RESUMO
Blood-oxygen-level dependent (BOLD) signals are widely used in functional magnetic resonance imaging (fMRI) as a proxy measure of brain activation. However, because these signals are blood-related, they are also influenced by other physiological processes. This is especially true in resting state fMRI, during which no experimental stimulation occurs. Previous studies have found that the amplitude of resting state BOLD is closely related to regional vascular density. In this study, we investigated how some of the temporal fluctuations of the BOLD signal also possibly relate to regional vascular density. We began by identifying the blood-bound systemic low-frequency oscillation (sLFO). We then assessed the distribution of all voxels based on their correlations with this sLFO. We found that sLFO signals are widely present in resting state BOLD signals and that the proportion of these sLFOs in each voxel correlates with different tissue types, which vary significantly in underlying vascular density. These results deepen our understanding of the BOLD signal and suggest new imaging biomarkers based on fMRI data, such as amplitude of low-frequency fluctuation (ALFF) and sLFO, a combination of both, for assessing vascular density.
RESUMO
Resting state functional connectivity analysis is a widely used method for mapping intrinsic functional organization of the brain. Global signal regression (GSR) is commonly employed for removing systemic global variance from resting state BOLD-fMRI data; however, recent studies have demonstrated that GSR may introduce spurious negative correlations within and between functional networks, calling into question the meaning of anticorrelations reported between some networks. In the present study, we propose that global signal from resting state fMRI is composed primarily of systemic low frequency oscillations (sLFOs) that propagate with cerebral blood circulation throughout the brain. We introduce a novel systemic noise removal strategy for resting state fMRI data, "dynamic global signal regression" (dGSR), which applies a voxel-specific optimal time delay to the global signal prior to regression from voxel-wise time series. We test our hypothesis on two functional systems that are suggested to be intrinsically organized into anticorrelated networks: the default mode network (DMN) and task positive network (TPN). We evaluate the efficacy of dGSR and compare its performance with the conventional "static" global regression (sGSR) method in terms of (i) explaining systemic variance in the data and (ii) enhancing specificity and sensitivity of functional connectivity measures. dGSR increases the amount of BOLD signal variance being modeled and removed relative to sGSR while reducing spurious negative correlations introduced in reference regions by sGSR, and attenuating inflated positive connectivity measures. We conclude that incorporating time delay information for sLFOs into global noise removal strategies is of crucial importance for optimal noise removal from resting state functional connectivity maps.
RESUMO
Functional near infrared spectroscopy (fNIRS) is a promising method for monitoring cerebral hemodynamics with a wide range of clinical applications. fNIRS signals are contaminated with systemic physiological interferences from both the brain and superficial tissues, resulting in a poor estimation of the task related neuronal activation. In this study, we use the anatomical resolution of functional magnetic resonance imaging (fMRI) to extract scalp and brain vascular signals separately and construct an optically weighted spatial average of the fMRI blood oxygen level-dependent (BOLD) signal for characterizing the scalp signal contribution to fNIRS measurements. We introduce an extended superficial signal regression (ESSR) method for canceling physiology-based systemic interference where the effects of cerebral and superficial systemic interference are treated separately. We apply and validate our method on the optically weighted BOLD signals, which are obtained by projecting the fMRI image onto optical measurement space by use of the optical forward problem. The performance of ESSR method in removing physiological artifacts is compared to i) a global signal regression (GSR) method and ii) a superficial signal regression (SSR) method. The retrieved signals from each method are compared with the neural signals that represent the 'ground truth' brain activation cleaned from cerebral systemic fluctuations. We report significant improvements in the recovery of task induced neural activation with the ESSR method when compared to the other two methods as reflected in the Pearson R(2) coefficient and mean square error (MSE) metrics (two tailed paired t-tests, p<0.05). The signal quality is enhanced most when ESSR method is applied with higher spatial localization, lower inter-trial variability, a clear canonical waveform and higher contrast-to-noise (CNR) improvement (60%). Our findings suggest that, during a cognitive task i) superficial scalp signal contribution to fNIRS signals varies significantly among different regions on the forehead and ii) using an average scalp measurement together with a local measure of superficial hemodynamics better accounts for the systemic interference inherent in the brain as well as superficial scalp tissue. We conclude that maximizing the overlap between the optical pathlength of superficial and deeper penetration measurements is of crucial importance for accurate recovery of the evoked hemodynamic response in fNIRS recordings.
