Dynamical cluster analysis of cortical fMRI activation

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• Functional magnetic resonance imaging (fMRI) is crucial for mapping brain activity.
• Existing clustering methods for fMRI data have limitations in detecting activation sites and response shapes.
• Analyzing hemodynamic responses requires robust and adaptable analytical tools.

Purpose of the Study:

• To introduce and evaluate a novel dynamical cluster analysis (DCA) method for fMRI data.
• To compare the performance of DCA against standard clustering algorithms like k-means.
• To demonstrate DCA's ability to detect localized cortical blood oxygenation changes during voluntary movements.

Main Methods:

• fMRI data acquired during finger movements using a 1.5-T scanner and echo planar imaging.
• Application of a new dynamical cluster analysis (DCA) method based on signal time course similarity.
• Simultaneous calculation of multidimensional scaling (MDS) for online visualization.

Main Results:

• DCA outperforms k-means in terms of quantization error and result reproducibility.
• DCA automatically determines the number and shapes of representative signal time courses.
• The method successfully detected cortical activation loci and discriminated different hemodynamic response shapes and phases.
• Analysis of activation onset times revealed simultaneous and sequential activation in SMA, M1, and S1 areas.

Conclusions:

• DCA is a robust and effective method for analyzing fMRI data, offering advantages over traditional clustering techniques.
• DCA provides detailed insights into the dynamics of hemodynamic responses during motor tasks.
• The method enhances the understanding of brain activation patterns and temporal dynamics in specific cortical regions.