Optimal artifact suppression in simultaneous electrocorticography stimulation and recording for bi-directional...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electrocorticogram (ECoG)-based brain-computer interfaces (BCIs) aim to restore function in neurological deficits.
• Simultaneous recording and stimulation in BCIs are hindered by strong stimulation artifacts.
• Ultra-low power operation for implantable BCIs exacerbates artifact issues.

Purpose of the Study:

• Develop a novel method for suppressing stimulation artifacts before they reach the analog front-end.
• Address challenges in bi-directional BCI operation for neurological restoration.
• Prevent saturation of analog front-end amplifiers in ultra-low power BCI systems.

Main Methods:

• Devised an artifact suppression method using elementary biophysical principles.
• Employed a weak auxiliary stimulation between the primary stimulator and recording grid.
• Utilized a constrained optimization procedure to determine the auxiliary stimulating dipole's parameters.

Main Results:

• Achieved artifact suppression of 28.7 dB in simulations and 22.9 dB in brain phantom experiments.
• The optimized auxiliary dipole matched the optimal canceling dipole in both simulations and experiments.
• Demonstrated effective artifact mitigation prior to analog front-end processing.

Conclusions:

• A simple, constrained optimization-based method effectively suppresses stimulation artifacts.
• The technique prevents analog front-end amplifier saturation in BCIs.
• This method can complement other artifact mitigation strategies for enhanced performance.