Logo image
Automated analysis of cartilage morphology
Conference proceeding

Automated analysis of cartilage morphology

Satyananda Kashyap, Yin Yin and Milan Sonka
2013 IEEE 10th International Symposium on Biomedical Imaging, pp.1300-1303
04/2013
DOI: 10.1109/ISBI.2013.6556770

View Online

Abstract

Magnetic Resonance Imaging (MRI) allows early detection of and assessment of progression in Osteoarthritis (OA). It is known that the rate of cartilage loss varies in individual cartilage plates. No known completely automated algorithms exist to reliably segments MRI image data and accurately delineate cartilage plates. We report an automated method for cartilage segmentation and definition of cartilage plates. We present a modified Layered Optimal Graph Image Segmentation of Multiple Objects and Surfaces (LOGISMOS) with a robust steerable-feature-based design and demonstrate improvements of segmentation accuracy. Our new method was tested on 60 data sets provided by the MICCAI segmentation challenge 2010. The mean and standard deviation of surface positioning errors are reported for individual cartilage plate regions.
Magnetic Resonance Imaging Osteoarthritis Bars Bones Cartilage Plates graph-based segmentation Image segmentation Knee Segmentation LOGISMOS Standards Steerable Features Surface morphology

Details

Metrics

32 Record Views
Logo image