Decoding two-dimensional movement trajectories using electrocorticographic signals in humans

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain-computer interfaces (BCIs) offer communication and control for severely paralyzed individuals.
• Current BCIs often rely on intracortical microelectrodes, facing challenges with long-term stability.
• Decoding kinematic parameters from brain signals is crucial for BCI development.

Purpose of the Study:

• To investigate the feasibility of using electrocorticography (ECoG) for decoding human movement kinematics.
• To compare the accuracy of ECoG-based decoding with established intracortical electrode methods.
• To identify novel ECoG signal features for improved BCI performance.

Main Methods:

• Recorded human brain activity using subdural ECoG electrodes during arm movement.
• Decoded kinematic parameters from ECoG signals.
• Introduced and analyzed a new ECoG feature, the local motor potential (LMP).
• Assessed feature tuning properties, including cosine tuning.

Main Results:

• Kinematic parameters were decoded from ECoG signals with accuracy comparable to intracortical recordings in non-human primates.
• The novel local motor potential (LMP) feature demonstrated high information content for movement decoding.
• ECoG signal features exhibited cosine tuning, previously observed only in intracortical recordings.

Conclusions:

• Electrocorticography (ECoG) provides a viable, stable, and less invasive alternative to intracortical electrodes for brain-computer interface (BCI) systems.
• ECoG-based BCIs show promise for restoring communication and control in individuals with severe paralysis.
• ECoG recordings offer valuable insights into human motor control mechanisms.