Backscatter measurement of cancellous bone using the ultrasound transit time spectroscopy

Recently, ultrasound transit time spectroscopy (UTTS) was proposed as a promising method for bone quantitative ultrasound measurement. Studies have showed that UTTS could estimate the bone volume fraction and other trabecular bone structure in ultrasonic through-transmission measurements. The goal of this study was to explore the feasibility of UTTS to be adapted in ultrasonic backscatter measurement and further evaluate the performance of backscattered ultrasound transit time spectrum (BS-UTTS) in the measurement of cancellous bone density and structure. First, taking ultrasonic attenuation into account, the concept of BS-UTTS was verified on ultrasonic backscatter signals simulated from a set of scatterers with different positions and intensities. Then, in vitro backscatter measurements were performed on 26 bovine cancellous bone specimens. After a logarithmic compression of the BS-UTTS, a linear fitting of the log-compressed BS-UTTS versus ultrasonic propagated distance was performed and the slope and intercept of the fitted line for BS-UTTS were determined. The associations between BS-UTTS parameters and cancellous bone features were analyzed using simple linear regression. The results showed that the BS-UTTS could make an accurate deconvolution of the backscatter signal and predict the position and intensity of the simulated scatterers eliminating phase interference, even the simulated backscatter signal was with a relatively low signal-to-noise ratio. With varied positions and intensities of the scatterers, the slope of the fitted line for the log-compressed BS-UTTS versus ultrasonic propagated distance (i.e., slope of BS-UTTS for short) yield a high agreement (r2 = 99.84%-99.96%) with ultrasonic attenuation in simulated backscatter signal. Compared with the high-density cancellous bone, the low-density specimen showed more abundant backscatter impulse response in the BS-UTTS. The slope of BS-UTTS yield a significant correlation with bone mineral density (r = 0.87; p < 0.001), BV/TV (r = 0.87; p < 0.001), and cancellous bone microstructures (r up to 0.87; p < 0.05). The intercept of BS-UTTS was also significantly correlated with bone densities (r = -0.87; p < 0.001) and trabecular structures (|r|=0.43-0.80; p < 0.05). However, the slope of the BS-UTTS underestimated attenuation when measurements were performed experimentally. In addition, a significant non-linear relationship was observed between the measured attenuation and the attenuation estimated by the slope of the BS-UTTS. This study demonstrated that the UTTS method could be adapted to ultrasonic backscatter measurement of cancellous bone. The derived slope and intercept of BS-UTTS could be used in the measurement of bone density and microstructure. The backscattered ultrasound transit time spectroscopy might have potential in the diagnosis of osteoporosis in the clinic.

Mechanisms of Interaction of Ultrasound With Cancellous Bone: A Review

Ultrasound is now a clinically accepted modality in the management of osteoporosis. The most common commercial clinical devices assess fracture risk from measurements of attenuation and sound speed in cancellous bone. This review discusses fundamental mechanisms underlying the interaction between ultrasound and cancellous bone. Because of its two-phase structure (mineralized trabecular network embedded in soft tissue-marrow), its anisotropy, and its inhomogeneity, cancellous bone is more difficult to characterize than most soft tissues. Experimental data for the dependencies of attenuation, sound speed, dispersion, and scattering on ultrasound frequency, bone mineral density, composition, microstructure, and mechanical properties are presented. The relative roles of absorption, scattering, and phase cancellation in determining attenuation measurements in vitro and in vivo are delineated. Common speed of sound metrics, which entail measurements of transit times of pulse leading edges (to avoid multipath interference), are greatly influenced by attenuation, dispersion, and system properties, including center frequency and bandwidth. However, a theoretical model has been shown to be effective for correction for these confounding factors in vitro and in vivo. Theoretical and phantom models are presented to elucidate why cancellous bone exhibits negative dispersion, unlike soft tissue, which exhibits positive dispersion. Signal processing methods are presented for separating "fast" and "slow" waves (predicted by poroelasticity theory and supported in cancellous bone) even when the two waves overlap in time and frequency domains. Models to explain dependencies of scattering on frequency and mean trabecular thickness are presented and compared with measurements. Anisotropy, the effect of the fluid filler medium (marrow in vivo or water in vitro), phantoms, computational modeling of ultrasound propagation, acoustic microscopy, and nonlinear properties in cancellous bone are also discussed.

Modeling ultrasound propagation through material of increasing geometrical complexity

Ultrasound is increasingly being recognized as a neuromodulatory and therapeutic tool, inducing a broad range of bio-effects in the tissue of experimental animals and humans. To achieve these effects in a predictable manner in the human brain, the thick cancellous skull presents a problem, causing attenuation. In order to overcome this challenge, as a first step, the acoustic properties of a set of simple bone-modeling resin samples that displayed an increasing geometrical complexity (increasing step sizes) were analyzed. Using two Non-Destructive Testing (NDT) transducers, we found that Wiener deconvolution predicted the Ultrasound Acoustic Response (UAR) and attenuation caused by the samples. However, whereas the UAR of samples with step sizes larger than the wavelength could be accurately estimated, the prediction was not accurate when the sample had a smaller step size. Furthermore, a Finite Element Analysis (FEA) performed in ANSYS determined that the scattering and refraction of sound waves was significantly higher in complex samples with smaller step sizes compared to simple samples with a larger step size. Together, this reveals an interaction of frequency and geometrical complexity in predicting the UAR and attenuation. These findings could in future be applied to poro-visco-elastic materials that better model the human skull.

Transducer impulse response correction for a deconvolution derived ultrasound transit time spectrum

Propagation of ultrasound through a complex composite sample may exhibit phase interference between two or more sonic-rays if differences in transit time are less than the pulse length. The transit time spectrum of a test sample, equivalent to its impulse response, was derived through active-set deconvolution of ultrasound signals with, and without, the test sample. The aim of this study was to test the hypothesis that in cases where only the transmit ultrasound transducer's digitally-coded excitation signal is available, hence not the input ultrasound signal without the test sample, incorporation of the transducer impulse response may increase both accuracy and precision of ultrasound transit time spectroscopy. A digital 1 MHz sinusoid signal was used to create an ultrasound pulse that was propagated through a 5 step-wedge acrylic sample immersed in water. Transit time spectra were obtained through deconvolution utilising an ultrasound input signal, along with a digital input signal, with and without incorporation of the transducer impulse response. Incorporation of the transducer impulse response reduced a quantitative measure of noise-to-signal ratio by a factor of 12. The paper has demonstrated the potential for increased accuracy and precision of transit time spectroscopy when the transducer impulse response is incorporated within active-set deconvolution analysis.

Establishing sheep as an experimental species to validate ultrasound-mediated blood-brain barrier opening for potential therapeutic interventions

Rationale: Treating diseases of the brain such as Alzheimer's disease (AD) is challenging as the blood-brain barrier (BBB) effectively restricts access of a large number of potentially useful drugs. A potential solution to this problem is presented by therapeutic ultrasound, a novel treatment modality that can achieve transient BBB opening in species including rodents, facilitated by biologically inert microbubbles that are routinely used in a clinical setting for contrast enhancement. However, in translating rodent studies to the human brain, the presence of a thick cancellous skull that both absorbs and distorts ultrasound presents a challenge. A larger animal model that is more similar to humans is therefore required in order to establish a suitable protocol and to test devices. Here we investigated whether sheep provide such a model.

Methods: In a stepwise manner, we used a total of 12 sheep to establish a sonication protocol using a spherically focused transducer. This was assisted by ex vivo simulations based on CT scans to establish suitable sonication parameters. BBB opening was assessed by Evans blue staining and a range of histological tests.

Results: Here we demonstrate noninvasive microbubble-mediated BBB opening through the intact sheep skull. Our non-recovery protocol allowed for BBB opening at the base of the brain, and in areas relevant for AD, including the cortex and hippocampus. Linear time-shift invariant analysis and finite element analysis simulations were used to optimize the position of the transducer and to predict the acoustic pressure and location of the focus.

Conclusion: Our study establishes sheep as a novel animal model for ultrasound-mediated BBB opening and highlights opportunities and challenges in using this model. Moreover, as sheep develop an AD-like pathology with aging, they represent a large animal model that could potentially complement the use of non-human primates.

The dependence of broadband ultrasound attenuation on phase interference in thin plates of variable thickness and curvature: a comparison of experimental measurement and computer simulation

The measurement of broadband ultrasound attenuation describes the linear increase in ultrasound attenuation with frequency (dB/MHz); this is generally performed at the calcaneus, consisting of a high proportion of metabolically active cancellous bone. Although broadband ultrasound attenuation is not routinely implemented within clinical management since it cannot provide a reliable estimation of bone mineral density and hence clinical definition of osteopenia and osteoporosis, it offers a reliable means to predict osteoporotic fracture risk. One of the potential factors that can influence the accuracy of broadband ultrasound attenuation measurement is the effect of cortical end plates. This study aimed to explore this, performing a comparison of experimental study and computer simulation prediction. A total of three categories of thin discs were three-dimensional (3D) printed to replicate cortical shells of (1) variable constant thickness (planar), (2) variable constant thickness (curved), and (3) variable thickness. A through-transmission technique was used, where two single-element, unfocused, 1 MHz broadband transducers, as utilised clinically, were positioned coaxially in a cylindrical holder and immersed in water. Both quantitative and qualitative analyses demonstrated that broadband ultrasound attenuation measurements of the 'planar' and 'curved' discs were not statistically different (p-values > 0.01). A cyclic relationship between broadband ultrasound attenuation and disc thickness was observed; this was replicated within a computer simulation of phase interference created by a double-reflection echo within each disc (R² = 97.0%). Variable-thickness discs provided broadband ultrasound attenuation measurements ranging between 31.6 ± 0.1 and 40.60 ± 0.1 dB/MHz. Again applying the double-reflection echo simulation, a high level of agreement between experimental and simulation was recorded (R² = 93.4%). This study indicates that the cortical end plate can significantly affect the broadband ultrasound attenuation measurement of cancellous bone as a result of phase interference and, therefore, warrants further investigation to minimise its effect on clinical assessment.

Bone volume fraction and structural parameters for estimation of mechanical stiffness and failure load of human cancellous bone samples; in-vitro comparison of ultrasound transit time spectroscopy and X-ray μCT

Conventional mechanical testing is the 'gold standard' for assessing the stiffness (N mm^-1) and strength (MPa) of bone, although it is not applicable in-vivo since it is inherently invasive and destructive. The mechanical integrity of a bone is determined by its quantity and quality; being related primarily to bone density and structure respectively. Several non-destructive, non-invasive, in-vivo techniques have been developed and clinically implemented to estimate bone density, both areal (dual-energy X-ray absorptiometry (DXA)) and volumetric (quantitative computed tomography (QCT)). Quantitative ultrasound (QUS) parameters of velocity and attenuation are dependent upon both bone quantity and bone quality, although it has not been possible to date to transpose one particular QUS parameter into separate estimates of quantity and quality. It has recently been shown that ultrasound transit time spectroscopy (UTTS) may provide an accurate estimate of bone density and hence quantity. We hypothesised that UTTS also has the potential to provide an estimate of bone structure and hence quality. In this in-vitro study, 16 human femoral bone samples were tested utilising three techniques; UTTS, micro computed tomography (μCT), and mechanical testing. UTTS was utilised to estimate bone volume fraction (BV/TV) and two novel structural parameters, inter-quartile range of the derived transit time (UTTS-IQR) and the transit time of maximum proportion of sonic-rays (TTMP). μCT was utilised to derive BV/TV along with several bone structure parameters. A destructive mechanical test was utilised to measure the stiffness and strength (failure load) of the bone samples. BV/TV was calculated from the derived transit time spectrum (TTS); the correlation coefficient (R²) with μCT-BV/TV was 0.885. For predicting mechanical stiffness and strength, BV/TV derived by both μCT and UTTS provided the strongest correlation with mechanical stiffness (R²=0.567 and 0.618 respectively) and mechanical strength (R²=0.747 and 0.736 respectively). When respective structural parameters were incorporated to BV/TV, multiple regression analysis indicated that none of the μCT histomorphometric parameters could improve the prediction of mechanical stiffness and strength, while for UTTS, adding TTMP to BV/TV increased the prediction of mechanical stiffness to R²=0.711 and strength to R²=0.827. It is therefore envisaged that UTTS may have the ability to estimate BV/TV along with providing an improved prediction of osteoporotic fracture risk, within routine clinical practice in the future.

Solid volume fraction estimation of bone:marrow replica models using ultrasound transit time spectroscopy

The acceptance of broadband ultrasound attenuation (BUA) for the assessment of osteoporosis suffers from a limited understanding of both ultrasound wave propagation through cancellous bone and its exact dependence upon the material and structural properties. It has recently been proposed that ultrasound wave propagation in cancellous bone may be described by a concept of parallel sonic rays; the transit time of each ray defined by the proportion of bone and marrow propagated. A Transit Time Spectrum (TTS) describes the proportion of sonic rays having a particular transit time, effectively describing the lateral inhomogeneity of transit times over the surface aperture of the receive ultrasound transducer. The aim of this study was to test the hypothesis that the solid volume fraction (SVF) of simplified bone:marrow replica models may be reliably estimated from the corresponding ultrasound transit time spectrum. Transit time spectra were derived via digital deconvolution of the experimentally measured input and output ultrasonic signals, and compared to predicted TTS based on the parallel sonic ray concept, demonstrating agreement in both position and amplitude of spectral peaks. Solid volume fraction was calculated from the TTS; agreement between true (geometric calculation) with predicted (computer simulation) and experimentally-derived values were R(2)=99.9% and R(2)=97.3% respectively. It is therefore envisaged that ultrasound transit time spectroscopy (UTTS) offers the potential to reliably estimate bone mineral density and hence the established T-score parameter for clinical osteoporosis assessment.

Pulse-echo ultrasound transit time spectroscopy: A comparison of experimental measurement and simulation prediction

Considering ultrasound propagation through complex composite media as an array of parallel sonic rays, a comparison of computer-simulated prediction with experimental data has previously been reported for transmission mode (where one transducer serves as transmitter, the other as receiver) in a series of 10 acrylic step-wedge samples, immersed in water, exhibiting varying degrees of transit time inhomogeneity. In this study, the same samples were used but in pulse-echo mode, where the same ultrasound transducer served as both transmitter and receiver, detecting both ‘primary’ (internal sample interface) and ‘secondary’ (external sample interface) echoes. A transit time spectrum was derived, describing the proportion of sonic rays with a particular transit time. A computer simulation was performed to predict the transit time and amplitude of various echoes created, and compared with experimental data. Applying an amplitude-tolerance analysis, 91.7% ± 3.7% of the simulated data were within ±1 standard deviation of the experimentally measured amplitude-time data. Correlation of predicted and experimental transit time spectra provided coefficients of determination (R2%) ranging from 100.0% to 96.8% for the various samples tested. The results acquired from this study provide good evidence for the concept of parallel sonic rays. Furthermore, deconvolution of experimental input and output signals has been shown to provide an effective method to identify echoes otherwise lost due to phase cancellation. Potential applications of pulse-echo ultrasound transit time spectroscopy include improvement of ultrasound image fidelity by improving spatial resolution and reducing phase interference artefacts.