Premium
Modeling fiber type grouping by a binary markov random field
Author(s) -
Venema Henk W.
Publication year - 1992
Publication title -
muscle and nerve
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.025
H-Index - 145
eISSN - 1097-4598
pISSN - 0148-639X
DOI - 10.1002/mus.880150617
Subject(s) - markov chain , binary number , type (biology) , markov random field , fiber , field (mathematics) , mathematics , statistical physics , computer science , artificial intelligence , statistics , physics , materials science , biology , segmentation , pure mathematics , arithmetic , paleontology , image segmentation , composite material
A new approach to the quantification of fiber type grouping is presented, in which the distribution of histochemical type in a muscle corss section is regarded as a realization of a binary Markov random field (BMRF). Methods for the estimation of the parameters of this model are discussed. The first order BMRF, which is used in this article, contains 2 parameters: α and β. The parameter β is of prime importance, as it is an interaction parameter which governs the degree of type grouping. The value of this parameter is estimated for 9 muscle biopsies. The interpretation of the results is discussed.