Share Email Print

Proceedings Paper

Comparison and combination of scaling index method and Minkowski functionals in the analysis of high resolution magnetic resonance images of the distal radius in vitro
Author(s): Irina N. Sidorenko; Jan Bauer; Roberto Monetti; Dirk Mueller; Ernst J. Rummeny; Felix Eckstein; Christoph W. Raeth
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

High resolution magnetic resonance (HRMR) imaging can reveal major characteristics of trabecular bone. The quantification of this trabecular micro architecture can be useful for better understanding the progression of osteoporosis and improve its diagnosis. In the present work we applied the scaling index method (SIM) and Minkowski Functionals (MF) for analysing tomographic images of distal radius specimens in vitro. For both methods, the correlation with the maximum compressive strength (MCS) as determined in a biomechanical test and the diagnostic performance with regard to the spine fracture status were calculated. Both local SIM and global MF methods showed significantly better results compared to bone mineral density measured by quantitative computed tomography. The receiver operating characteristic analysis for differentiating fractured and non-fractured subjects revealed area under the curve (AUC) values of 0.716 for BMD, 0.897 for SIM and 0.911 for MF. The correlation coefficients with MCS were 0.6771 for BMD, 0.843 for SIM and 0.772 for MF. We simulated the effect of perturbations, namely noise effects and intensity variations. Overall, MF method was more sensitive to noise than SIM. A combination of SIM and MF methods could, however, increase AUC values from 0.85 to 0.89 and correlation coefficients from 0.71 to 0.82. In conclusion, local SIM and global MF techniques can successfully be applied for analysing HRMR image data. Since these methods are complementary, their combination offers a new possibility of describing MR images of the trabecular bone, especially noisy ones.

Paper Details

Date Published: 19 March 2008
PDF: 9 pages
Proc. SPIE 6914, Medical Imaging 2008: Image Processing, 69144V (19 March 2008); doi: 10.1117/12.769395
Show Author Affiliations
Irina N. Sidorenko, Max-Planck-Institut für extraterrestrische Physik (Germany)
Jan Bauer, Technische Univ. Muenchen (Germany)
Roberto Monetti, Max-Planck-Institut für extraterrestrische Physik (Germany)
Dirk Mueller, Technische Univ. Muenchen (Germany)
Ernst J. Rummeny, Technische Univ. Muenchen (Germany)
Felix Eckstein, Paracelsus Medical Private Univ. (Austria)
Christoph W. Raeth, Max-Planck-Institut für extraterrestrische Physik (Germany)

Published in SPIE Proceedings Vol. 6914:
Medical Imaging 2008: Image Processing
Joseph M. Reinhardt; Josien P. W. Pluim, Editor(s)

© SPIE. Terms of Use
Back to Top