Open AccessApplication of texture analysis to muscle MRI: 1- What kind of information should be expected from texture analysis?
Author(s) -
Jacques D. de Certaines,
Thibaut Larcher,
Dorota Duda,
Noura Azzabou,
PierreAntoine Eliat,
Luis Escudero,
António Pinheiro,
Guanyu Yang,
Jean-Louis Coatrieux,
Eduard Snezkho,
Alexey Shukelovich,
Manuela Pereira,
R A Lerski
Publication year - 2015
Publication title -
hal (le centre pour la communication scientifique directe)
Language(s) - English
DOI - 10.1140/s40366-015-0017-1
Subject(s) - texture (cosmology) , computer science , artificial intelligence , computer vision , pattern recognition (psychology) , image (mathematics)
International audienceSeveral previous clinical or preclinical studies using computerized texture analysis of MRImages have demonstrated much more clinical discrimination than visual image analysis bythe radiologist. In muscular dystrophy, a discriminating power has been already demonstratedwith various methods of texture analysis of magnetic resonance images (MRI-TA).Unfortunately, a scale gap exists between the spatial resolutions of histological and MRimages making a direct correlation impossible. Furthermore, the effect of the varioushistological modifications on the gray level of each pixel is complex and cannot be easilyanalyzed. Consequently, clinicians will not accept the use of MRI-TA in routine practice ifTA remains a “black box” without clinical correspondence at a tissue level. A goal thereforeof the multicenter European COST action MYO-MRI is to optimize MRI-TA methods inmuscular dystrophy and to elucidate the histological meaning of MRI textures