About this Event
Abstract: Shared access to well-curated large, globally inclusive datasets is critical for the development, validation, and translation of novel biomarkers and prognostic tools in the clinic. Nonetheless, geographical separation and data governance policies often pose technical and privacy issues that complicate and constrain en masse integration and analysis of multicentric datasets. This talk will show progress in learning and validating brain biomarkers and prognostic models of age related neurodegenerative diseases with decentralized datasets. It will discuss the associated hurdles and make a case for investment in neuroinformatic infrastructure for managing studies with heterogeneous computational setups and governance policies. It will feature lightweight neuroinformatic tools - Nipoppy and Neurobagel, developed to address the needs of changing neuroscience research prompted by shift towards decentralized paradigms. The Nipoppy framework facilitates standardization of study-level data workflows across sites, allowing global researchers to share consistent, derived datasets that are often subject to less restrictive policy. The Neurobagel tools enable semantic harmonization, and privacy-aware data discovery crucial for mega-analyses and replication experiments. The talk will conclude with a federated analysis case-study performed in a multicentric ENIGMA-PD consortium, followed with a discussion on community building efforts to promote replicable, inclusive research at scale to improve our understanding of brain disorders and clinical outcomes.
Bio: Nikhil is a computational neuroscientist with industry and academic research experience in neuroimaging and machine-learning. He's interested in development of robust biomarkers and machine-learning models for prognosis in neurodegenerative diseases. Currently he focus on development of neuroinformatic tools and implementation of reproducible computational workflows to assess and improve replication of imaging biomarkers and ML models. Concurrently, he also involved in open-science and data harmonization projects relating to curation of diverse, open datasets representative of global populations.
At present, he's working with Dr. Jean-Baptiste Poline in the ORIGAMI lab at McGill University. He has completed his PhD thesis on prognostic applications for Alzheimer’s disease using MR imaging and machine-learning techniques under Dr. Mallar Chakravarty at the University of Toronto. Subsequently, he's worked as a researcher on a collaborative project between McGill and UMass universities assessing reproducibility of computational pipelines estimating cortical surfaces measures. He then went on to work at the Allen Institute, where he developed segmentation techniques and network spread models for quantifying cellular pathologies of Alzheimer’s disease in high-resolution microscopy data.
Within the larger research community, he's involved in multiple open-science and sustainable-neuroimaging initiatives. He develops and teaches several “open” workshops (e.g. MAIN and educational courses. I am also a current treasurer for the Sustainability & Environmental Action SIG within the Organization for Human Brain Mapping (OHBM) working on promoting sustainable research practices and conferencing.
Unable to attend in person? Join us virtually
For more information, visit The Neuro's website
Event Venue & Nearby Stays
Montreal Neurological Institute – Hospital, 3801 Rue University, Montréal, Canada
USD 0.00
