Adaptive Singular Value Decomposition-Based Hyperecho Suppression for Diagnostic Power Doppler Ultrasound Image Processing

Abstract

The power Doppler (PD) ultrasound imaging provides high-quality, noninvasive visualization of blood flow and has the potential to be used for circulation screening. However, its application in human arterial imaging remains challenging due to the presence of complex hyperechoic moving structures (HMSs). In this study, we propose an adaptive singular value decomposition (SVD) filtering strategy for HMS suppression. The proposed method used a k-means clustering algorithm directly on prebeamformed IQ data to segment HMS and non-HMS regions, followed by an adaptive SVD filtering strategy tailored to each tissue type. Compared to the existing SVD filtering methods, the proposed approach can effectively suppress the HMS artifacts. In addition, the human carotid artery imaging experiments demonstrate significant improvement in HMS suppression throughout cardiac cycles and across various imaging locations. With such capability, we believe that the proposed strategy will be a useful tool in applying PD for the 3-D imaging of human blood vessels.