Determining significant connectivity by 4D spatiotemporal wavelet packet resampling of functional neuroimaging data

Area of Science:

• Neuroimaging
• Functional Connectivity Analysis
• Signal Processing

Background:

• Neuroimaging data contains inherent background spatial correlations from factors like scanner noise and preprocessing.
• Distinguishing true functional connectivity from these background correlations is crucial for accurate brain region analysis.
• Existing methods may not adequately address the spatiotemporal nature of these correlations.

Purpose of the Study:

• To develop and validate a novel 4D spatiotemporal wavelet packet resampling method.
• To generate surrogate neuroimaging data that preserves background spatial correlations while excluding true functional relationships.
• To improve the accuracy of functional connectivity detection in fMRI datasets.

Main Methods:

• A 4D spatiotemporal wavelet packet resampling technique was developed.
• The method generates surrogate data preserving average background spatial correlation across slices and time series.
• An amplitude adjustment algorithm was extended to match surrogate data amplitude to original data.

Main Results:

• The developed resampling method effectively isolates and preserves background spatial correlations.
• The technique successfully excludes correlations arising from true functional relationships.
• The method was applied to resting-state and motor task fMRI datasets, demonstrating its utility.

Conclusions:

• The 4D wavelet resampling method provides a robust approach to account for background spatial correlation in neuroimaging.
• This technique enhances the precision of identifying significant functional connectivity in fMRI data.
• The method offers an improvement over existing wavelet-based approaches and extends their applicability to 4D data.