Real-time implementation of a novel algorithm for ultrasound freehand elastography of breast lesions

摘要

The classification of breast lesions represents a main concern in current senology diagnostic practice. B-mode imaging plays a major role, but the specificity is still too low, since no information on tissue stiffness is directly provided. An original freehand elastography method, based on a Fourier domain displacement estimator, has been recently proposed and proved capable of producing off-line robust estimates of phantoms elasticity. In order to permit in-vivo examinations of breast lesions, a real-time version of the proposed method has been implemented in the research scanner ULA-OP, designed at the University of Florence. In a preliminary test on patients the proposed method detected 36 lesions (11 softer, 14 harder and 11 having the same elasticity than the surrounding tissue). The same lesions were classified as malignant (5) and benign (31) by an experienced sonographer through B-mode analysis.