|
An important ongoing question for higher education is how to successfully mentor the next generation of scientists
and engineers. It has been my privilege to have been mentored by one of the best, Dr Robert F. Wagner and his
colleagues at the CDRH/FDA during the mid 1980s. Bob introduced many of us in medical ultrasonics to statistical
imaging techniques. These ideas continue to broadly influence studies on adaptive aperture management
(beamforming, speckle suppression, compounding), tissue characterization (texture features, Rayleigh/Rician
statistics, scatterer size and number density estimators), and fundamental questions about how limitations of
the human eye-brain system for extracting information from textured images can motivate image processing. He
adapted the classical techniques of signal detection theory to coherent imaging systems that, for the first time in
ultrasonics, related common engineering metrics for image quality to task-based clinical performance. This talk
summarizes my wonderfully-exciting three years with Bob as I watched him explore topics in statistical image
analysis that formed a rational basis for many of the signal processing techniques used in commercial systems
today. It is a story of an exciting time in medical ultrasonics, and of how a sparkling personality guided and
motivated the development of junior scientists who flocked around him in admiration and amazement.
|
