Spatial statistical methods for detecting small fiber neuropathies

Epidermal nerve fibers (ENFs) are thin sensory nerve fibers in the epidermis, which is the outmost part of the skin. William Kennedy´s group at the University of Minnesota consists of leading experts in ENF studies. They have observed mainly by visual inspection that both the number of ENFs per area and the spatial structure of them in samples taken from a subject with some small fiber neuropathy, such as diabetic neuropathy, differ from samples taken from healthy subjects. More precisely, the diseased ENF patterns seem to have less ENFs per area and they tend to be more clustered than the healthy patterns. Current clinical practice focuses on observed numbers of ENFs but further analysis of the spatial distribution of ENFs suggests potential to detect small fiber neuropathies in earlier stages when observed ENF counts remain within normative range. Therefore, we will focus on the spatial structure of ENFs in healthy and neuropathic samples. In order to detect and understand differences between healthy and neuropathic patterns, we will construct stochastic spatial processes that generate patterns similar to those observed in the data. In statistical terms, we will need to construct new spatial and spatio-temporal (parametric) models that are able to capture the details of the ENF structure, and find inference tools for the models. The project will be done in close collaboration with William Kennedy and Gwen Wendelschafer-Crabb, who will also provide us with data

Start date 01/01/2014
End date The project is closed: 31/12/2017
​William Kennedy, University of Minnesota
Gwen Wendelschafer-Crabb, University of Minnesota
Peter Guttorp, University of Washington
Mari Myllymäki, Aalto University, Finland

Funded by

  • Swedish Research Council (VR) (Public, Sweden)
​Life Science

Published: Tue 18 Jun 2019.