Fast and Robust Clutter Filtering in Ultrafast Echocardiography

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Signal Processing

Background:

• Singular value decomposition (SVD) is standard for ultrasound clutter filtering but is computationally intensive and sensitive to motion.
• Randomized SVD (rSVD) accelerates filtering for stationary tissues but its efficacy in ultrafast echocardiography and blood flow imaging is unproven.

Purpose of the Study:

• To evaluate randomized SVD (rSVD) as a fast and robust clutter filter for ultrafast echocardiograms.
• To assess rSVD's performance in power Doppler analysis of myocardial and blood dynamics.

Main Methods:

• Acquired in vivo ultrafast echocardiograms from nine human hearts using cascaded synthetic aperture imaging.
• Applied rSVD for clutter filtering prior to power Doppler analysis.
• Calculated contrast ratio (CR) and contrast-to-noise ratio (CNR) to evaluate image quality.

Main Results:

• rSVD accelerated clutter filtering by 12-fold compared to full-SVD.
• rSVD demonstrated significantly improved local contrast (mean CNR, p < 0.001) and comparable overall effectiveness (mean CR, p = 0.20).
• rSVD showed superior robustness in low signal-to-noise ratio scenarios, with smaller CR and CNR standard deviations.

Conclusions:

• rSVD is a computationally efficient and robust alternative to full-SVD for clutter filtering in ultrafast echocardiography.
• rSVD enhances blood flow dynamics analysis by improving contrast resolution and stability.
• The findings support the use of rSVD for ultrafast echocardiographic blood dynamics assessment.