Grant: Canada Research Chairs

The availability of multimodal data for the study of the brain in health and disease has exploded in the past decade, with a number of international initiatives to make data more accessible to the scientific community. These large datasets have the potential to address the unreliability of underpowered neuroimaging studies, as well as open new avenues for large sample analyses, including exciting multimodal analyses. Additionally, research data are becoming increasingly more multimodal and longitudinal, with some studies acquiring, over multiple years, a wide range of data types including behavioral measures, neurocognition, neuroimaging, audio/video recordings, smartphone and wearable activity trackers data, and fluid biomarkers. The availability and richness of these datasets provides exciting opportunities in the quest to better understand brain structure and function, but also presents special challenges for researchers without significant computational expertise or resources to analyze this data. The overarching theme of the Chair is to build an ecosystem of free open source software and freely available derived data to leverage the power of these source datasets while making analyses more accessible to neuroscientists and clinicians. The program revolves around two themes: 1. The development of novel neuroimaging methods to study the human brain morphometry, microstructure, and structural connectivity from Magnetic Resonance Imaging. 2. The design and implementation of a flexible suite of software tools and data standards to manage, process and explore multimodal longitudinal data in a wide range of clinical contexts. Throughout the program, technology will be designed to support rigorous, reproducible, accessible, interoperable, and less biased analyses, in particular with regards to ethnoracial background. We anticipate several contributions in the scientific and societal domains. The developed software will include many novel methods such as new semi-automated and automated labeling algorithms, diffusion MRI QC procedures, and a flexible multimodal data management ecosystem. This ecosystem will be applicable to many clinical domains and help define new standards for data management and processing. Finally, our focus on reducing bias, increasing transparency and reproducibility will help make neuroscience studies more generalizable, and ultimately support the development of new treatments for individuals suffering from brain disorders.