Evaluation of texture parameters for the quantitative description of multimodal nonlinear optical images from atherosclerotic rabbit arteries
; Mostaço-Guidolin, Leila B.
; Ko, Alex C.-T.
; Popescu, Dan P.
; Smith, Michael S.D.
; Kohlenberg, Elicia K.
; Shiomi, Masashi
; Major, Arkady
Sowa, Michael G.
NRC Institute for Biodiagnostics; National Research Council Canada
Physics in Medicine and Biology
The composition and structure of atherosclerotic lesions can be directly related to the risk they pose to the patient. Multimodal nonlinear optical (NLO) microscopy provides a powerful means to visualize the major extracellular
components of the plaque that critically determine its structure. Textural
features extracted from NLO images were investigated for their utility in
providing quantitative descriptors of structural and compositional changes
associated with plaque development. Ten texture parameters derived from
the image histogram and gray level co-occurrence matrix were examined that
highlight specific structural and compositional motifs that distinguish early
and late stage plaques. Tonal–texture parameters could be linked to key
histological features that characterize vulnerable plaque: the thickness and
density of the fibrous cap, size of the atheroma, and the level of inflammation
indicated through lipid deposition. Tonal and texture parameters from NLO
images provide objective metrics that correspond to structural and biochemical
changes that occur within the vessel wall in early and late stage atherosclerosis.