Cortical and subcortical mechanisms of brain-machine interfaces

Area of Science:

• Neuroscience
• Brain-Machine Interfaces (BMIs)
• Neuroimaging

Background:

• Current Brain-Machine Interface (BMI) systems typically focus on motor and premotor cortex activity.
• The whole-brain networks, including subcortical regions, involved in BMI control remain largely uncharacterized.
• Understanding these networks is crucial for developing more intuitive and embodied BMI control.

Purpose of the Study:

• To comprehensively map the brain regions active during real-time, on-line BMI control.
• To investigate the overlap between BMI control networks and motor imagery networks.
• To identify neural correlates of the sense of agency experienced during BMI use.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was employed to record brain activity.
• Participants performed an electroencephalography (EEG)-based BMI task inside the fMRI scanner.
• Analysis focused on identifying activated regions during BMI control and correlating them with subjective agency.

Main Results:

• An extensive cortical-subcortical network was identified during motor-imagery BMI operation.
• Activated regions included sensorimotor cortices, posterior parietal cortex, insula, and lateral occipital cortex.
• The basal ganglia and anterior cingulate cortex were specifically associated with the subjective sense of BMI control.

Conclusions:

• BMI control engages a broader neural network than previously understood, extending beyond sensorimotor areas.
• Distinct regions, including basal ganglia and ACC, contribute to the sense of agency in BMI use.
• Findings offer insights for designing BMIs that provide a more natural and embodied user experience.