Biphasic Electrostimulation Artifact Model for ECAP Extraction in Spinal Cord Stimulation.

Area of Science:

• Biomedical Engineering
• Neuroscience
• Signal Processing

Background:

• Evoked compound action potential (ECAP) is crucial for monitoring neural activity in closed-loop spinal cord stimulation (SCS).
• Low signal-to-noise ratio in ECAP recordings, caused by stimulation artifacts, hinders accurate neural activity assessment.
• Existing methods struggle to effectively isolate neural signals from stimulation artifacts.

Purpose of the Study:

• To propose and validate the Biphasic Electrostimulation Artifact Model (BEAM) for accurate ECAP extraction.
• To develop an enhanced quantification approach using a bilinear growth curve model for evaluating ECAP extraction efficacy.
• To demonstrate the clinical utility of BEAM in reducing stimulation artifacts without distorting neural information.

Main Methods:

• Developed the BEAM model based on the principle of additivity and boundary conditions reflecting the electrical environment and temporal dynamics.
• Introduced a bilinear growth curve model to calculate the E-score for quantitative comparison of ECAP extraction methods.
• Validated the BEAM approach using clinical data from nine-month spinal cord stimulation therapy.

Main Results:

• The BEAM method achieved the highest E-scores compared to other methods.
• BEAM successfully reduced stimulation artifacts in clinical ECAP recordings.
• The model accurately characterized artifact morphology across diverse neurostimulation scenarios without distorting neural activity information.

Conclusions:

• BEAM offers a robust solution for extracting ECAP signals by effectively modeling and removing stimulation artifacts.
• This method enhances the reliability of neural activity monitoring in closed-loop SCS.
• BEAM has the potential to enable novel interpretations of neural activation for advanced closed-loop control strategies.