Catie Chang1, Biyu J. He2, Jeff H. Duyn3
1Advanced MRI section, NINDS , National Institutes of Health, Bethesda, MD, United States; 2NINDS, National Institutes of Health, Bethesda, MD, United States; 3Advanced MRI section, NINDS, National Institutes of Health, Bethesda, MD, United States
It was recently demonstrated that the BOLD signal exhibits scale-free temporal dynamics whose autocorrelation/long-range memory is reflected in the Hurst exponent (H). Here, we demonstrate that H and BOLD signal amplitude decrease as a function of increased cognitive effort (working memory load) across widespread regions of the brain. Furthermore, we characterize the sensitivity of H to the manner in which fMRI data are pre-processed. Importantly, removing noise due to physiological processes and head motion was found to enhance the significance of load-dependent changes in the Hurst exponent, increasing the likelihood that the observed dynamic changes are of neural origin.