Abstract #3816

Regional optimization of physiological noise models improves functional connectivity measurements in resting-state fMRI at 7T

Sandro Nunes¹, Marta Bianciardi², Afonso Dias¹, Luís M. Silveira³, Lawrence L. Wald², and Patrícia Figueiredo¹

¹ISR-Lisboa/LARSyS and Department of Bioengineering, Instituto Superior Técnico – Universidade de Lisboa, Lisbon, Portugal, ²Department of Radiology, A.A. Martinos Center for Biomedical Imaging, MGH and Harvard Medical School, Boston, MA, United States, ³INESC-ID, Instituto Superior Técnico – Universidade de Lisboa, Lisbon, Portugal

We develop physiological noise models based on cardiac/respiratory recordings, with lag optimization at various levels of specificity (group, dataset, regional and voxel), where regional optimization was achieved by clustering the lagged BOLD responses across the brain. We compare these models, both in terms of the spurious variance explained in the data and the specificity and reproducibility of functional connectivity measurements from three well-known resting-state networks in rs-fMRI at 7T. Voxelwise models explain the most variance in the data; however, connectivity strength specificity and test-retest reproducibility indicate that optimization at the regional/cluster level produces the most accurate networks.

This abstract and the presentation materials are available to members only; a login is required.

Join Here