Related Experiment Video
Updated: Sep 11, 2025
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
15.8K
Selective inference for fMRI cluster-wise analysis, issues, and recommendations for critical vector selection: A
Angela Andreella1, Anna Vesely2, Wouter Weeda3
1Department of Economics, Ca' Foscari University of Venice, Venice, Italy.
Imaging Neuroscience (Cambridge, Mass.)
|August 13, 2025
Summary
pARI generally outperforms Notip for cluster-wise brain imaging analysis when using recommended settings. Both permutation-based methods for simultaneous inference on active voxels have unique strengths and weaknesses.
Area of Science:
- Neuroimaging
- Statistical analysis
- Brain mapping
Background:
- Simultaneous inference is crucial for cluster-wise brain imaging analysis.
- Two permutation-based methods, Notip and pARI, address inference on active voxel proportions.
- Existing comparisons lack an extensive evaluation of Notip versus pARI.
Purpose of the Study:
- To provide a comprehensive comparison between Notip and pARI.
- To evaluate their performance in cluster-wise brain imaging analysis.
- To identify the advantages and disadvantages of each method.
Main Methods:
- Permutation-based statistical methods.
- Comparison of Notip and pARI under recommended settings.
- Analysis of critical vector definition strategies (external randomization vs. a priori determination).
Main Results:
- pARI demonstrated superior performance compared to Notip.
- Both methods utilize ordered p-value vectors but differ in family definition.
- Notip uses external data randomization; pARI uses a priori definition.
Conclusions:
- pARI is recommended over Notip for cluster-wise brain imaging analysis under standard conditions.
- Each method possesses distinct advantages and limitations.
- Further research may explore hybrid approaches or optimized settings.
