Ultrasound Small Vessel Imaging With Block-Wise Adaptive Local Clutter Filtering

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Signal Processing

Background:

• Robust clutter filtering is crucial for effective ultrasound small vessel imaging.
• While eigen-based methods excel in small animals, they face challenges in human imaging due to complex tissue and noise.
• Existing techniques struggle with spatially-varying characteristics in in vivo human scans.

Purpose of the Study:

• To develop a novel block-wise adaptive singular value decomposition (SVD) clutter filtering technique for improved in vivo human ultrasound imaging.
• To enhance the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of small vessel perfusion images.
• To address the limitations of conventional global SVD filtering in complex human tissues.

Main Methods:

• The proposed method segments plane wave ultrasound data into overlapped local spatial blocks.
• Singular value decomposition (SVD) is applied within each block to separate tissue, blood, and noise signals.
• Adaptive singular value cutoff thresholds are determined locally, and results are combined for enhanced image quality.

Main Results:

• The block-wise adaptive SVD method achieved over a two-fold increase in SNR and a three-fold increase in CNR (in dB) compared to global SVD.
• Significant improvements were observed in suppressing depth-dependent background noise and near-field clutter.
• The study systematically investigated the impact of block size and overlap on imaging quality and computational cost.

Conclusions:

• The novel block-wise adaptive SVD technique offers superior clutter filtering for ultrasound small vessel imaging in humans.
• This method effectively improves SNR and CNR, outperforming conventional global SVD approaches.
• The findings demonstrate a promising advancement for in vivo human ultrasound perfusion imaging.