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

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Multi-task batteries for individual brain mapping: Experimental design and implementation

Bassel Arafat¹, Caroline Nettekoven¹, Jörn Diedrichsen²; ¹University of Western Ontario, ²Johns Hopkins University

Presenter: Jörn Diedrichsen

While the characterization of individual human brain organization with functional magnetic resonance imaging (fMRI) has in the past relied heavily on resting-state data, it has been shown that a more powerful identification of functional brain organization can be achieved with batteries including a broad set of tasks. Following practical considerations, these multi-task datasets are often designed such that each imaging run includes only a small number of similar tasks or conditions, such that most task-task contrasts have to be made across fMRI runs. Here we show that a design in which all tasks are measured repeatedly within the same imaging run is statistically superior both for estimating tasks-rest contrasts, as well as any task-to-task contrast. An interspersed multi-task design leads to more predictive brain parcellations and connectivity models, even though the design requires participants to constantly switch between tasks. We present a flexible Python toolbox that implements 20+ common tasks with this design, and that automatizes the generation of multi-task batteries for fMRI experimentation. Furthermore, we provide a framework for sharing and integrating pre-processed data across a growing number of multi-task datasets.

Topic Area: Brain Networks & Neural Dynamics

Extended Abstract: Full Text PDF