Walker, William F.
Associate Professor, Biomedical Engineering
PO Box 800759
Medical Ultrasonic Imaging
The main thrusts of my research are the development of new methods that use ultrasonic waves to interrogate previously unexplored tissue properties, the development of new signal processing methods to enhance ultrasound image quality, and the development of novel system hardware to enable new experimental approaches or open new realms of clinical application.
In one project, termed radiation force imaging, we utilize a series of ultrasonic pulses to apply local forces within tissue. We then process the echoes received from these pulses to measure the viscoelastic response of the tissue. This data forms the basis of a number of new image types that depict the complex mechanical properties of tissues in vivo. This technique has shown early potential for imaging the vitreous body of the eye to predict future retinal detachment and for performing dynamic measurement of blood clot formation.
In another project we utilize a software-modified clinical GE scanner to interrogate the angular scatter properties of tissue. We electronically move the transmit and receive apertures in equal and opposite directions to obtain data at a variety of scattering angles, while maintaining a uniform system response across angles to enable coherent processing. This technique offers a new source of contrast in soft tissues and has shown the potential to improve the visualization of calcification.
In two other projects we are applying techniques from the realm of statistical signal processing to speed and enhance ultrasound beamformer design and to improve ultrasound image resolution and contrast through adaptive imaging.
We have two active projects in the area of hardware development. In the first we are modifying a Philips SONOS 5500 imaging system to enable simultaneous acquisition of data from all 128 channels of this phased array scanner, over a period of more than one cardiac cycle. Once complete, we will be the only research group in the world with this capability. We anticipate that such data will open a broad array of signal processing approaches. In another project we are working to develop an integrated beamformer capable of addressing a fully populated 1024 element array. The high level of integration used in this system will enable production at an extremely low cost and will open new realms of clinical application.