TH-AB-209-06: An Investigation of the Feasibility of Malignant Lesion Detection During Routine Breast Cancer Screening Using a Prototype Acousto-Optic Transmission Ultrasound Imaging System

Abstract

Purpose: To investigate the feasibility of malignant lesion detection during routine breast cancer screening using an acousto-optic (AO) transmission ultrasound imaging system, particularly in premenopausal women with high breast density. Methods: A full-field, single-projection ultrasound imaging system was developed that uses a high-resolution AO detector to convert the acoustic intensity transmitted through the compressed breast into a visual image by virtue of the AO effect in nematic liquid crystals. In this work, a comprehensive system model was proposed to describe the AO imaging process, including the generation of the incident acoustic field by the transducer, the propagation of the field through the breast tissue, and the conversion of the transmitted acoustic field intensity into a visual image. Validation of the imaging model was achieved through comparison of actual AO breast phantom images with simulated images based on the proposed model. Malignant lesion detectability studies were subsequently performed in simulation using homogeneous and heterogeneous numerical breast phantoms. Results: Comparison of actual AO breast phantom images with simulated images based on the proposed system model showed strong agreement, with an RMSE less than 4%. Lesion detectability studies using homogeneous numerical breast phantoms demonstrated excellent visibility for breast lesions as small as 0.5 cm and source frequencies on the order of 4 MHz. Though lesion detectability proved immune to spatial variations in mass density (2%) and attenuation (15%) in the breast parenchyma, minor variations in parenchymal sound speed (1.3%) resulted in substantial refraction artifacts that compromised the diagnostic utility of the prototype system. Conclusions: Differences in acoustic absorption may overcome refraction and enhance lesion visibility at conventional breast ultrasound frequencies (10 to 15 MHz). However, provided the FDA limit on the incident acoustic intensity (1 W/cm2) is employed, such high-frequency imaging will require improvements in the current AO detector sensitivity (10−6 W/cm2). Department of Defense (DOD) Breast Cancer Research Program IDEA Award W81XWH-11-1-0332; National Institutes of Health (NIH) Grant T32 EB002103-24 from the National Institute of Biomedical Imaging and Bioengineering (NIBIB)