Higher Order Singular Value Decomposition Filter for Contrast Echocardiography

Area of Science:

• Medical Imaging
• Cardiovascular Ultrasound
• Signal Processing

Background:

• Contrast-enhanced echocardiography is clinically relevant for coronary circulation assessment but limited by poor image quality.
• Current methods like multipulse contrast schemes (MPCSs) and singular value decomposition (SVD) filters struggle with motion and propagation artifacts, hindering contrast agent visibility in the myocardium.

Purpose of the Study:

• To develop and evaluate a novel signal processing scheme using higher-order SVD (HOSVD) for amplitude modulation/amplitude-modulated pulse inversion (AM/AMPI) echoes.
• To improve contrast detection and image quality in contrast-enhanced echocardiography compared to existing techniques.

Main Methods:

• Proposed a processing scheme employing HOSVD on AM/AMPI echoes, treating complementary pulse echoes as a separate dimension.
• Implemented a method to remove tissue scattering signals by rank reduction in the HOSVD algorithm for enhanced contrast detection.
• Conducted in vitro flow phantom and in vivo porcine heart experiments to compare the proposed HOSVD method with conventional AM and SVD filters.

Main Results:

• In vitro studies demonstrated that HOSVD on AM pulsing achieved a higher contrast-to-background ratio (CBR) than conventional AM (10 dB higher) and SVD filter (14 dB higher).
• In vivo porcine heart experiments showed HOSVD improved CBR by up to 19 dB in open-chest acquisitions compared to AM.
• HOSVD also improved contrast ratio (CR) by 3 dB in closed-chest acquisitions.

Conclusions:

• The proposed HOSVD processing scheme significantly enhances contrast agent detection and image quality in echocardiography.
• This method offers a promising advancement for routine clinical assessment of coronary circulation using contrast-enhanced ultrasound.
• HOSVD provides superior performance over conventional methods, addressing key limitations in current clinical tools.