Enhancing breast nonlinearity parameter imaging using full angular spatial compounding

The presence of abnormalities in breast tissue constitutes a critical public health issue, emphasizing the necessity for advanced diagnostic methods to improve early detection. Hence, breast ultrasound has emerged as a crucial diagnostic tool offering unique soft tissue information through acoustic parameters analysis. The acoustic nonlinearity parameter, B/A, in breast pathologies exhibits different values compared to healthy tissue. Thus, the B/A imaging has potential to complement conventional B-mode ultrasound diagnosis. A recent method to estimate the B/A is the depletion method (DM); however, the B/A images obtained using the DM are limited by a large bias and variance trade-off. In the present work, a new technique that integrates the DM with the full angular spatial compounding (FASC) framework is proposed. By combining data from multiple viewing angles, FASC enhances image quality by reducing the variance and increasing contrast, thereby, yielding a more precise and accurate representation. The analysis was performed both with and without the application of total variation (TV) regularization to further refine the B/A estimates. Assessment of the effectiveness of the FASC approach was conducted using three samples (I, II, III) in silico with different inclusion configurations. Simulations show a reduction in the average normalized root mean square error of background and inclusion regions from 0.92, 0.79 and 0.94, respectively, using single-view estimates, to 0.31, 0.25, and 0.24 using FASC, and further to 0.16, 0.09, and 0.07 when combining FASC and TV regularization. These results suggest the potential to estimate the B/A using the depletion method in combination with FASC and TV regularization frameworks to enhance the bias-variance trade-off.