Ju Zhang, Presenter: Nothing to Disclose

Jacqui Hislop-Jambrich PhD, Abstract Co-Author: Employee, Toshiba Corporation

Duane Malcolm, Abstract Co-Author: Nothing to Disclose

C David L Thomas, Abstract Co-Author: Nothing to Disclose

Poul Nielsen, Abstract Co-Author: Nothing to Disclose

The femur and hip joint in particular are complicated structures that have both clinical and anthropological significance. The variability of surface structures among individuals especially in terms of gender makes the extraction of consistently reproducible measurements non-trivial and time consuming. Automatic image segmentation and meshing methods allow precisely-defined measurements to be taken from CT-volumes as part of an automated pipeline. We present initial findings from such a pipeline for obtaining morphometric measurements of the hip. We believe that the importance of this work lies in the eventual creation of a comprehensive databank that will be of use in the development of prosthetic devices and the tracking of disease and evolutionary morphometry.

A 16-row MDCT was used to acquire images on 55 human cadavers (24 male, 31 female). The outer femoral surface was automatically segmented and meshed with sub-voxel accuracy. Femoral head area, femoral axis length, neck angle, neck width, and subtrochanteric width were automatically measured on the mesh according to mathematical definitions based on mesh geometry. Errors with respect to manual measurements were between 2.4% and 7.6% on average. All automatic measurements except for neck angle showed significant differences between genders (p-value<0.001), which was reflected in the manual measurements.

We present this initial evaluation of five measurements with a view toward the creation of a macro-structural atlas of bones in the hip. The automated system shows good promise in terms of accuracy and sensitivity compared to manual measurements. We describe the variability of these measurements in our homogenous population with specific reference to gender to provide an overview of what is possible using current technology.

Automatic assessment of proximal femur morphometry has shown that variations are sufficient to warrant  the creation of a bank of detailed morphometric assessments of the hip. We believe that detailed knowledge of this complicated structure may be used to support the development of prosthetic devices and assist in diagnosing complex hip-based disorders.

Zhang, J, Hislop-Jambrich, J, Malcolm, D, Thomas, C, Nielsen, P, One Size in No Way Fits All - Quantifying Hip Variations by Automatic Morphometric Measurements from CT.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13026384.html