Multiplane Deconvolution in Orthographic Ultrasonic Transmission Imaging

Abstract

Orthographic ultrasonic transmission imaging has received renewed interest and demonstrated significant potential clinical applicability during the past few years. The ultrasonic camera developed at SRI for the Orthopaedic Clinic of the University of Müenster in West Germany is an example of a real-time ultrasonic transmission imaging system. Using a spatially and temporally incoherent method of insonification, which eliminates diffraction artifacts and smoothly blurs out-of-focus structures, the SRI camera produces sharply focused images within the focal plane. Although well-focused images can be obtained in orthographic transmission imaging, the overlapping of out-of-focus structures frequently detracts from image quality. To overcome this difficulty and to extend the applicability of ultrasonic transmission imaging, we proposed to apply image-processing techniques to remove out-of-focus structures through multiplane deconvolution. We used a linear model for the imaging process and investigated different algorithms for the multiplane deconvolution. Our simulation results show that an algorithm based on the minimum mean square error appears to be the most effective. We have also applied the algorithm to improve the images collected with an experimental ultrasonic transmission imaging system. In this paper, we briefly describe the principles of multiplane deconvulution and present simulation and experimental results that illustrate its effectiveness in improving the image quality.KeywordsFocal PlaneTransmission ImagingOptimal FilterNoise Power Spectral DensityUltrasonic TransmissionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.