- Email: [email protected]
2084792 Investigation of the Effects of Myocardial Anisotropy for Shear Wave Elastography Using Acoustic Radiation Force and Harmonic Vibration
Abstracts
sensitive the microstructure difference. The model-based parameter estimates yielded significantly higher tumor cell clustering effect for EHS than for 4T1, consistent with the observed difference in microstructure of the two tumor types. Conclusions: The proposed models are capable of characterizing tumor microstructure by predicting the spatial distribution of tumor cells, which could lead to improved diagnosis and classification capability. [Supported by NIH CA111289].
2086603 Fast and Slow Wave Detection in Cancellous Bone In Vitro Using Bandlimited Deconvolution and Prony’s Method Keith Wear,1 Yoskhiki Nagatani,2 Katsunori Mizuno,2 Mami Matsukawa2 1Center for Devices and Radiological Health, Food and Drug Administration, Silver Spring, MD, United States; 2Doshisha University, Kyoto, Japan Objectives: Through-transmission measurements in cancellous bone often reveal two longitudinal waves. The velocities and amplitudes of ‘‘fast’’ and ‘‘slow’’ waves are related to bone microstructure and composition, and therefore may provide diagnostic information related to fracture-risk. One commercial system provides diagnostic measurements based on properties of fast and slow waves in the wrist (Breban et al., Bone, 46, 1620-1625, 2010). Unfortunately, these waves often overlap in time and frequency domains. The OBJECTIVES: of this work was to test signal processing algorithms for separating fast and slow waves in through-transmission signals measured in cancellous bone. Methods: Signal loss and phase velocity for fast and slow waves were measured as functions of propagation depth in a bovine cancellous bone sample. Fast and slow waves were investigated for thicknesses ranging from 7 to 12 mm using bandlimited deconvolution (Wear, J. Acoust. Soc. Am., 135, 2102-2112, 2014) and the modified least-squares Prony’s method with curve fitting (MLSP+CF) (Wear, J. Acoust. Soc. Am., 133, 2490-2501, 2013). Results: Bandlimited deconvolution consistently isolated two waves with linear-with-frequency attenuation coefficients, as evidenced by high correlation coefficients between attenuation coefficient and frequency: 0.997 6 0.002 (fast wave) and 0.986 6 0.013 (slow wave) (mean 6 standard deviation). Average root-mean-squared (RMS) differences between the two algorithms for phase velocities were 5 m/s (fast wave, 350 kHz) and 13 m/s (slow wave, 750 kHz). Average RMS differences for signal loss were 1.6 dB (fast wave, 350 kHz) and 0.4 dB (slow wave, 750 kHz). Phase velocities for thickness 5 10 mm were 1726 m/s (fast wave, 350 kHz) and 1455 m/s (slow wave, 750 kHz). Conclusions: Results show support for the model of two waves with linear-with frequency attenuation, successful isolation of fast and slow waves, and good agreement between bandlimited deconvolution and MLSP+CF.
2084792 Investigation of the Effects of Myocardial Anisotropy for Shear Wave Elastography Using Acoustic Radiation Force and Harmonic Vibration Matthew Urban,1 Bo Qiang,1 Ivan Nenadic,1 Shigao Chen,1 James Greenleaf1 1Physiology and
S97
Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN, United States Objectives: Various methods have been developed that use either acoustic radiation force (ARF) or external vibration (EV) to evaluate the material properties of the myocardium. The complicated myocardial structure presents a unique challenge to wave propagation methods because the muscle fibers change direction through the thickness of the left ventricular wall. Methods: An analysis of the effects of myocardial anisotropy for different types of shear wave excitations was conducted. A finite element model in the shape of cylinder with 100 mm diameter was constructed that contained 20 layers of 1 mm thickness with the top and bottom layers in contact with water. Each layer was defined as an elastic, transversely isotropic medium and the orientation of the muscle fibers was changed for each layer ranging from -50 to 80 from top to bottom. Each layer had a shear wave velocity (SWV) along the fibers of 5.0 m/s and a SWV across the fibers of 2.5 m/s. Harmonic excitations at 30, 50, 100, and 200 Hz and an impulsive force were used in a line source in the middle of the model to compare the differences in the wave propagation with different frequency content. Data were analyzed to obtain the group SWV from the motion data every 5 to evaluate the local anisotropy. Results: We evaluated the measured orientation of the fibers in each layer by evaluating the angle with the highest SWV. The 30 and 50 Hz results showed no variation in the measured orientation angle in the layers. The 100 and 200 Hz results showed some variation of the orientation with respect to the layer. The impulse simulation results showed good agreement with the true orientations except near the top and bottom boundaries. The values of SWV were found to have different levels of bias depending on the excitation. Conclusions: This multi-layered anisotropic model demonstrates how to resolve different anisotropic layers in the myocardium using ARF or EV while also revealing that using lower frequencies (30 or 50 Hz) results in measurements that are less sensitive to anisotropy variation through the thickness of the myocardial wall. We also studied the biases in SWV values measured with different excitations. 2083648 Robust and Ultrafast Detection of Shear Wave Motion Using Coded Excitation Pengfei Song,1 Matthew William Urban,1 Armando Manduca,1 James F. Greenleaf,1 Shigao Chen1 1Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN, United States Objectives: Recent advancements of plane wave imaging (PWI) greatly facilitated the development of shear wave elastography (SWE), thanks to the ultrafast frame rate (FR) and large field-of-view (FOV) of PWI. However, PWI suffers from poor penetration due to lack of focusing, and cannot provide robust shear wave (SW) motion detection in deep tissues and obese patients. This study systematically investigated the feasibility of using coded excitation (CE) PWI for SW detection, with the hypothesis that CE can provide superior detection penetration and SW signal-to-noiseratio (SNR) as compared to the conventional short pulse (CSP) signals.
sensitive the microstructure difference. The model-based parameter estimates yielded significantly higher tumor cell clustering effect for EHS than for 4T1, consistent with the observed difference in microstructure of the two tumor types. Conclusions: The proposed models are capable of characterizing tumor microstructure by predicting the spatial distribution of tumor cells, which could lead to improved diagnosis and classification capability. [Supported by NIH CA111289].
2086603 Fast and Slow Wave Detection in Cancellous Bone In Vitro Using Bandlimited Deconvolution and Prony’s Method Keith Wear,1 Yoskhiki Nagatani,2 Katsunori Mizuno,2 Mami Matsukawa2 1Center for Devices and Radiological Health, Food and Drug Administration, Silver Spring, MD, United States; 2Doshisha University, Kyoto, Japan Objectives: Through-transmission measurements in cancellous bone often reveal two longitudinal waves. The velocities and amplitudes of ‘‘fast’’ and ‘‘slow’’ waves are related to bone microstructure and composition, and therefore may provide diagnostic information related to fracture-risk. One commercial system provides diagnostic measurements based on properties of fast and slow waves in the wrist (Breban et al., Bone, 46, 1620-1625, 2010). Unfortunately, these waves often overlap in time and frequency domains. The OBJECTIVES: of this work was to test signal processing algorithms for separating fast and slow waves in through-transmission signals measured in cancellous bone. Methods: Signal loss and phase velocity for fast and slow waves were measured as functions of propagation depth in a bovine cancellous bone sample. Fast and slow waves were investigated for thicknesses ranging from 7 to 12 mm using bandlimited deconvolution (Wear, J. Acoust. Soc. Am., 135, 2102-2112, 2014) and the modified least-squares Prony’s method with curve fitting (MLSP+CF) (Wear, J. Acoust. Soc. Am., 133, 2490-2501, 2013). Results: Bandlimited deconvolution consistently isolated two waves with linear-with-frequency attenuation coefficients, as evidenced by high correlation coefficients between attenuation coefficient and frequency: 0.997 6 0.002 (fast wave) and 0.986 6 0.013 (slow wave) (mean 6 standard deviation). Average root-mean-squared (RMS) differences between the two algorithms for phase velocities were 5 m/s (fast wave, 350 kHz) and 13 m/s (slow wave, 750 kHz). Average RMS differences for signal loss were 1.6 dB (fast wave, 350 kHz) and 0.4 dB (slow wave, 750 kHz). Phase velocities for thickness 5 10 mm were 1726 m/s (fast wave, 350 kHz) and 1455 m/s (slow wave, 750 kHz). Conclusions: Results show support for the model of two waves with linear-with frequency attenuation, successful isolation of fast and slow waves, and good agreement between bandlimited deconvolution and MLSP+CF.
2084792 Investigation of the Effects of Myocardial Anisotropy for Shear Wave Elastography Using Acoustic Radiation Force and Harmonic Vibration Matthew Urban,1 Bo Qiang,1 Ivan Nenadic,1 Shigao Chen,1 James Greenleaf1 1Physiology and
S97
Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN, United States Objectives: Various methods have been developed that use either acoustic radiation force (ARF) or external vibration (EV) to evaluate the material properties of the myocardium. The complicated myocardial structure presents a unique challenge to wave propagation methods because the muscle fibers change direction through the thickness of the left ventricular wall. Methods: An analysis of the effects of myocardial anisotropy for different types of shear wave excitations was conducted. A finite element model in the shape of cylinder with 100 mm diameter was constructed that contained 20 layers of 1 mm thickness with the top and bottom layers in contact with water. Each layer was defined as an elastic, transversely isotropic medium and the orientation of the muscle fibers was changed for each layer ranging from -50 to 80 from top to bottom. Each layer had a shear wave velocity (SWV) along the fibers of 5.0 m/s and a SWV across the fibers of 2.5 m/s. Harmonic excitations at 30, 50, 100, and 200 Hz and an impulsive force were used in a line source in the middle of the model to compare the differences in the wave propagation with different frequency content. Data were analyzed to obtain the group SWV from the motion data every 5 to evaluate the local anisotropy. Results: We evaluated the measured orientation of the fibers in each layer by evaluating the angle with the highest SWV. The 30 and 50 Hz results showed no variation in the measured orientation angle in the layers. The 100 and 200 Hz results showed some variation of the orientation with respect to the layer. The impulse simulation results showed good agreement with the true orientations except near the top and bottom boundaries. The values of SWV were found to have different levels of bias depending on the excitation. Conclusions: This multi-layered anisotropic model demonstrates how to resolve different anisotropic layers in the myocardium using ARF or EV while also revealing that using lower frequencies (30 or 50 Hz) results in measurements that are less sensitive to anisotropy variation through the thickness of the myocardial wall. We also studied the biases in SWV values measured with different excitations. 2083648 Robust and Ultrafast Detection of Shear Wave Motion Using Coded Excitation Pengfei Song,1 Matthew William Urban,1 Armando Manduca,1 James F. Greenleaf,1 Shigao Chen1 1Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN, United States Objectives: Recent advancements of plane wave imaging (PWI) greatly facilitated the development of shear wave elastography (SWE), thanks to the ultrafast frame rate (FR) and large field-of-view (FOV) of PWI. However, PWI suffers from poor penetration due to lack of focusing, and cannot provide robust shear wave (SW) motion detection in deep tissues and obese patients. This study systematically investigated the feasibility of using coded excitation (CE) PWI for SW detection, with the hypothesis that CE can provide superior detection penetration and SW signal-to-noiseratio (SNR) as compared to the conventional short pulse (CSP) signals.