Spine disc MR image analysis using improved independent component analysis based active appearance model and Markov random field.
- Author:
Shijie HAO
1
;
Shu ZHAN
;
Jianguo JIANG
;
Hong LI
;
Rosse IAN
Author Information
1. School of Computer and Information, Hefei University of Technology, Hefei 230009, China.
- Publication Type:Journal Article
- MeSH:
Algorithms;
Humans;
Image Interpretation, Computer-Assisted;
methods;
Intervertebral Disc;
pathology;
Intervertebral Disc Displacement;
diagnosis;
pathology;
Lumbar Vertebrae;
pathology;
Magnetic Resonance Imaging;
Markov Chains
- From:
Journal of Biomedical Engineering
2010;27(1):6-15
- CountryChina
- Language:Chinese
-
Abstract:
As there are not many research reports on segmentation and quantitative analysis of soft tissues in lumbar medical images, this paper presents an algorithm for segmenting and quantitatively analyzing discs in lumbar Magnetic Resonance Imaging (MRI). Vertebrae are first segmented using improved Independent component analysis based active appearance model (ICA-AAM), and lumbar curve is obtained with Minimum Description Length (MDL); based on these results, fast and unsupervised Markov Random Field (MRF) disc segmentation combining disc imaging features and intensity profile is further achieved; finally, disc herniation is quantitatively evaluated. The experiment proves that the proposed algorithm is fast and effective, thus providing doctors with aid in diagnosing and curing lumbar disc herniation.
