Standardized reporting of neuroimaging results with NIDM in SPM, FSL and AFNI

International audience; The growing awareness of underpowered and irreproducible research [1] has highlighted the necessity for data sharing in neuroimaging. Encouragingly, in the past ten years, the number of publicly available datasets has greatly increased, principally thanks to consortiums dedicated to data sharing such as the Human Connectome Project [4], the International Data sharing Initiative [5] or more recently the Consortium for Reliability and Reproducibility [3] and study forrest [10]. But while there is an increasing interest in sharing raw or pre-processed data (e.g. after motion correction, registration to a template), statistical images supporting neuroimaging publications are rarely made available. Furthermore, despite the availability of guidelines [7], ambiguous or incomplete methodological reporting is still commonplace [2] making image-based meta-analysis and reproducibility studies particularly challenging. In [6], we introduced the Neuroimaging Data Model (NIDM), a domain-specific extension of the recently-approved W3C recommendation, PROV-DM [9]. Here, we introduce NIDM-Results, a standardised representation of “mass univariate” neuroimaging results for the three major software analysis packages: SPM, FSL, and AFNI.