Multiple sclerosis (MS) is characterized by a progressive worsening of disability over time. As many regulatory-cleared disease-modifying treatments aiming to slow down this progression are now available, a clear need has arisen for a personalized and data-driven approach to treatment optimization in order to more efficiently slow down disease progression and eventually, progressive disability worsening. This strongly depends on the availability of biomarkers that can detect and differentiate between the different forms of disease worsening, and on predictive models to estimate the disease trajectory for each patient under certain treatment conditions. To this end, we here describe a multicenter, retrospective, observational study, aimed at setting up a harmonized database to allow the development, training, optimization, and validation of such novel biomarkers and AI-based decision models. Additionally, the data will be used to develop the tools required to better monitor this progression and to generate further insights on disease worsening and progression, patient prognosis, treatment decisions and responses, and patient profiles of patients with MS.