Toward brain-heart computer interfaces: a study on the classification of upper limb movements using multisystem...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Cardiovascular Dynamics

Background:

• Brain-computer interfaces (BCIs) traditionally use brain signals for tasks.
• The interplay between brain and cardiovascular dynamics is vital for locomotion but often overlooked in BCIs.
• Integrating heartbeat information into BCIs can enhance their functionality.

Purpose of the Study:

• To exploit multidimensional features of brain-heart interplay (BHI) for classifying upper limb movements.
• To develop a brain-heart computer interface (BHCI) system.
• To investigate the neurophysiological role of cortical-level functional interplay on cardiac activity in upper limb control.

Main Methods:

• Collected data from 26 healthy volunteers performing 90 movements.
• Utilized a synthetic physiological data generation framework for BHI assessment.
• Employed sequential forward selection and k-nearest neighbors for classification of BHI features.

Main Results:

• The proposed BHCI system automatically distinguished between rest and movement classes with 90% average accuracy.
• Demonstrated the feasibility of using BHI features for classifying three distinct upper limb movement types.
• Identified the crucial role of functional interplay from the cortical level to the heart in neural control.

Conclusions:

• The integration of functional brain-heart interplay (BHI) shows significant potential for improving BCIs.
• This approach offers neurophysiological insights into motor control mechanisms.
• Advanced BHCI systems could be developed by incorporating BHI insights.