Romy Lorenz | Probing the laminar circuitry of the prefrontal cortex during the attentional modulation of working memory

The neural mechanisms of selectively attending to information held in working memory are not well understood. This study aims to investigate the layer-specificity of information processing during the attentional modulation of working memory by using multivariate pattern analyses (MVPA) and psychophysiological interaction analyses (PPI). The pre-registered hypotheses formulated for this study were derived by combining insights from studies about selective attention in working memory and recent findings about the dissociable functional role of distinct cortical layers in working memory.

In our study, we will have 25 subjects perform an image-based retro-cue paradigm while undergoing high-resolution GE-BOLD fMRI with 2D-EPI readout at 7T (0.8 mm isotropic voxel resolution, 2s TR). In this paradigm, a retro-cue indicates whether the first or second set of images will be relevant for subjects’ subsequent behavioural response in the probe period and thus guides subjects' attention to one of the image sets to be maintained over the delay period.

Subjects are invited for two experimental visits, each lasting up to 90 minutes, in which they undergo anatomical and functional MRI. In the first visit, subjects will undergo anatomical imaging and functional localizer runs using low-resolution whole-brain fMRI. The location of peak activity from the analysis of the functional localiser run is then used to position the acquisition slab for the experimental runs acquired with high-resolution fMRI on the second visit. In the second visit, subjects will undergo six experimental runs in the scanner. We will focus our main analyses on regions in the left prefrontal cortex, as in our pilot study involving five subjects, we saw stronger decoding in the left compared to the right prefrontal cortex. The classifiers of interest we will compute are (1) attended faces vs. attended houses during the cued delay period, (2) unattended faces vs. unattended houses during the cued delay period, (3) left vs. right button press during the probe period.