Final Activity Report Summary - QUSOB (Ultrasonic assessment of osteoporosis in cancellous bone)
The project was focused on coupled experimental, numerical and theoretical studies of ultrasonic wave propagation in cancellous bones, aiming at the validation of theoretical models of elastic wave propagation in such materials. It was proposed to achieve the goal by:
(i) analysis of the data obtained from ultrasonic studies, numerical simulations of wave propagation in real 3D trabecular bone structures, reconstructed from micro-computed tomography; and
(ii) analysis of existing/developed models of propagation of elastic waves in heterogeneous materials.
Within the project, 26 slabs of cancellous bones were prepared from fresh human femoral condyles. The main part of the project was concentrated on the ultrasonic (US) measurements performed on intact specimens (i.e. with marrow inside) and subsequently on water- and alcohol-saturated samples (i.e. the marrow was removed). One of the key point of the studies was the comparison of the site matched ultrasonic data (signals, wave parameters, etc.) obtained from the same specimen at different frequencies (0.5 1, and 2 MHz) and saturated with different fluids.
It was found that, the changes of absorption conditions (by replacement marrow by water in the pores) in the fluid or at the fluid / trabeculae interphases (decrease of the friction) did not lead to expected decrease of the values of attenuation coefficient. Moreover, significant alteration of scattering conditions (acoustic impedances at the fluid / trabeculae by replacement of the marrow/water by alcohol) also was not reflected in the attenuation.
Second important part of the project was focused on the evaluation of structural parameters of the cancellous bones based on the data acquired from micro-computed tomography. The reconstructed 3D models of bone micro-architecture, obtained from the same bone regions of interest (ROI's) as the US studies, were used twofold: (i) to calculate structural parameters and (ii) as an input data for numerical simulation of elastic wave propagation. The set of structural parameters obtained from the ROI's, site matched with US studies, shows less than 10 % bone volume fraction (BV/TV) of this area. Moreover, the range of values of the structure model index (1.5 - 2) and degree of anisotropy (1.3 - 2.3) indicate that the structure is not highly oriented.
The results of numerical simulations (signals transmitted through 3D bone micro structures) were compared with results of ultrasonic experiments. The main conclusion of these studies is that the attenuation coefficient increases both in the experiments and simulations as BV/TV increase. The simulated and measured phase velocities are very close, particularly for higher (1 & 2 MHz) frequencies.
Until the end of the project, there was not possible to validate or invalidate theoretical model. The mostly used macroscopic Biot's model (which doesn't include scattering effects) is very sensitive to the fluid viscosity. Change of the fluid viscosity (three orders of magnitude) is reflected in theoretical attenuation, while in the experiments practically no difference is observed. In contrary, the attenuation predicted by the scattering model (Faran model) is sensitive to the elastic properties of the fluid filling pores. Accordingly to such models, attenuation of alcohol saturated bones should be higher than of water-filled specimens, while in the experiments rather opposite behaviour is observed.
In summary, despite the lack of clear identification of the attenuation mechanisms in cancellous bones characterised by less than 10 % values of BV/TV, the studies concerning this issue will be continued. The ultrasonic data were not fully exploited yet due to unforeseen problem of the separation of mixed waveforms of the transmitted radio-frequency signals measured in the more dense ROI's. The waveform overlapping preclude any reliable analysis within these ROI's, therefore the problem of wave separation is currently under investigation.
(i) analysis of the data obtained from ultrasonic studies, numerical simulations of wave propagation in real 3D trabecular bone structures, reconstructed from micro-computed tomography; and
(ii) analysis of existing/developed models of propagation of elastic waves in heterogeneous materials.
Within the project, 26 slabs of cancellous bones were prepared from fresh human femoral condyles. The main part of the project was concentrated on the ultrasonic (US) measurements performed on intact specimens (i.e. with marrow inside) and subsequently on water- and alcohol-saturated samples (i.e. the marrow was removed). One of the key point of the studies was the comparison of the site matched ultrasonic data (signals, wave parameters, etc.) obtained from the same specimen at different frequencies (0.5 1, and 2 MHz) and saturated with different fluids.
It was found that, the changes of absorption conditions (by replacement marrow by water in the pores) in the fluid or at the fluid / trabeculae interphases (decrease of the friction) did not lead to expected decrease of the values of attenuation coefficient. Moreover, significant alteration of scattering conditions (acoustic impedances at the fluid / trabeculae by replacement of the marrow/water by alcohol) also was not reflected in the attenuation.
Second important part of the project was focused on the evaluation of structural parameters of the cancellous bones based on the data acquired from micro-computed tomography. The reconstructed 3D models of bone micro-architecture, obtained from the same bone regions of interest (ROI's) as the US studies, were used twofold: (i) to calculate structural parameters and (ii) as an input data for numerical simulation of elastic wave propagation. The set of structural parameters obtained from the ROI's, site matched with US studies, shows less than 10 % bone volume fraction (BV/TV) of this area. Moreover, the range of values of the structure model index (1.5 - 2) and degree of anisotropy (1.3 - 2.3) indicate that the structure is not highly oriented.
The results of numerical simulations (signals transmitted through 3D bone micro structures) were compared with results of ultrasonic experiments. The main conclusion of these studies is that the attenuation coefficient increases both in the experiments and simulations as BV/TV increase. The simulated and measured phase velocities are very close, particularly for higher (1 & 2 MHz) frequencies.
Until the end of the project, there was not possible to validate or invalidate theoretical model. The mostly used macroscopic Biot's model (which doesn't include scattering effects) is very sensitive to the fluid viscosity. Change of the fluid viscosity (three orders of magnitude) is reflected in theoretical attenuation, while in the experiments practically no difference is observed. In contrary, the attenuation predicted by the scattering model (Faran model) is sensitive to the elastic properties of the fluid filling pores. Accordingly to such models, attenuation of alcohol saturated bones should be higher than of water-filled specimens, while in the experiments rather opposite behaviour is observed.
In summary, despite the lack of clear identification of the attenuation mechanisms in cancellous bones characterised by less than 10 % values of BV/TV, the studies concerning this issue will be continued. The ultrasonic data were not fully exploited yet due to unforeseen problem of the separation of mixed waveforms of the transmitted radio-frequency signals measured in the more dense ROI's. The waveform overlapping preclude any reliable analysis within these ROI's, therefore the problem of wave separation is currently under investigation.