A new method for assessing the performance of signal processing filters in suppressing the side lobe level

Ultrasonography

Mok Kun Jeong¹  Sung Jae Kwon² 

[1] Department of Electronic Engineering, Daejin University, Pocheon, Korea;Division of Human IT Convergence Engineering, Daejin University, Pocheon, Korea;

关键词: beamforming;    contrast;    resolution;    side lobe;    ultrasound;   

DOI  :  10.14366/usg.20032

来源: DOAJ

【 摘 要 】

Purpose This study aimed to propose a new ground truth ultrasound imaging method and to confirm its efficacy when applied to side lobe suppression filtering. Methods Using a computer simulation, we synthesized a side lobe-free image (i.e., with no side lobe whatsoever) by separating the main and side lobe signals in the construction of point target, speckled cyst, and pseudo-kidney images. During signal processing, we assessed the quality of the filtered image by comparing it with a ground truth image (i.e., the main lobe image). Results We examined the effect of reducing the side lobe by applying aperture apodization, side lobe estimation and reduction filtering, and minimum variance beamforming, which are widely used as side lobe suppression techniques. Despite the drawback of decreased resolution, the commonly used apodization method increases the contrast, which improves ultrasound image quality and enables a better diagnosis. Although side lobe estimation and reduction filtering and minimum variance beamforming are demanding in terms of computational resources, they can considerably improve ultrasound images. Compounding of ultrasound images processed by various signal processing methods increases the resolution and contrast, while reducing the speckle noise. Conclusion Although it appears that the proposed method can only be used for computer-generated radiofrequency data, this method can improve ultrasound image quality by identifying the characteristics of signal processing filters for side lobe suppression and applying appropriately adjusted filters to in vivo human imaging data.

【 授权许可】

Unknown