City Research Online

Detection capabilities of energy dispersive X-Ray diffraction with respect to in-vitro trabecular bone mineral loss

Allday, A. W. (2004). Detection capabilities of energy dispersive X-Ray diffraction with respect to in-vitro trabecular bone mineral loss. (Unpublished Doctoral thesis, City, University of London)


This study concerns the measurement of bone density pertaining to the management of osteoporosis. Bone density measurements may be absolute quantities for comparison to population norms or comparative measurements over time to ascertain bone loss rate. They provide important data to help physicians develop therapy programmes, determine fracture risk and to advise on preventative measures to those in higher risk groups.

This report primarily describes an investigation into bone mineral measurement utilising scattered x-radiation known as energy dispersive x-ray diffraction (EDXRD). Osteoporosis typically manifests itself in regions of trabecular bone to a greater extent than cortical bone and, as such, osteoporosis levels may be determined with greater accuracy by measuring trabecular bone density in isolation. EDXRD was used to accomplish this by defining a scattering volume within a region of trabecular bone.

The first part of this work is concerned with characterisations of the x-ray tube and detection system, which were found to be suitable for the EDXRE) experiments. Then follows the calculation of EDXRD experimental geometry with respect to the bone samples used, and construction of bone + marrow mix phantoms. Sample and phantom material dimensions were set at 15 x 15 x 15 mm3. Initial EDXRD and dual energy x-ray diffraction (DEXA) experiments using these phantoms are then described. Methods for the preparation and use of trabecular bone cubes for the main densitometry study were developed. This included the location of sample extraction from excised femoral heads (involving the medial trabecular region); the removal of sample marrow lipid (using a sequence of detergent and ethanol); a technique to simulate bone mineral density loss (using ethylenediamineterta-acetic acid (EDTA)); a method to assess demineralisation homogeneity (microdensitometry of sample radiographs); a method of calculating the density of EDXRD scanned volumes from within the bone cube samples (microdensitometry); and a procedure to introduce bone marrow substitute into the samples (using a vacuum pump and magnetic resonance imaging to determine success). The EDXRD diffractometer used for the main study is described, followed by a method to correct measured EDXRD scattered spectra for attenuation by the samples and any additional attenuation added to simulate clinical sites. This correction was achieved using ratios of transmitted spectra with and without the attenuating material in place.

Following this preparatory work, two EDXRD and DEXA studies are presented that involved trabecular bone sample cubes. Precision, accuracy and correlation of each modality was investigated. The first study used bone density measurements of one component bone sample systems (trabecular bone only with no marrow). With a potential maximum of 1.000, best case results for EDXRD were found to be 0.9510 (for precision), 0.9573 (volume measurement accuracy) and 0.8659 (correlation), and for DEXA, 0.9703 (precision), 0.6186 (volume measurement accuracy) and 0.9797 (correlation). Additional attenuation was detrimental to EDXRD precision (0.8909) and volume measurement accuracy (0.8044), with slight improvement to correlation (0.8799), whereas DEXA precision was marginally improved (0.9764), volume measurement accuracy deteriorated significantly (0.1688) and correlation deteriorated marginally (0.9685). This study was then extended by using demineralised one component bone samples.

The second study consisted of two component bone sample system measurements (trabecular bone with marrow substitute in the trabecular spaces), along with additional attenuation to simulate typical calcaneus, radius and femoral head measurement sites.

Best case calcaneus results for EDXRD were 0.9692 (precision), 0.9156 (volume measurement accuracy) and 0.9469 (correlation), and for DEXA, 0.9782 (precision), 0.4761 (volume measurement accuracy) and 0.9785 (correlation). Best case radius results for EDXRD were 0.9357 (precision), 0.8584 (volume measurement accuracy) and 0.9356 (correlation), and for DEXA, 0.9776 (precision), 0.1541 (volume measurement accuracy) and 0.9837 (correlation). Best case femoral head results for EDXRD were 0.9357 (precision), 0.5888 (volume measurement accuracy) and 0.9464 (correlation), and for DEXA, 0.9866 (precision) and 0.9785 (correlation). DEXA volume measurement accuracy for the femoral head simulation was too low to calculate. DEXA precision (which was fairly consistent for each attenuation condition) outperformed that of EDXRD (which deteriorated with increased attenuation). DEXA correlation (fairly consistent for each attenuation condition) outperformed that of EDXRD (also fairly consistent). EDXRD volume measurement accuracy (which deteriorates with increased attenuation) clearly outperforms that of DEXA. (DEXA accuracy is presented here in terms of measuring specific bone volumes, which the method is unable to achieve and should not therefore be considered a failure of DEXA.)

An investigation into radiation effective dose necessary for useful EDXRD results was performed using thermoluminescent dosimetry (TLD), based on the method employed in this project for the two component bone sample system. Results suggest an EDXRD effective dose (568 pSv at 70 kVp, 800 mAs) of around 142 times that of DEXA (4 pSv from typical published data).

An examination of bone sample strength using the one component bone samples concluded the main study. Elastic moduli were calculated from results of cube compression in all three orientations. Trabecular bone cube orientation was shown to have a differing influence on elastic modulus for different samples, suggesting bone strength is determined by trabecular architecture in addition to density. R2 correlation between measured bone densities and elastic moduli in the medial trabecular region direction (0.0814) was found to be approximately half of that for the other two orientations (0.1698 and 0.1568).

The above results suggest that EDXRD is most suited to the measurement of absolute bone density values for specific bone volumes but that DEXA demonstrates better results in terms of precision and correlation. This, along with a greatly reduced dose cost, makes DEXA much more suited to the clinical environment. However suggestions are offered which include EDXRD dose reduction methods for potential clinical applications

Publication Type: Thesis (Doctoral)
Subjects: R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine
R Medicine > RZ Other systems of medicine
Departments: School of Health & Psychological Sciences > Midwifery & Radiography
School of Health & Psychological Sciences > School of Health & Psychological Sciences Doctoral Theses
Doctoral Theses
Text - Accepted Version
Download (14MB) | Preview



Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login