Temporal alignment of electrocorticographic recordings for upper limb movement

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Brain-computer interfaces (BCIs) rely on detecting movement-related brain activity.
• Traditional methods average neural responses, assuming consistent timing across trials.
• Complex movements like reaching and grasping exhibit significant trial-to-trial variability in timing and speed, hindering accurate signal averaging.

Purpose of the Study:

• To develop and evaluate an alignment method that accounts for movement variability in electrocorticography (ECoG) data.
• To enhance the visualization and detection of movement-related neural components for improved BCI design.

Main Methods:

• Utilized electrocorticographic (ECoG) recordings from four subjects performing upper limb reaching and retrieving tasks.
• Employed arm speed as a reference to align neural activity across trials.
• Applied a non-linear transformation to the temporal axes for alignment.

Main Results:

• The alignment method successfully compensated for temporal variabilities in movement execution.
• Resulting average spectrograms showed superior visualization of movement-related neural activity compared to unaligned data.
• Demonstrated enhanced detection of neural components crucial for BCI applications.

Conclusions:

• Movement-related neural activity in complex tasks is subject to significant temporal variability.
• An arm-speed-based alignment method effectively addresses this variability, improving neural signal analysis.
• This approach offers a more robust method for extracting movement-related brain signals for BCI development.