RSNA 2003 Scientific Papers > 3D Structure Analysis of High-Resolution Magnetic ...
  Scientific Papers
  SESSION: Musculoskeletal (Metabolic Bone Diseases: Osteoporosis)

3D Structure Analysis of High-Resolution Magnetic Resonance Imaging of the Proximal Femur in Relationship with Biomechanical Bone Strength in Vitro

  DATE: Tuesday, December 02 2003
  START TIME: 03:40 PM
  END TIME: 03:47 PM
  CODE: J21-850

Dirk Mueller MD
Munich Germany
Thomas Link MD
Roberto Monetti PhD
Gregor Morfill PhD
Ernst Rummeny MD
Christoph Raeth PhD

Magnetic resonance (MR), high resolution
Magnetic resonance (MR), image processing

Purpose: In this study we applied the newly developed 3D-based scaling index method to high resolution magnetic resonance (HR-MR) images of the proximal femur. These trabecular bone structure measures were compared with bone mineral density (BMD) as well as standard morphological 2D parameters in the prediction of bone strength in vitro.

Methods and Materials: Coronal HR-MR images of the proximal femur were obtained in 30 fresh human proximal femur specimens using a 1.5 T MR scanner equipped with 22 mT/m gradients and a prototype 2 element phased array coil. A 3D-spin-echo sequence with a matrix of 384 x 512 was used with a field of view (FOV) of 75 x 100 mm, yielding a voxel size of 0.195 x 0.195 x 0.9 mm and 0.195 x 0.195 x 0.3 mm. After segmentation with a region growing technique the structure analysis was performed using algorithms based on our recently introduced local 3D scaling index method as well as morphological 2D parameters. In addition BMD measurements were obtained using dual-energy-X-ray-absorptiometry (DXA). All specimens were tested destructively in a biomechanical materials testing machine to determine the maximum compressive strength (MCS).

Results: Using the 3D structure measures obtained by the local 3D scaling index method R-values up to 0.73 (p<0.01) were calculated. Comparable correlations were determined between BMD and MCS with R-values up to 0.74 (p<0.01). R-values up to 0.69 (p<0.01) were computed for the standard 2 D parameters and MCS. Using multivariate regression analysis and combining structure parameters and BMD, improved correlations versus MCS significantly (up to R = 0.94; p< 0.01).

Conclusion: The results of this study suggest that in an experimental setting structure parameters based on our new local 3D scaling index method determined in high resolution MR images of the proximal femur can be used to predict the mechanical properties of trabecular bone in vitro. However, due to limitations of the depiction of the trabecular bone structure using high resolution MRI even the 3D- technique may not be ideally suited to assess bone strength, though it provides substantial additional information to BMD as shown by multivariate regression analysis.




Questions about this event email: