Computational protocol: Impact of aberrant cerebral perfusion on resting-state functional MRI: A preliminary investigation of Moyamoya disease

Similar protocols

Protocol publication

[…] Rs-fMRI data were preprocessed using a MATLAB toolbox, the Data Processing Assistant for Resting-State (DPARSF) V2.1 Basic Edition,[] which was based on statistical parametric mapping software functions (SPM8; http://www.fil.ion.ucl.ac.uk/spm/software/spm8) and REST software (http://restfmri.net/forum/index.php). For each participant, the first five time-points were discarded to avoid transient signal changes before magnetization reached a steady-state and to allow subjects to become accustomed to the fMRI scanning noise. The rs-fMRI data were corrected for the acquisition delay among the slices. It has been previously been showed in various studies that spurious correlations appear due to head motion artifact.[] Therefore, we preferred the Friston 24-parameter model (including 6 head motion parameters of the current time point and the preceding time point, and the 12 corresponding square items)[] for motion correction, because higher-order regression models may perform better at the individual-subject level.[] Furthermore, a threshold for head motion during the present study was established at 3 mm and 3 degrees. Following motion correction, all data were spatially normalized to the Montreal Neurological Institute (MNI) template and resampled to 3 × 3 × 3 mm3. Then, the processed images were spatially smoothed with a 6-mm full-width half-maximum Gaussian kernel, followed by linear detrending to remove any residual drift. Twenty-seven nuisance signals were removed from the time series of each voxel via linear regression, including the global signal, white matter signal, cerebrospinal fluid signal and twenty-four head motion parameters. This regression procedure was utilized to reduce spurious variance unlikely to reflect neural activity.The procedure for calculating ALFF was based on previous studies.[,] The filtered time series were transformed to the frequency domain using fast Fourier transform (FFT) (parameters: taper percent = 0, FFT length = shortest). Since the power of a given frequency was proportional to the square of the amplitude of this frequency component, the square root was calculated by the power spectrum and averaged across 0.01–0.08 Hz at each voxel. For normalization, the ALFF of each voxel was divided by the global mean ALFF values for each subject. For comparison, the blood supply area of regions of interest (ROIs) of ALFF images should be consistent with those of PWI which were supplied by the middle cerebral artery (see below). In addition, in order to minimize the potential bias introduced by the posterior circulation, the bilateral frontal lobes were finally chosen as ROIs of ALFF images. The left and right frontal masks, which were extracted from WFU PickAtlas toolbox 3.0 (http://www.ansir.wfubmc.edu), were overlaid onto the global mean ALFF maps by using MRIcron software, respectively, and eventually the bilateral frontal lobe ALFF values of each patient were obtained (). […]

Pipeline specifications

Software tools DPABI, SPM, MRIcron
Applications Magnetic resonance imaging, Functional magnetic resonance imaging
Organisms Homo sapiens
Diseases Brain Diseases, Moyamoya Disease
Chemicals Oxygen