Topic：Application of Sparse Radon Transform in Quantitative Bone Ultrasound
Time：March 26, 15:00—17:00
Location：Auditorium, SIST Building
Speaker： Prof. Lawrence Le
Abstract:
Multichannel analysis of dispersive ultrasonic energies requires a reliable mapping of the data from the time-distance domain to the frequency-wavenumber or frequency-phase velocity domain. The mapping is usually performed with the classical 2-D Fourier transform (FT). The extracted dispersion trajectories of the guided modes lack the resolution in the transformed domain to discriminate wave modes. The resolving power associated with the FT is closely linked to the spatial dimension of the recorded data. This is especially true for bone ultrasound using guided waves where the spatial acquisition for an axial transmission configuration is very often restricted by the limited dimension of the ultrasound probe, the number of channels, the irregularity of the acquisition surface, and the accessibility to the skeletal site. This is well documented in the literature.
In 2014, we brought the notion of Radon transform [Nguyen et al, Ultrasonics 54:1178-1185; Tran et al, Ultrasound Med Biol 40:2715-2727; Tran et al, J Acoust Soc Am 136:248-259] to the attention of bone ultrasound community and provided a Radon-based solution to solve the resolution problem. In this presentation, we will revisit the linear high-resolution Radon transform (RT) to filter and reconstruct wavefields, and to image the dispersive energies of the recorded wavefields through long bones. The RT is posed as an inverse problem, which allows implementation of the regularization strategy to enhance the focusing power. We will use l1-norm regularization to illustrate the advantages and robustness of the high-resolution RT algorithm using the simulated, ex-vivo, and in-vivo data. The method also accommodates unevenly-spaced records, effectively attenuates random noise, enhances the signal-to-noise ratio, reconstructs the missing records, and improves the coherency of the guided wave modes. The proposed transform presents a powerful signal enhancement and imaging tool to process ultrasonic wavefields and extract dispersive guided wave energies under limited aperture.
At the end, Dr. Le will also introduce the internship and graduate opportunities in the University of Alberta
Bio:
Dr. Lawrence Le received his PhD in earth physics from the University of Alberta, Edmonton, Canada, in 1991. He held a NSERC postdoctoral position in Schlumberger-Doll Research Lab, Ridgefield, Connecticut. In 1994, he made a big switch in his career to start his medical physics residency in the Department of Radiology and Diagnostic Imaging. In 1999, he completed a MBA degree in finance and technology commercialization. He is currently a professor in medical physics and a director of the graduate program in the department. He directs the Ultrasonic Bone Tissue Characterization and Imaging group. He guides his students to use vigorous geophysical and signal processing principles to study long bones and invert the data for bone properties.