27041
Multi-Modal Data Platform for Autism Research
The growing quantity and diversity of data collections in autism research has created a demand for neuroinformatics platforms harmonizing multi-modal cohorts across large-scale distributed collaborations. Providing researchers with integrated analysis tools to leverage these combined data assets in a robust, sustainable and transparent manner is pivotal to the future of neurodevelopmental research. Data sharing initiatives such as the NIMH Data Archive (NDA) and research networks critically rely on secure online platforms to manage distributed datasets spanning diverse modalities of brain, behavior and genotypic data.
The ‘MNI Ecosystem’ (Das 2016, fig.1) consisting of LORIS (Das 2012), CBRAIN (Sherif 2014) and BrainBrowser (Sherif 2015) provides an open-source neuroinformatics environment designed to address these challenges, hosting large-scale projects such as the NIH MRI study of Pediatric Development (NIHPD), the IBIS autism network, and developing initiatives such as the Canadian National Autism Neuroinformatics Platform.
Objectives:
LORIS, CBRAIN and BrainBrowser are open-source software packages developed at the Montreal Neurological Institute (MNI), designed to integrate data capture, analysis and visualization within a unified context. This ‘MNI Ecosystem’ combines demographic, neuroimaging, clinical/behavioral, biobanking and summary genetic data via web-accessible databasing and a user-friendly online suite of data management tools. Transparency, flexibility and interoperability with other data systems are core principles of this neuroinformatic ecosystem.
Methods:
LORIS is a web-based data capture and management system designed for seamless interoperability with other clinical/behavioral, imaging and genomic data platforms. Modality-specific tools native to LORIS include secure online surveys, data uploaders, mobile-friendly data input and quality control workflows. Customizable instruments are programmed with complex scoring and multi-language support. BrainBrowser, a web-based visualization tool embedded within LORIS, provides an interactive 3D viewer for neuroimaging scans. Data capture systems for MRI, EEG and modalities such as eye-tracking can load data directly into LORIS, and summary genomic and biobanking data are integrated via extensible modules.
Integrated querying tools enable users to design, curate and save custom datasets, exporting them to processing suites and platforms such as CBRAIN, a high-performance computing portal. Once a dataset is exported to CBRAIN, software pipelines analyze and return post-processed results to be stored in the LORIS database.
Results:
In autism and neurodevelopmental research, LORIS serves as the core infrastructure for large-scale longitudinal projects such as the IBIS autism research network and the NIHPD database. These initiatives have contributed significantly to the interoperability and core data standard development of the NIMH Data Archive (NDA).
In the context of autism and related disorders, the Canadian National Autism Neuroinformatics Platform is deploying a combined gene-brain-behavior data platform in LORIS and CBRAIN. This initiative plans to leverage existing data collections with new multi-modal cohorts, to amplify the translational impact of these data assets across the Canadian research community.
Conclusions:
Combined in a single platform, LORIS, CBRAIN and BrainBrowser provide a robust neuroinformatics infrastructure for acquisition, validation and analysis of data collected from many domains. Customizable for large-scale multi-modal ASD cohorts, the MNI ecosystem platform enables powerful and transparent leveraging of resources and tools for cross-disciplinary computational approaches in autism and neurodevelopmental research.