Share Email Print
cover

Proceedings Paper

Cerebella segmentation on MR images of pediatric patients with medulloblastoma
Author(s): Zu Y. Shan; Qing Ji; John Glass; Amar Gajjar; Wilburn E. Reddick
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In this study, an automated method has been developed to identify the cerebellum from T1-weighted MR brain images of patients with medulloblastoma. A new objective function that is similar to Gibbs free energy in classic physics was defined; and the brain structure delineation was viewed as a process of minimizing Gibbs free energy. We used a rigid-body registration and an active contour (snake) method to minimize the Gibbs free energy in this study. The method was applied to 20 patient data sets to generate cerebellum images and volumetric results. The generated cerebellum images were compared with two manually drawn results. Strong correlations were found between the automatically and manually generated volumetric results, the correlation coefficients with each of manual results were 0.971 and 0.974, respectively. The average Jaccard similarities with each of two manual results were 0.89 and 0.88, respectively. The average Kappa indexes with each of two manual results were 0.94 and 0.93, respectively. These results showed this method was both robust and accurate for cerebellum segmentation. The method may be applied to various research and clinical investigation in which cerebellum segmentation and quantitative MR measurement of cerebellum are needed.

Paper Details

Date Published: 29 April 2005
PDF: 7 pages
Proc. SPIE 5747, Medical Imaging 2005: Image Processing, (29 April 2005); doi: 10.1117/12.594661
Show Author Affiliations
Zu Y. Shan, St. Jude Children's Research Hospital (United States)
Qing Ji, St. Jude Children's Research Hospital (United States)
John Glass, St. Jude Children's Research Hospital (United States)
Amar Gajjar, St. Jude Children's Research Hospital (United States)
Wilburn E. Reddick, St. Jude Children's Research Hospital (United States)


Published in SPIE Proceedings Vol. 5747:
Medical Imaging 2005: Image Processing
J. Michael Fitzpatrick; Joseph M. Reinhardt, Editor(s)

© SPIE. Terms of Use
Back to Top