Abstract
Visual working memory (VWM) recruits a widespread circuit, including regions within intraparietal sulcus, precentral sulcus, inferior frontal sulcus, anterior insula, and pre-supplementary motor area. This anatomical configuration, which emphasizes structures buried within sulci or fissures, presents challenges for the application of fNIRS, which measures near-infrared light reflected from brain structures, due to its low sensitivity to deeper tissue and low spatial resolution compared to fMRI. Some VWM regions such as those in and around intraparietal sulcus, precentral sulcus, and inferior frontal sulcus may be strongly observable in some individuals, but not others due to individual differences in cortical folding patterns and/or functional organization, while other regions on the medial surface or in the insula may be largely unobservable. Here, we perform detailed examination of the impact of anatomical and functional sources of variance on the application of fNIRS to study VWM. fMRI was used to map VWM functional brain circuitry in 17 healthy individuals, using a stair-cased 2-back paradigm reporting the spatial frequency of large Gabor patches. The fMRI activation map was projected to fNIRS channel space through Monte Carlo photon modeling as a simulated reference point. A subset of the fMRI subjects are participating in fNIRS experiments using the same paradigm in order to examine the degree to which actual individual differences reflect those predicted from the modeling. These investigations provide a basis for establishing a set of best practices for the application of fNIRS to the study of visual working memory.