Parameter Estimation of Human Nerve C-Fibers using Matched Filtering and Multiple Hypothesis Tracking : Complete report

We describe a multiple target tracking (MTT) application in microneurography that estimate conduction velocity changes and recovery constants of human nerve C-fibers. Results using the tracking system on real data are presented. Action potentials (APs) were recorded from C-fibers in the peroneal nerve of awake human subjects. The APs were detected by a matched filter constituting a maximum likelihood constant false alarm rate (ML-CFAR) detector. Using the multiple hypothesis tracking (MHT) method, the detected APs (targets) in each trace (scan) were associated to individual nerve fibers (tracks) by their typical conduction latencies in response to electrical stimulation. The measurements were one-dimensional (range only) and the APs were spaced in time with intersecting trajectories. In general, the AP amplitude of each C-fiber differed for different fibers. Amplitude estimation was therefore incorporated into the tracking algorithm to improve the performance. The target trajectory was modeled as an exponential decay with three unknowns. These parameters were estimated iteratively by applying the simplex method on the parameters that enter nonlinearly and the least squares method on the parameters that enter linearly. Keywords—Detection, matched filtering, target tracking, MHT, multiple hypothesis tracking, parameter estimation, microneurography