Quantification of Load Dependent Brain Activity in Parametric N-Back Working Memory Tasks using Pseudo-continuous Arterial Spin Labeling (pCASL) Perfusion Imaging

Abstract

Brain activation and deactivation induced by N-back working memory tasks and their load effects have been extensively investigated using positron emission tomography (PET) and blood-oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI). However, the underlying mechanisms of BOLD fMRI are still not completely understood and PET imaging requires injection of radioactive tracers. In this study, a pseudo-continuous arterial spin labeling (pCASL) perfusion imaging technique was used to quantify cerebral blood flow (CBF), a well understood physiological index reflective of cerebral metabolism, in N-back working memory tasks. Using pCASL, we systematically investigated brain activation and deactivation induced by the N-back working memory tasks and further studied the load effects on brain activity based on quantitative CBF. Our data show increased CBF in the fronto-parietal cortices, thalamus, caudate, and cerebellar regions, and decreased CBF in the posterior cingulate cortex and medial prefrontal cortex, during the working memory tasks. Most of the activated/deactivated brain regions show an approximately linear relationship between CBF and task loads (0, 1, 2 and 3 back), although several regions show non-linear relationships (quadratic and cubic). The CBF-based spatial patterns of brain activation/deactivation and load effects from this study agree well with those obtained from BOLD fMRI and PET techniques. These results demonstrate the feasibility of ASL techniques to quantify human brain activity during high cognitive tasks, suggesting its potential application to assessing the mechanisms of cognitive deficits in neuropsychiatric and neurological disorders.

Keywords: N-back working memory; cerebral blood flow; load effects; pseudo-continuous arterial spin labeling (pCASL).

Figure 1

Paired t-maps between the three active working memory task conditions and the 0 back baseline vigilance task condition based on quantified CBF. p < 0.05 corrected. ‘L’ denotes the left hemisphere of the brain and ‘R’ denotes the right hemisphere.

Figure 2

Trend t-maps across the four working memory conditions based on quantified CBF. p < 0.05 corrected. ‘L’ denotes the left hemisphere of the brain and ‘R’ denotes the right hemisphere.

Figure 3

The average quantitative CBF values in the TNN and TPN ROIs (A), in the ROIs which exhibited linearly decreasing CBF (B), in the cortical ROIs which exhibited linearly increasing CBF (C), in the subcortical ROIs which exhibited linearly increasing CBF (D), and in the cerebellar ROIs which exhibited linearly increasing CBF (E) from 0b to 3b. Bars labeled in blue, cyan, pink and red demonstrate the CBF values under 0b, 1b, 2b and 3b, separately. TNN: task negative network; TPN: task positive network. PCC: posterior cingulate cortex; MPFC: medial prefrontal cortex; laTEMP: left anterior temporal gyrus; raTEMP: right anterior temporal gyrus. lIPL: left inferior parietal lobule. rIPL: right inferior parietal lobule. lMFG: left middle frontal gyrus; rMFG: right middle frontal gyrus; SMA: supplementary motor area; lpreCG: left precentral cingulate gyrus; ldlPFC: left dorsolateral prefrontal cortex; rdlPFC: right dorsolateral prefrontal cortex; laINS: left anterior insula; raINS: right anterior insula. lCAD: left caudate; rCAD: right CAD; lTHA: left thalamus; rTHA: right thalamus. lCEB/FG: left cerebellum/fusiform gyrus; rCEB/FG: right cerebellum/fusiform gyrus.

Figure 4

The average quantitative CBF values in the cortical ROIs exhibiting a quadratic trend of CBF (A) and in the cortical ROIs exhibiting a cubic trend of CBF (B) from 0b to 3b. Bars labeled in blue, cyan, pink and red demonstrate the CBF values under 0b, 1b, 2b and 3b, separately. lIPL: left inferior parietal lobule. rIPL: right inferior parietal lobule.