Experimental and theoretical modelling of cancellous bone structure for the ultrasonic assessment of osteoporosis.

This study describes the use of the Stereolithography (SL) process to produce what are probably the first realistic mimics of trabecular bone, with a precise predefined structure, for use in the modelling of BUA in the calcaneus. This is part of the ongoing investigation of the role of ultrasound in the assessment of Osteoporosis.It has been shown that Broadband Ultrasonic Attenuation (BUA) measurements of the calcaneus have a similar diagnostic accuracy to ionising radiation measurements of spinal bone density, in predicting fracture risk due to Osteoporosis. There is evidence that BUA measurements are related to structure as well as bone mineral density and if this structural information can be isolated, osteoporotic fracture risk predictions will be enhanced.The actual ultrasonic attenuation mechanisms occurring in cancellous bone (thought to be mainly scattering) are not fully understood at present and further research is hampered by the lack of suitable samples. The available bone samples tend to be of elderly or diseased origin and so there is, therefore, a need for a bone mimic with similar acoustic properties to cancellous bone whose structure can be carefully controlled. Previous attempts to model cancellous bone using perspex and a composite epoxy resin did not produce acoustic characteristics sufficiently close enough to natural tissue and so an alternative approach was required.The Stereolithography (SL) process is a form of rapid prototyping that allows complex solid objects to be made directly from 3D computer models, by laser scanning of liquid light cured resins. The smallest wall thickness of this system is 0.3mm A literature search of calcaneal trabecular bone was unable to supply the detailed structural information required to produce a design for a cancellous bone model, and so a modified histomorphometrical analysis procedure was used.Two cancellous core samples, one high density and one low density, were analysed using digital imaging techniques and the majority of the trabeculea in both models were found to be less than the 0.3 mm resolution of the SL system. This ruled out a full 3D reconstruction of the samples and so a model, similar to but not exactly like, cancellous bone was designed, based on the skeleton of the low density sample.The basic model, using castor oil as a marrow mimic, was found to be stable and the ultrasonic characteristics were compatible with natural tissue.In order to look at the relationship between ultrasonic parameters and porosity, a family of models, based on the basic design, was produced. The porosity of the models was 70% (basic design) 50%, 30%, 80% and 85%.BUA showed a non linear relationship with porosity with a minima at 0 and 100%, and a maxima at 70%. This is very similar to natural tissue which has a peak BUA at 75%. Velocity showed a linear relationship with porosity and the values, within experimental limits, followed the theoretical model. These results validated the use of the SL process as a modelling medium.Two further models, with a different structure but the same porosity as the basic model, were produced, and the three models showed a 7% spread of BUA, confirming the structural dependence of BUA.Using the precise structural details extracted from the computer model, the best structural indicator of BUA for the SL models was found to be Trabecular Surface Area, with a high correlation of r -2= 0.92.This would seem to confirm that the attenuation mechanism in cancellous bone is governed by architectural complexity as well as porosity.