Neural decoding using gyral and intrasulcal electrocorticograms

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electrocorticography (ECoG) of the primary motor cortex (M1) is a key technology for brain-computer interfaces (BCIs).
• Intrasulcal ECoG, recorded within the central sulcus of M1, is less explored compared to gyral ECoG.
• Understanding intrasulcal ECoG's utility is crucial for advancing BCI systems.

Purpose of the Study:

• To evaluate the effectiveness of intrasulcal ECoG for neural decoding of upper limb movements.
• To compare the performance of intrasulcal ECoG with gyral ECoG for BCI applications.
• To determine the anatomical importance of M1 regions for movement prediction.

Main Methods:

• Recorded intrasulcal and gyral ECoG from five patients performing or imagining upper limb movements.
• Utilized univariate statistical analysis to assess signal variability across movement classes.
• Employed a support vector machine for single-trial neural decoding of movement intentions.

Main Results:

• Intrasulcal ECoG signals exhibited significant variability correlating with different movement classes.
• Neural decoding achieved 80-90% accuracy in predicting movement classes.
• Intrasulcal ECoG on the motor cortex bank outperformed gyral ECoG and intrasulcal ECoG on the sensory bank.

Conclusions:

• Intrasulcal ECoG on M1 is highly effective for inferring upper limb movements.
• This technique shows significant promise for developing practical and high-performance BCI systems.
• Early decoding before movement onset highlights the potential for real-time BCI control.