitive tests with the state variable, and an instantaneous clinical prediction model.

Let (textbf{x}(t)) be a set of dynamical variables tracking the course of the disease, linked to some biomarkers, cognitive tests, or other measurements, and (textbf{y}) be a set of relevant covariates that may affect the disease evolution or manifestation, such as sex, age, therapeutic interventions, or genetics. The model relies on the assumption that the rate of change at a given point in time depends only on the current values of the dynamical variables and covariates. There are no assumptions about the number and observation times of each biomarker, neuropsychological and clinical…

Read more…