Brad Sutton, Technical Director of the Biomedical Imaging Center and Abel Bliss Faculty Scholar in the College of Engineering, will lecture on “Ultrafast functional MRI: A tool for examining spurious correlations in fMRI connectivity.”
Abstract:
Functional MRI can non-invasively probe the internal communications in the brain, allowing us to map out network connectivity by looking for temporally correlated functional signals from different brain regions. However, physiological noise (cardiac pulsations and respiratory motion) along with motion of the head can cause significant signals in fMRI and lead to spurious correlations, potentially masking smaller functional signals.
Making matters worse, it has been seen that groups of people can have similar and repeatable characteristics in their motion and physiological signals during the fMRI scan, leading to reproducible group effects in functional connectivity. Finally, with widespread use of standardized imaging protocols, there may even be reproducibility in confounding correlations due to the way in which fMRI data is sampled.
In this talk, Brad will provide a very early look at a new technique to sample fMRI data with a very fast, whole brain 3D acquisition. Brad will explain how the technique works and show some example data that demonstrates a potential use to separate the sampling of physiological signals to reduce their contributions to the connectivity estimate.
