Segmentation of the cartilage in the rib cage in 3d MRI

Authors:
Yolanda H. Noorda;Lambertus W. Bartels;Josien P. W. Pluim
Affiliations:
Image Sciences Institute, UMC Utrecht, Utrecht, The Netherlands;Image Sciences Institute, UMC Utrecht, Utrecht, The Netherlands;Image Sciences Institute, UMC Utrecht, Utrecht, The Netherlands
Venue:
MICCAI'12 Proceedings of the 4th international conference on Abdominal Imaging: computational and clinical applications
Year:
2012

Citing 5
Cited 0

Muliscale Vessel Enhancement Filtering

MICCAI '98 Proceedings of the First International Conference on Medical Image Computing and Computer-Assisted Intervention
Fast Radial Symmetry for Detecting Points of Interest

IEEE Transactions on Pattern Analysis and Machine Intelligence
Automated model-based rib cage segmentation and labeling in CT images

MICCAI'07 Proceedings of the 10th international conference on Medical image computing and computer-assisted intervention
Automatic human knee cartilage segmentation from multi-contrast MR images using extreme learning machines and discriminative random fields

MLMI'11 Proceedings of the Second international conference on Machine learning in medical imaging
Automatic segmentation of the articular cartilage in knee MRI using a hierarchical multi-class classification scheme

MICCAI'05 Proceedings of the 8th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

Interventional non-invasive MR-guided techniques for treatment of liver tumors, such as HIFU, could benefit greatly from automatic cartilage detection. In this paper, segmentation of the cartilage in the rib cage is performed in 3D MR images. This is a challenging task, due to the poor contrast between cartilage and muscle, and the non-uniform intensity of the cartilage. Our segmentation algorithm is based on feature selection by analyzing orientation and vesselness, automatic sternum localization using anatomical knowledge, skeletonization and ridge finding, and level set evolution. We show that our algorithm is capable of detecting all visible cartilage structures in the scans. Gaps and false positives may occur, due to lack of contrast or the presence of non-cartilage structures with similar features. However, the segmentation is accurate, even for regions with low contrast, with an average error of the boundary of 1.1 mm.