Goal: This study compares two strategies to compensate for attenuation in the estimation of the backscatter coefficient σ of breast cancer Methods: Attenuation compensation was performed using local attenuation estimates from a constrained Spectral-Log-Difference (CSLD) method, and average estimates from a dynamic programming (DP)-based regularized method. First, bias and linearity of σ estimates from both strategies were evaluated in a Gammex 410SCG phantom with inclusions with known σ imaged at 8 MHz on a Verasonics Vantage 128 scanner. Then, the strategies were compared in terms of the contrast-to-noise ratio (CNR) between adipose tissue and invasive ductal carcinoma. σ estimates were obtained from 10 patients with biopsy-confirmed IDCs imaged invivo with a 12L4 transducer on a Siemens Acuson S2000 as part of an IRB-approved protocol at the National Institute of Cancerology in Mexico City. DP σ estimates provided smaller bias and better linearity compared to estimates of CSLD. In the in vivo study, DP σ estimates produced negative IDC-vs-adipose CNR that was significantly different from zero (p = 0.002) and agreed with the hypoechoic appearance of IDC. [Work supported by CONACYT Ciencia de Frontera 1311307, ASA international student support, and UNAM PAPIIT IA102320. We thank Siemens Mexico for scanner loan.]

No abstract