SPARC: A Hybrid Computational Approach to Classify Vagal C Fiber Functions

Abstract

Listen

Translate article

Neuroscience conventionally categorizes all unmyelinated fibers into an undifferentiated C fiber classification, implying a common function and impeding further classification. This is highly problematic, especially for autonomic nerves like the abdominal vagus, which contains >99% unmyelinated C fibers. We require a modern classification framework to discriminate differences among these important, underappreciated messengers. One important and easily measurable property of C fibers is their conduction speed, which relates directly to their morphology or caliber. We present a hybrid computational modeling approach, integrating biophysical (HH‐type) and classical (e.g., Gasser, Erlanger and Grundfest) methods, to enable the classification of C fiber function according to fiber caliber/speed.The Hybrid Nerve Response Modeling and Classification System is a tool to classify C fiber functions with respect to fiber caliber and radial location within a nerve. Biophysical models of HH axons contribute a bottom‐up model of fiber activation that explains the influence of the electrode configuration and other factors on fiber activation threshold and the measured compound action potential (CAP) profile. Alternatively, modeled CAPs can be generated through superposition of individual axon contributions (single fiber action potentials; SFAPs), relating fiber caliber distribution data to SFAP shape and conduction properties; this classical model of the CAP accurately predicts the experimentally measured maximal CAP shape. Using published and novel nerve ultrastructure data, we unified the biophysical and classical modeling approaches in a new simulation framework that predicts graded CAP profiles in response to bipolar, constant‐current stimuli delivered through various cuff electrode configurations.In this work, we demonstrate how the Hybrid Model can be used to discern the spatial and size distribution of unmyelinated C fibers from the CAP of the ventral gastric branch of the rat vagus nerve. Through our web‐based user interface (Fig. 1), we will show how our system can predict the activation threshold of unmyelinated fibers in any nerve cross section as a function of the individual fiber size and spatial location within a common set of electrode geometries. We will provide examples of how this information may be used to accelerate our delineation of the various functions of C fibers using measured associations between nerve and organ responses to stimulation.Support or Funding InformationFunding AcknowledgmentsThis work was supported by NIH SPARC OT2 OD023847 (PI: Powley), NIH SPARC OT2 OD025340 (PI: Grill) and NIH SPARC OT2 OD026585 01 (PI: Havton)[LEFT] Web‐based interface for the Hybrid Model. [FAR RIGHT] Segmented electron micrograph of n = 4776 C fibers from a rat ventral vagal gastric branch. The user interface lets users track the location (e.g., spatial location in the nerve cross section and temporal location in the CAP response) and relative activation thresholds of C fibers with respect to observed changes in the downstream physiology of the organs and tissues they supply.Figure 1