PsyMRI is aimed at a multicenter-based utilization of already existing resting-state magnetic resonance data with a focus on major depressive disorder (MDD) and neurodegenerative diseases.
By sharing datasets between collaborating sites, problems concerning the acquisistion of suitable subjects can be overcome. Compared to healthy controls the measurement of patients via magnetic resonance imaging is complicated due to inclusion criteria. Moreover the number of patients in some uncommon psychiatric or neurodegenerative diseases is substantially limited in general.
Exchanging datasets that have been collected in different scanning sites allows for a higher sample size to be included in studies. Furthermore statistical validations of hypotheses and group effects as well as replications of results and meta-analyses can be carried out on larger datasets.
Aggregating existing datasets is therefore thought to advance sctientific knowledge and promote progress concerning diagnosis and treatment of psychiatric and neurodegenerative diseases like depression and dementia.
Membership is achieved by signing the agreement for the multicenter use of existing imaging data and compliance with the minimum requirements specified therein. Rules concerning authorship in publications based on the consortial dataset are also included in this agreement. Local datasets can only be included if they have been acquired with approval from the local ethics comittee and included subjects gave written informed consent.
Ownership of the data remains with the providing member of the Consortium. PsyMRI is not a free data sharing initiative as included datasets are only accessible for members of the Consortium.
The submission of rs-MRI data and accompanying metadata is managed via the web-based data storage system XNAT (Extensible Neuroimaging Archive Toolkit ) where compressed DICOM raw data and spreadsheet documents can be uploaded. Datasets are then stored on a physically separated server. For integration into the XNAT structure, datasets will be validated and subjected to an anonymization process (defacing). Optionally preprocessing procedures of the MRI data can be centralized and automated. Access to the files in the XNAT structure is then only granted to full members of the Consortium.