Evaluation of the Effect of Clutter Reduction in Attenuation Coefficient Estimation

Abstract

Quantitative ultrasound (QUS) aims to provide objective measurements of tissue properties, thereby overcoming the limitations of conventional subjective assessments. A relevant clinical application of QUS is the assessment of metabolic dysfunction-associated steatotic liver disease (MASLD). However, in abdominal imaging, tissue heterogeneity increases the likelihood of acoustic interference, known as clutter, a type of image degradation caused by multiple scattering, reverberation, or off-axis reflections that introduce unwanted signals into the data received by the transducer. Clutter has a negative effect on the estimation of the attenuation coefficient (AC). In this study, the ADMIRE algorithm was evaluated to reduce clutter and improve the robustness of AC estimates.Results in simulations with known values (0.4-0.6 dB/cm-MHz) show that ADMIRE reduced the standard deviation by up to 56.7% (from ±0.3 to ±0.13) and the overestimated maximum AC value decreased by approximately 83.3% compared to the ground truth value (from around 1.2 to 0.7 dB/cm-MHz), which reduced the mean error from 35% to 13.3%. In clinical data from healthy livers, the algorithm consistently improved accuracy, reducing the standard deviation by up to 27.8% and producing values within the range expected according to the literature (0.56-0.63 dB/cm-MHz).