Correction of multipath interference using CLEAN and spatial location diversity

Previous studies indicate that ultrasonic wavefront distortion inside the female breast falls into two categories, incoherent scattering and coherent multipath. Scattering reduces target strength, broadens image lobe and lowers the image contrast while refraction creates coherent multipath interference that produces false targets in the image. Time-delay type compensation algorithms are useful to minimize scattered energy and to improve image contrast to a large extent (15 dB). Further improvement of contrast resolution requires compensation algorithms that can remove false targets. Multipath interference is a much more difficult problem than scattering. Because it is not a stationary, stochastic process, two distinct steps are required to deaberrate refraction-contaminated image. It is first necessary to recognize which target is the true target and which is the false one caused by interference. The authors demonstrate that when a cluster of true and false targets is imaged from several different array locations the true target remains in position and changes very little in size and shape. The images of false targets, on the other hand, are highly sensitive to the direction from which they are viewed. Therefore, they change sufficiently in location, size and shape to classify them as false. In this way the recognition problem is solved. The second key problem is the elimination of the multipath arrivals. The authors show that coherent CLEAN technique can be used as a tool to solve the second problem.