Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session B: Wednesday, August 13, 1:00 – 4:00 pm, de Brug & E‑Hall

Systematic differences between brain organisation revealed by task and rest

Caroline Nettekoven¹, Ali Shahbazi, Bassel Arafat¹, Matea Skenderija, Ana Luisa Pinho, Jörn Diedrichsen²; ¹University of Western Ontario, ²Johns Hopkins University

Presenter: Caroline Nettekoven

Capturing brain organization using functional Magnetic Resonance Imaging (fMRI) is a fundamental goal in neuroscience. Brain parcellation, gradient, or connectivity models all rely on the voxel by voxel correlation matrices. However, it is unclear which behavioral paradigms and analysis approaches result in a correlation matrix estimate that would be most predictive of regional co-activation patterns across a variety of cognitive states. This study systematically compares resting-state and task-based fMRI data along with commonly used filtering and analysis methods. Using task-based and resting-state fMRI data from 17 subjects, we show that task-based data, particularly when subjected to a GLM, is biased by the particular task contrast being used. However, when including an increasing number of diverse tasks, this bias disappears. We show that such multi-task approach results in better prediction for novel tasks than resting-state data, while the latter is easier to acquire and more reliable. We conclude that task-based approaches, especially with diverse task sets, are better suited for understanding brain function across many behavioral states.

Topic Area: Brain Networks & Neural Dynamics

Extended Abstract: Full Text PDF