- Email: [email protected]
2263
P142
Ultrasound in Medicine and Biology
metastases economically and effectively, also reducing the number of more expensive imaging procedures.
Volume 32, Number 5S, 2006 Conclusions: The method using the mean normal stress as unknown is effective. 2262
DIAGNOSTIC US 2260 Transient ultrasound radiation force elastography: A preliminary comparison with surface palpation elastography Bamber JC, Lima DM, Duck FA, Shipley JA, Xu L, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, United Kingdom of Great Britain and Northern Ireland; Bath Royal United Hospital, United Kingdom of Great Britain and Northern Ireland Objectives: Ultrasound elasticity imaging is being developed for evaluating the mechanical properties of soft tissues in vivo, but each research group tends to specialize in a specific approach. Our objective was to assess the advantages and disadvantages of generating tissue strain by deep loading with transient acoustic radiation force, versus the use of nearly static surface loading. Methods: Experiments with gelatin phantoms and finite element models were employed to measure the performance of pseudo-static freehand elastography and radiation force elastography. Results: Relative to surface loading, radiation force images are less susceptible to artefacts associated with boundary conditions, elastic modulus inhomogeneities and pre-existing tissue motion. Furthermore, they may possess a reduced rate of decay of strain signal-to-noise ratio with depth, and demonstrate an improved contrast-transfer-efficiency, particularly for inclusions that have negative modulus contrast or that are disconnected from the background by a low friction boundary. It takes longer to scan using radiation force loading and the elastograms may show contrast for ultrasonic attenuation and absorption, as well as for tissue stiffness. Conclusions: Transient deep loading possess advantages over surface loading for elastography but both approaches may have a role to play, and further comparative evaluation in vivo is needed. 2261 Multidimensional reconstruction/imaging of shear modulus distribution of living soft tissues Sumi C, Sophia University, Japan Objectives: Previously, we reported a shear modulus reconstruction method utilizing the typical value of Poisson’s ratio. However, as reconstruction errors were confirmed due to the difference between the original value and the set value, we proposed a method to reconstruct Poisson’s ratio as well. Furthermore, we proposed to reconstruct density as well to deal with dynamic deformation. However, due to the tissue incompressibility, the reconstruction of shear modulus, Poisson’s ratio and density become unstable. Methods: In this report, to stabilize these reconstructions, we report a new reconstruction method using the mean normal stress as unknown. This method also allows stable reconstructions of shear modulus and density under the condition that Poisson’s ratio remains unknown. The effectiveness is verified through 2D phantom experiments and 3D simulations. [The cubic phantom (50 mm sides) includes a spherical inclusion (radius, 5 mm) at the center (depth, 25.0 mm) having different shear modulus and Poisson’s ratio from those of the surrounding medium. Density is uniformly set. Noise-filled measurement data of displacement vector were simulated by adding white noise to the calculated raw displacement data.] Results: The new method yielded stable quantitative reconstructions for both the simulations and phantom experiments.
On useful ultrasound image processing techniques for the diagnosis of fine blood vessels Ito M, Yamada A, Nakamura K, Kato K, Kobayashi A, Kuroshima N, Tokyo Denki University, Japan; Tokyo Univ. Agriculture and Tech., Japan; Mitaka Kohki Co., Japan; Aloka Co., Japan; Micro Design Inc., Japan; Teikyo Univ., School of Medicine, Japan Objectives: In order to extract the contours of vessels of various sizes, robust three-dimensional imaging techniques are essential to the resultant successive scanned data. Our aim is to access to the vessels of about 300 to 500 microns. We focus on the speckle reduction and edge enhancement by adaptive morphological filters. Methods: The system with a high-frequency probe of 20 to 40 MHz, scan the region of 10⫻10⫻30 (mm in size). The accessed volume data have 128 cross sections of 512⫻512 pixels of 256 gray levels. Ultrasound longitudinal cross sectional images of blood vessels as well as 3D contours can be reconstructed after a series of pre- and postprocessing. Results: We developed an adaptive three dimensional morphological filter, which smooth and enhance a series of cross sectional images in 3D space. It makes the imaging system independent of scanning conditions such as position and other factors. Conclusions: The system can be used for the diagnosis of fine vessels in a lesion. The developed filters are designed to process ultrasound images adaptively and are, thus useful for detecting fine vessel in noisy ultrasound images with low resolution. This research is supported by MEXT.HAITEKU (2002–2006) and Kanto Bureau of Economy, Trade and Industry. 2263 Tissue harmonic imaging for musculoskeletal ultrasound Gauthier TP, van Holsbeeck MT, Philips Ultrasound, United States of America; Henry Ford Hospital, United States of America Objectives: To assess the diagnostic performance of Tissue Harmonic Imaging (THI) for musculoskeletal imaging. Methods: A numerical simulation was used to predict harmonic waveform shape given ultrasound scanners signal path limitation. It was further used to design ideal THI waveforms for musculoskeletal imaging. Patients were then scanned with a compact linear transducer featuring THI. Side-by-side comparisons between conventional imaging and THI were performed by scanning pathologies using either imaging modes. Dual display was used to monitor both conventional and THI performance on same imaging planes. Results: When THI was used, image quality was significantly improved overall. Direct comparisons demonstrated lower overall clutter level with superior axial, contrast and lateral resolution leading to better border delineation of structures. THI was utilized for a more definitive diagnosis of cases such as meniscus tears, TFC abnormalities, rotator cuff tears, needle visualization during procedures and bursal abnormalities. Conclusions: THI may be used to improve image quality and help diagnostic. Musculoskeletal ultrasound typically involves scanning superficial structures, which implies that acoustic waveforms won’t propagate over long distances to allow for nonlinear propagation to occur. This study therefore supports the need for carefully designed acoustic waveforms to make THI work to image the musculoskeletal system.
Ultrasound in Medicine and Biology
metastases economically and effectively, also reducing the number of more expensive imaging procedures.
Volume 32, Number 5S, 2006 Conclusions: The method using the mean normal stress as unknown is effective. 2262
DIAGNOSTIC US 2260 Transient ultrasound radiation force elastography: A preliminary comparison with surface palpation elastography Bamber JC, Lima DM, Duck FA, Shipley JA, Xu L, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, United Kingdom of Great Britain and Northern Ireland; Bath Royal United Hospital, United Kingdom of Great Britain and Northern Ireland Objectives: Ultrasound elasticity imaging is being developed for evaluating the mechanical properties of soft tissues in vivo, but each research group tends to specialize in a specific approach. Our objective was to assess the advantages and disadvantages of generating tissue strain by deep loading with transient acoustic radiation force, versus the use of nearly static surface loading. Methods: Experiments with gelatin phantoms and finite element models were employed to measure the performance of pseudo-static freehand elastography and radiation force elastography. Results: Relative to surface loading, radiation force images are less susceptible to artefacts associated with boundary conditions, elastic modulus inhomogeneities and pre-existing tissue motion. Furthermore, they may possess a reduced rate of decay of strain signal-to-noise ratio with depth, and demonstrate an improved contrast-transfer-efficiency, particularly for inclusions that have negative modulus contrast or that are disconnected from the background by a low friction boundary. It takes longer to scan using radiation force loading and the elastograms may show contrast for ultrasonic attenuation and absorption, as well as for tissue stiffness. Conclusions: Transient deep loading possess advantages over surface loading for elastography but both approaches may have a role to play, and further comparative evaluation in vivo is needed. 2261 Multidimensional reconstruction/imaging of shear modulus distribution of living soft tissues Sumi C, Sophia University, Japan Objectives: Previously, we reported a shear modulus reconstruction method utilizing the typical value of Poisson’s ratio. However, as reconstruction errors were confirmed due to the difference between the original value and the set value, we proposed a method to reconstruct Poisson’s ratio as well. Furthermore, we proposed to reconstruct density as well to deal with dynamic deformation. However, due to the tissue incompressibility, the reconstruction of shear modulus, Poisson’s ratio and density become unstable. Methods: In this report, to stabilize these reconstructions, we report a new reconstruction method using the mean normal stress as unknown. This method also allows stable reconstructions of shear modulus and density under the condition that Poisson’s ratio remains unknown. The effectiveness is verified through 2D phantom experiments and 3D simulations. [The cubic phantom (50 mm sides) includes a spherical inclusion (radius, 5 mm) at the center (depth, 25.0 mm) having different shear modulus and Poisson’s ratio from those of the surrounding medium. Density is uniformly set. Noise-filled measurement data of displacement vector were simulated by adding white noise to the calculated raw displacement data.] Results: The new method yielded stable quantitative reconstructions for both the simulations and phantom experiments.
On useful ultrasound image processing techniques for the diagnosis of fine blood vessels Ito M, Yamada A, Nakamura K, Kato K, Kobayashi A, Kuroshima N, Tokyo Denki University, Japan; Tokyo Univ. Agriculture and Tech., Japan; Mitaka Kohki Co., Japan; Aloka Co., Japan; Micro Design Inc., Japan; Teikyo Univ., School of Medicine, Japan Objectives: In order to extract the contours of vessels of various sizes, robust three-dimensional imaging techniques are essential to the resultant successive scanned data. Our aim is to access to the vessels of about 300 to 500 microns. We focus on the speckle reduction and edge enhancement by adaptive morphological filters. Methods: The system with a high-frequency probe of 20 to 40 MHz, scan the region of 10⫻10⫻30 (mm in size). The accessed volume data have 128 cross sections of 512⫻512 pixels of 256 gray levels. Ultrasound longitudinal cross sectional images of blood vessels as well as 3D contours can be reconstructed after a series of pre- and postprocessing. Results: We developed an adaptive three dimensional morphological filter, which smooth and enhance a series of cross sectional images in 3D space. It makes the imaging system independent of scanning conditions such as position and other factors. Conclusions: The system can be used for the diagnosis of fine vessels in a lesion. The developed filters are designed to process ultrasound images adaptively and are, thus useful for detecting fine vessel in noisy ultrasound images with low resolution. This research is supported by MEXT.HAITEKU (2002–2006) and Kanto Bureau of Economy, Trade and Industry. 2263 Tissue harmonic imaging for musculoskeletal ultrasound Gauthier TP, van Holsbeeck MT, Philips Ultrasound, United States of America; Henry Ford Hospital, United States of America Objectives: To assess the diagnostic performance of Tissue Harmonic Imaging (THI) for musculoskeletal imaging. Methods: A numerical simulation was used to predict harmonic waveform shape given ultrasound scanners signal path limitation. It was further used to design ideal THI waveforms for musculoskeletal imaging. Patients were then scanned with a compact linear transducer featuring THI. Side-by-side comparisons between conventional imaging and THI were performed by scanning pathologies using either imaging modes. Dual display was used to monitor both conventional and THI performance on same imaging planes. Results: When THI was used, image quality was significantly improved overall. Direct comparisons demonstrated lower overall clutter level with superior axial, contrast and lateral resolution leading to better border delineation of structures. THI was utilized for a more definitive diagnosis of cases such as meniscus tears, TFC abnormalities, rotator cuff tears, needle visualization during procedures and bursal abnormalities. Conclusions: THI may be used to improve image quality and help diagnostic. Musculoskeletal ultrasound typically involves scanning superficial structures, which implies that acoustic waveforms won’t propagate over long distances to allow for nonlinear propagation to occur. This study therefore supports the need for carefully designed acoustic waveforms to make THI work to image the musculoskeletal system.