PT - JOURNAL ARTICLE AU - Cauda, Franco AU - Costa, Tommaso AU - Fava, Luciamo AU - Palermo, Sara AU - Bianco, Francesca AU - Duca, Sergio AU - Geminiani, Giuliano AU - Tatu, Karina AU - Keller, Roberto TI - Predictability of Autism, Schizophrenic and Obsessive Spectra Diagnosis: Toward a Damage Network Approach AID - 10.1101/014563 DP - 2015 Jan 01 TA - bioRxiv PG - 014563 4099 - http://biorxiv.org/content/early/2015/01/29/014563.short 4100 - http://biorxiv.org/content/early/2015/01/29/014563.full AB - Schizophrenia, obsessive-compulsive and autistic disorders are traditionally considered as three separate psychiatric conditions each with specific symptoms and pattern of brain alterations. This view can be challenged since these three conditions have the same neurobiological origin, stemming from a common root of a unique neurodevelopmental tree.The aim of this meta-analytic study was to determine, from a neuroimaging perspective, whether i) white matter and gray matter alterations are specific for the three different spectrum disorders, and the nosographical differentiation of three spectra is supported by different patterns of brain alterations. ii) it might be possible to define new spectra starting from specific brain damage. iii) it is possible to detect a “brain damage network” (a connecting link between the damaged areas that relates areas constantly involved in the disorder).Three main findings emerged from our meta-analysis: The three psychiatric spectra do not appear to have their own specific damage.It is possible to define two new damage clusters. The first includes substantial parts of the salience network, and the second is more closely linked to the auditory-visual, auditory and visual somatic areas.It is possible to define a "Damage Network" and to infer a hierarchy of brain substrates in the pattern of propagation of the damage.These results suggest the presence of a common pattern of damage in the three pathologies plus a series of variable alterations that, rather than support the sub-division into three spectra, highlight a two-cluster parcellation with an input-output and more cognitive clusters.