Deciphering the Origins of Evoked Compound Action Potentials in Peripheral Nerves: Mechanism and Implication

Introduction
Accurate assessment of nerve activation location and volume is crucial for optimizing neuromodulation therapy but remains a significant challenge. When stimulation is applied to the nerve bundle (peripheral nerves or spinal cord) using an electrode, an electric field recruits a segment of the nerve(s), termed the activated volume. Within this volume, nerve fibers produce action potentials that propagate in both directions . Action potentials from the same nerve fiber traveling in opposite directions within the activated volume collide and nullify each other . At the boundaries, action potentials travel outward, reaching recording electrodes and being recorded as evoked compound action potentials (eCAPs). However, current eCAP analysis methods struggle to accurately assess the location and volume of activated nerves.
Methods
We introduce the 'activation origin' framework, inspired by the observation that higher stimulation amplitudes reduce eCAP latency. This reduction is attributed to the expanded activation volume, causing neural recruitment nearer to the recording site and thus shortening the eCAP travel distance and the apparent latency of the eCAPs. By calculating eCAP travel time and velocity with multiple recording electrodes, we can determine the total distance traveled by the eCAPs (velocity × time). This allow us to trace these signals back to their origins and define the boundary of the activated volume.
Results
Spatiotemporal eCAP data were collected from rodent and swine vagus nerves. Our calculations confirmed accurate cathode positions at threshold stimulations. Monopolar stimulations showed that, as stimulation increased, eCAP origins moved away from the cathode. During anodic stimulation, virtual cathodes were identified near, but not directly on, the anode, at 1.8-2.5 times the threshold activated with cathodic stimulation, corroborating Ranck's (1975) findings. These findings revealed depolarization and hyperpolarization regions created by anodes and cathodes. Bipolar stimulations showed complex interactions, demonstrating multiple activation zones based on biophysical principles.
Conclusion
We developed a method to visualize and quantify activation locations and volumes, with several implications: 1. Improving programming practices by enabling practitioners to visualize how parameters impact activation. 2. Bridging the gap between computational models and eCAPs by personalizing models with patient-derived activation zones. 3. Enabling closed-loop stimulation via constant volume activation rather than eCAP amplitude, making it more representative of true neural responses regardless of patient posture or electrode impedance. 4. As only neural signals possess neuro-plausible origins, the calculation can authenticate neural signals by distinguishing them from EMG and motion artifacts, a major challenge in human eCAP, without using artifact-eliminating controls.