Mapping and decoding cortical engagement during motor imagery, mental arithmetic, and silent word generation using...

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Quantifying brain activity during mental tasks is crucial for brain-computer interfaces (BCIs).
• Magnetoencephalography (MEG) offers a non-invasive method to measure neural activity.
• Beta band oscillations are known to be involved in motor control and cognitive processes.

Purpose of the Study:

• To investigate the utility of beta band power decrements for mapping cortical engagement during various mental tasks.
• To assess the accuracy of machine learning classifiers in decoding mental tasks based on MEG data.
• To explore the potential of these neuroimaging markers for rehabilitation and clinical applications.

Main Methods:

• Analyzed MEG data from 18 participants performing motor imagery (hands, feet), mental arithmetic, and silent word generation.
• Estimated task-related cortical engagement using beta band (17-25 Hz) power decrements via frequency-resolved beamforming.
• Employed support vector machine and Gaussian-process classifiers to decode task types from neural data.

Main Results:

• Consistent beta power decreases observed in motor/premotor areas for motor imagery and in temporal/parietal/frontal regions for cognitive tasks.
• High classification accuracies achieved: 74% for motor imagery (hands vs. feet), 68% for cognitive tasks (word vs. subtraction).
• Motor vs. non-motor task classification yielded accuracies up to 85% (hands vs. word).

Conclusions:

• Beta band power decrements serve as reliable metrics for mapping and decoding cortical engagement during mental processes.
• These findings demonstrate the potential of MEG-based markers for BCIs, rehabilitation, and characterizing neurological impairments.
• Individual task performance can be assessed through within-subject correlations of beta decrements.