Computational protocol: Cortical Activation Patterns of Bodily Attention triggered by Acupuncture Stimulation

Similar protocols

Protocol publication

[…] The functional imaging data were preprocessed and analyzed using Nipype, which is a pipeline platform that joins software packages, including SPM8 (Welcome Department of Cognitive Neurology, London, United Kingdom, http://www.fil.ion.ucl.ac.uk/spm), Freesurfer (http://surfer.nmr. mgh.harvard.edu), and FSL (http://www.fmrib.ox.ac.uk/fsl/index.html) into a single workflow. In the preprocessing stage, a rigid-body transformation was used to realign the functional images to the mean EPI image, correcting for subject head movement and for slice timing. Outliers with movements >1 mm or with an intensity Z-threshold >3 standard deviations (SDs) from the mean were removed from the data using an artifact detection algorithm (http://www.nitrc.org/projects/artifact_detect). Surface-based analysis can provide more specific results for cortical structures based on the individually-reconstructed anatomical structure of cortical surfaces. Moreover, this analysis is potentially more sensitive because its search domain is small (no white matter and no cerebrospinal fluid). Cortical surface models can also simplify data visualization by revealing the pattern of activation throughout the whole cortex in one view. For the surfaced-based analysis, Freesurfer was used to segment each anatomical volume into gray and white matter structures and to perform cortical surface reconstruction. The mean functional image generated by realignment was registered to each subject’s reconstructed structural MRI data. The functional images were smoothed on the cortical surface using a Gaussian filter with a full-width at half-maximum (FWHM) of 4 mm. In addition, a whole-brain analysis was conducted that also encompassed subcortical regions and the cerebellum using AFNI software (NIMH, USA); the detailed methods and the resulting activation patterns are reported in .For each stimulus, a boxcar function was used to represent the onset of each event. These time series were then convolved with a canonical hemodynamic response function (HRF) to generate a simulated blood-oxygen-level dependent (BOLD) response. A standard hierarchical group model approach was used to fit the simulated response to scan time-courses. Contrast images were generated for each subject. Conditions were treated as fixed effects. A “summary statistics” procedure was used to model the group effects, performing one-sample t-tests across the individual contrast images. The model was applied with a t-value threshold of 2 and a cluster-threshold correction for multiple comparisons using a Monte Carlo simulation with 10,000 iterations, resulting in p < 0.05.To formally test whether any voxels were significantly activated (stimulation > baseline) during both conditions (genuine and pseudo), we tested the conjunction map under the “conjunction null” hypothesis. For this, we calculated a minimum Z-statistic image between two contrast images (genuine and pseudo) generated during the analysis of main effects of two conditions as a conjunction map of each individual. For the group-level statistical map of the conjunction analysis, a one sample t-test was performed across individual conjunction maps. This conjunction approach asks for the common neural correlates of both genuine acupuncture stimulation and pseudo-stimulation. A group-level statistical map of the difference comparing genuine- and pseudo-stimulation was created by performing paired t-tests between the two contrasts (genuine and pseudo) generated during the main-effect analysis of the two conditions across individuals. This difference map helps to reveal neural correlates of afferent signal processing. The resulting statistical parametric maps were corrected for multiple comparisons using a Monte Carlo simulation (with 10,000 iterations) using a cluster-wise probability threshold. Significant clusters were retained with a cluster-wise probability threshold of 0.01.An additional analysis was run to show correlations across individuals between sensitivity to external stimuli and inter-subject variability in differences of fMRI responses between genuine- and pseudo-stimulation. The map of covariates was created by performing an analysis of covariance (ANCOVA) on individual contrasts calculated by subtracting the contrast of the pseudo condition from the contrast of the genuine condition; the inverse of the demeaned PSE value in the two-point discrimination task was used as the covariate of interest. The covariate map was thresholded at p < 0.001 (Z > 3, uncorrected). To visualize the cortical activation map, statistical parametric maps were overlaid on a high resolution surface template, a default averaged template with high resolution (163842 vertices; 327680 faces) provided by Freesurfer. […]

Pipeline specifications

Software tools Nipype, SPM, FreeSurfer
Application Functional magnetic resonance imaging
Organisms Homo sapiens