PrecisionMapping_Statistics

Analysis code to accompany "Evaluating Approaches for Inference Testing of Whole-Brain Densely Sampled Single-Subject Task fMRI Data" paper

• Subject-level analysis code can be found on the following Github repository (https://github.com/BrightLab-ANVIL/WholeBrain_AuditoryMapping) on folder 1-subjectGLM.

1. Analysis for Inference Approaches

Whole-Brain Analyses

• x.WholeBrainVoxFDR.sh: Uses x.Subject_SigActivation.sh script to find voxels of significant positive activation with False Discovery Rate (FDR) correction

• x.WholeBrain_ClustFWE.sh: Example script to perform cluster-level inference testing with different thresholds

• x.Nonparam_Permutation.sh: Example script to perform non-parametric permutation inference testing with voxel_level (vox_based) or threshold free cluster enhancement (tfce) thresholding. Use mask_location = wholeBrain.

Region-Specific Analyses

• x.RegionSpecific_Top%tstat.sh: Uses DO_TopPer option to find voxels above a specified threshold (e.g. top 5% voxels) based on t-statistic value

• x.Nonparam_Permutation.sh: Example script to perform non-parametric permutation inference testing with voxel_level (vox_based) or threshold free cluster enhancement (tfce) thresholding. Use specific ROI for the mask_location loop (e.g. brainstem).

2. Stability Calculations

• Folds_COM.R: Example script to calculate the euclidean distance between the center of mass (COM) of an inference method's 40-minute data activation cluster and the COM of its respective 30-minute folds.

• Folds_DiceCoeff.R: Example script to calculate the dice coefficient score between an inference method's 40-minute data activation cluster and its respective 30-minute folds.