Improved visualization of endocardial borders with Short-Lag Spatial Coherence imaging of fundamental and harmonic ultrasound data

Abstract

Endocardial border visualization, a common task in echochardiography, is typically challenged by the presence of acoustic clutter. This study investigates endocardial border visibility in co-registered fundamental and harmonic data when utilizing the Short-Lag Spatial Coherence (SLSC) beamformer, a clutter reduction approach that we developed. Individual channel echo data were acquired from the left ventricle of 12 volunteers, after informed consent and IRB approval, to create matched image quadruplets of fundamental and harmonic B-mode and SLSC images. Contrast-to-noise ratios (CNR) were measured, and three cardiologists rated the visibility of endocardial segments in randomly ordered cine loops. The statistical significance of visibility ratings was determined with a Holm-Bonferroni correction at a significance level, α = 0.05. CNR increased approximately twofold in fundamental and harmonic SLSC images compared to fundamental and harmonic B-mode images. Fundamental and Harmonic SLSC imaging offered the greatest benefits when fundamental B-mode image quality was poor. Improvements in endocardial segment visibility in short-axis views ranged from 16-28% (α = 0.05) compared to fundamental B-mode images, while improvements in the apical four chamber views ranged from 22-35% (α = 0.05) compared to fundamental and harmonic B-mode images. Results suggest that SLSC and HSCI have the potential to increase endocardial border visualization and thereby improve cardiac assessments of poor-quality B-mode images.

Short-lag spatial coherence beamforming of photoacoustic images for enhanced visualization of prostate brachytherapy seeds

“…SLSC imaging improved lesion detectability in the presence of acoustic noise artifacts, when compared to conventional B-mode ultrasound imaging [15]. In addition, this beamformer was applied to in vivo ultrasound data to reduce clutter noise artifacts in cardiac [16][17][18], liver [19], thyroid [14], and vascular [15] images. Similar clutter reduction benefits were achieved when the SLSC beamformer was applied to photoacoustic data [20].…”

Postprocessing Approaches for the Improvement of Cardiac Ultrasound B-Mode Images: A Review

“…SLSC was extended to combine the harmonic signals (instead of just the fundamental) to form a new spatial coherence image. The approach was referred to as harmonic spatial coherence imaging (HSCI) [143]. HSCI demonstrated the potential to moderately improve on the SLSC performance.…”

CohereNet: A Deep Learning Architecture for Ultrasound Spatial Correlation Estimation and Coherence-Based Beamforming

“…SLSC beamforming is an advanced technique that is based on the fundamental van Cittert–Zernike theorem applied to backscattered ultrasound echo data [ 13 ] and therefore relies on the computation of multiple spatial coherence functions, which requires multiple cross-correlation calculations. Specifically, SLSC images are formed by calculating and then summing the spatial coherence of backscattered pressure waves received across the ultrasound transducer, demonstrating improvements over traditional ultrasound image quality in a variety of in vivo imaging applications, including breast [ 18 ], [ 19 ], liver [ 20 ], [ 21 ], fetal [ 22 ], [ 23 ], cardiac [ 24 ], [ 25 ], and thyroid [ 17 ] imaging. Benefits of SLSC images include improved visualization of anechoic targets of interest, including breast [ 18 ], [ 19 ] and thyroid cysts [ 17 ] as well as improved endocardial border detection [ 24 ], [ 25 ].…”

“…Specifically, SLSC images are formed by calculating and then summing the spatial coherence of backscattered pressure waves received across the ultrasound transducer, demonstrating improvements over traditional ultrasound image quality in a variety of in vivo imaging applications, including breast [ 18 ], [ 19 ], liver [ 20 ], [ 21 ], fetal [ 22 ], [ 23 ], cardiac [ 24 ], [ 25 ], and thyroid [ 17 ] imaging. Benefits of SLSC images include improved visualization of anechoic targets of interest, including breast [ 18 ], [ 19 ] and thyroid cysts [ 17 ] as well as improved endocardial border detection [ 24 ], [ 25 ]. However, clinical implementation was initially limited by the computational expense associated with multiple repeated correlation calculations.…”