H-scan ultrasound imaging with adaptive attenuation correction for improved detection of liver steatosis in human subjects

H-scan ultrasound (US) imaging is a tissue characterization technique that depicts the relative size of acoustic scatterers. However, frequency-dependent attenuation poses a major challenge for accurate analysis at tissue depth. Here we introduce a new approach for H-scan US imaging using adaptive attenuation correction (AC) to help overcome this issue. US data was acquired in phantom materials with known attenuation values and human liver with varying grades of steatosis (N = 10) using a Verasonics Vantage 256 and Philips EPIQ 7 US systems, respectively. Steatotic grade was calculated using magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) measures. The US signal was modeled as the sum of the true signal, attenuation fluctuation component (AFC), and random noise component (RNC). The new method involves developing a model that incorporates both the true signal along with the AFC and RNC. The approach also involves computing the noise power spectrum (PS) level, Fourier transform of the signal, and original PS of the signal. Preliminary results demonstrate a clear improvement in H-scan US image intensity at depth after adaptive AC. The study highlights the use of H-scan US imaging for liver tissue characterization, which can be used for the accurate detection and quantification of hepatic steatosis. Overall, H-scan US imaging with adaptive AC helps mitigate the confounding influence of acoustic attenuation at tissue depth, thereby leading to improved detection of subtle changes associated with early progression of fatty liver disease.