A system for analysis of human pain signals using a radar tracking approach

Master Thesis, Uppsala University, UPTEC F 98 078, Oct 1998.

Abstract:

In order to study peripheral neural activity correlating to pain, an analysis tool that facilitates the investigation of human unmyelinated (C-) fibers has been developed. The tool calculates important C-fiber data such as latency changes and recovery time constants, and helps the researcher to present and statistically process data.

By applying electrical stimuli repetitively at 0.25 Hz and additional stimuli, such as mechanical or chemical, at the receptive field of a studied C-fiber it is possible to estimate important data. The action potentials (APs) that are evoked by the stimuli are recorded. The recordings will usually contain APs from several fibers. Our tool has to determine which C-fiber an AP originated from to be able to get information about the single fibers.

To associate APs to a single C-fiber automatically, an algorithm from the radar tracking area, multiple hypothesis tracking (MHT), is used. The APs are treated as radar targets, and the algorithm tracks and calculates the most probable traces. After tracking, curves are fitted to each trace and C-fiber data are calculated. The analysis program consists of three main parts: detection, association (tracking) of APs and estimation of important constants.

Thesis Advisors:

Björn Hansson, Mikael Sternad

Related publications:

Report on the complete algorithm.
SPIE Conference paper on the detection and tracking.

Source:

Thesis			Slides  (PowerPoint format)
	PDF (1365 K),   Postscript (4975 K), compressed PS (gz) (946 K)		Technical (1147 K)		Medical (789 K)