Morphological contrast measure and contrast enhancement: One application to the segmentation of brain MRI

In this paper a morphological contrast measure is introduced. The quantification of the contrast is based on the analysis of the edges, which are associated with substantial changes in luminance. Due to this, the contrast measure is used to detect the image that presents a high visual contrast when...

Full description

Saved in:
Bibliographic Details
Published in:Signal processing 2007-09, Vol.87 (9), p.2125-2150
Main Authors: Mendiola-Santibañez, Jorge D., Terol-Villalobos, Iván R., Herrera-Ruiz, Gilberto, Fernández-Bouzas, Antonio
Format: Article
Language:English
Subjects:
Biological and medical sciences
Computerized, statistical medical data processing and models in biomedicine
Connected transformations
Contrast mappings
Medical management aid. Diagnosis aid
Medical sciences
Morphological contrast measure
MRI segmentation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper a morphological contrast measure is introduced. The quantification of the contrast is based on the analysis of the edges, which are associated with substantial changes in luminance. Due to this, the contrast measure is used to detect the image that presents a high visual contrast when a set of output images is analyzed. The set of output images is obtained by application of morphological contrast mappings with size criteria. These contrast transformations are defined under the notion of partitions generated by the set of flat zones of the image; therefore, they are connected transformations. In addition, an application to the segmentation of white and grey matter in brain magnetic resonance images (MRI) is provided. The detection of white matter is carried out by means of a contrast mapping with specific control parameters; subsequently, white and grey matter are separated and their ratio is calculated and compared with manual segmentations. Also, an example of segmentation of white and grey matter in MRI corrupted by 5% noise is presented in order to observe the performance of the morphological transformations proposed in this work.
ISSN:0165-1684
1872-7557
DOI:10.1016/j.sigpro.2007.02.008