Motor imagery of voluntary coughing: a functional MRI study using a support vector machine

Area of Science:

• Neuroscience
• Respiratory Physiology
• Cognitive Motor Science

Background:

• Investigating respiratory acts like coughing typically involves overt physical actions.
• Overt movements can introduce motion artifacts in neuroimaging studies, complicating data analysis.
• Motor imagery offers a potential alternative to reduce motion artifacts during respiratory research.

Purpose of the Study:

• To determine if motor imagery of voluntary coughing elicits similar spatiotemporal brain activity patterns as overt coughing.
• To evaluate the utility of motor imagery as a paradigm for studying respiratory functions.
• To compare the effectiveness of pattern classification (support vector machine) versus standard univariate analysis for detecting these patterns.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to scan 12 participants.
• Participants performed both overt coughing and imagined coughing tasks.
• Support vector machine (SVM) pattern classification was applied to analyze fMRI data.

Main Results:

• Brain activity patterns during imagined coughing showed greater similarity to overt coughing than to a resting baseline.
• Specific brain areas previously implicated in respiration exhibited these similar spatiotemporal patterns.
• SVM analysis demonstrated potential for higher sensitivity and specificity compared to univariate methods.

Conclusions:

• Motor imagery of voluntary coughing can effectively mimic the neural activity of overt coughing.
• Motor imagery is a promising, artifact-reducing paradigm for investigating respiratory control and function.
• Support vector machine analysis offers a sensitive and specific approach for analyzing brain activity in such tasks.