Ringdown reduction for an intraluminal ultrasound array using depth-dependent azimuthal filters

Abstract

Vibration of piezoelectric elements after electrical excitation creates ringdown artifacts in ultrasound images close to the transducer surface. For intraluminal applications, high quality images are often needed at depths close to the catheter surface. Consequently, ringdown artifacts pose a serious problem for intraluminal circumferential arrays. A current method to reduce these artificial echoes subtracts a precalibrated ringdown signal from each frame in a data set. The dynamic nature of the ringdown artifact, however, reduces the effectiveness of this subtraction algorithm with time following data collection. The ringdown signal has very distinctive characteristics, especially in the near field azimuthal spectrum. An in-depth study of those characteristics presents new alternatives to minimize ringdown artifacts. Using a set of depth-dependent azimuthal filters, we present a method to significantly reduce ringdown artifacts while minimally affecting the real signal. Rubber phantoms with embedded graphite particles were imaged close to the transducer surface to study potential improvements obtained with the azimuthal filters. Results show a reduction of the ringdown signal by at least 10 dB, while increasing the dynamic range between the real signal and the ringdown from 2 dB to 15 dB. The combination of azimuthal filtering with ringdown subtraction can greatly enhance contrast at ranges close to the transducer surface.