Automated In Vivo Sub-Hertz Analysis of Viscoelasticity (SAVE) for Evaluation of Breast Lesions

Automated In Vivo Sub-Hertz Analysis of Viscoelasticity (SAVE) for Evaluation of Breast Lesions 170 177 IEEE Transactions on Biomedical Engineering (TBME)

Many breast tumors found by palpation manifest as stiff masses. The stiffness of tumor is related to a mechanical property known as elasticity. Elastography methods use tissue elasticity as a biomarker to provide invaluable information that improves classification of the breast masses. However, naturally heterogeneous breast masses exhibit time varying deformation at constant loading revealing another mechanical property known as viscoelasticity. In this study, a creep-like test was implemented using an automated external compression device to characterize the viscoelasticity properties of the breast masses in vivo. While the breast was compressed by the automated device, ultrasound imaging data were acquired at a high frame rate to continuously track tissue response throughout different deformation phases. A two-step approach was employed to model tissue sudden response and subsequent gradual deformation as a purely elastic and a first order Kelvin-Voigt viscoelastic response, respectively. An automated region of interest selection method was developed to obtain imaging contrast features that tuned to the shape and location of the masses from which quantitative values based on different mechanical properties were obtained. Benign lesions had lower mean values of instantaneous and creep elasticity than malignant lesions. However, viscoelastic retardation time T1 in benign lesions was higher than malignant lesions. Among all properties,  provided contrast values that were significantly different in the two groups of benign and malignant lesions (p=0.008). Employing this method in a large group of patients will provide more insight about the utility of this method as a new biomarker for differentiating breast masses based on their viscoelasticity properties at low frequencies.