How can neuroimaging facilitate the diagnosis and stratification of patients with psychosis?

Early diagnosis and treatment of patients with psychosis are associated with improved outcome in terms of future functioning, symptoms and treatment response. Identifying neuroimaging biomarkers for illness onset and treatment response would lead to immediate clinical benefits. In this review we discuss if neuroimaging may be utilised to diagnose patients with psychosis, predict those who will develop the illness in those at high risk, and stratify patients. State-of-the-art developments in the field are critically examined including multicentre studies, longitudinal designs, multimodal imaging and machine learning as well as some of the challenges in utilising future neuroimaging biomarkers in clinical trials. As many of these developments are already being applied in neuroimaging studies of Alzheimer's disease, we discuss what lessons have been learned from this field and how they may be applied to research in psychosis.