A dimensional psychiatry approach to eating disorders using machine learning techniques on structural and functional neuroimaging data from multicenter studies.
My aim is to test the ‚dimensional‘ approach to classify psychiatric disorders. At present, the approach used in psychiatry is ‚categorical‘: each disorder is identified by a set of specific symptoms and is treated as a separated entity in therapy and research. To the contrary, the dimensional approach aims to identify basic mechanisms that can be disrupted to a different degree in each individual and, when this happens, might concur to an increased risk of psychiatric illness, across the whole spectrum of psychiatric disorders. Specifically, I will search for biological markers of these disrupted mechanisms in terms of alterations in brain structure and functional connectivity. I will focus on eating disorders and apply machine learning techniques to multicentric datasets of MRI images.