Current Research and Scholarly Interests
Our laboratory is an ultrasound engineering laboratory that develops ultrasonic beamforming and image reconstruction methods for diagnostic imaging applications. Our current focus is on beamforming methods that are capable of generating high-quality images in difficult-to-image patients and imaging conditions. We attempt to build these imaging methods into real-time imaging systems in order to apply them to clinical applications, such as pediatric, cardiac, liver, and fetal imaging. Another focus of our laboratory is the development of ultrasound molecular imaging platforms, including image reconstruction methodologies and targeted microbubbles for molecular signatures. We also employ beamforming concepts to enhance other areas of ultrasound research, such as therapeutic ultrasound and microbubble-mediated drug delivery or passive cavitation imaging.
A current project in our laboratory involve differentiable beamforming, which is a technique that iteratively estimates beamforming parameters to improve the reconstructed image. For example, we estimate the local sound speed and beamform the image using that speed of sound to correct for distortion in the ultrasound image (also known as aberration correction). We have also used the differentiable beamformer to solve other problems in ultrasound imaging, such as estimating the array shape of flexible ultrasound arrays. Another project involves the simulation of nonlinear, acoustic wave propagation under complex models of human anatomy and evaluate the impact of anatomy and acoustic parameters on the resulting images. Often, the anatomy and acoustic parameters are the source of aberration and diffuse reverberation of the wavefronts, both of which contribute to image noise. We have developed techniques that suppress the reverberation noise to improve downstream imaging techniques such as Doppler, radiation force imaging, or aberration correction. We often employ these techniques in clinical studies to observe the improvement in image quality.
In ultrsaound molecular imaging, we utilize signal processing and machine learning techniques to develop nondestructive, real-time ultrasound molecular imaging to permit clinical usage of the technique. In ultrasound molecular imaging, ligands are attached to microbubbles in order to bind to a specific biomarker of disease. In our case, we wish to avoid the destruction of the microbubble in the imaging process, which is commonly used to differentiate bound from free-floating microbubbles. We also wish to permit free-hand and real-time molecular imaging, which is currently unavailable with state-of-the-art ultrasound molecular imaging techniques. In addition to developing imaging capabilities, we have also developed small ligands that bind to cancer biomarkers such that the cancers can be detected and diagnosed under ultrasound.
