Increased sensitivity in identifying language-related functional connectivity using jackknife resampling analyses

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Brain Imaging

Background:

• Task-based fMRI (tbfMRI) often uses static correlation methods for functional connectivity (FC) analysis.
• Static methods may overlook transient neural interactions crucial for understanding brain function.

Purpose of the Study:

• To investigate if jackknife resampling enhances the detection of language-related FC networks in tbfMRI.
• To compare the sensitivity of jackknife resampling versus static correlation for identifying neural connections.

Main Methods:

• Analyzed surface-based FC networks in 172 healthy adults using data from the Human Connectome Project.
• Computed FC matrices across 68 cortical regions of interest, applying Bonferroni correction for statistical significance.
• Compared static correlation FC networks with those derived from jackknife resampling, using an edge consistency threshold.

Main Results:

• The static method identified 75 significant language-related FCs.
• Jackknife resampling identified all 75 static FCs plus 24 additional connections (p < 0.001).
• These additional connections involved key language regions like the middle temporal gyrus and posterior cingulate cortex.

Conclusions:

• Jackknife resampling significantly enhances the detection of robust, task-relevant functional connections.
• This technique offers a promising alternative for modeling language networks and improving neurocomputational representations.
• Improved FC detection has implications for both research and clinical applications in neuroscience.