Ultrafast ultrasound breast scanning for improved image quality and functional information in 3-D

Abstract

With an automated breast volume scanner (ABVS), a volumetric breast image can be constructed. Clinical studies show high sensitivity but also high recall-rates due to the detection of many lesions of uncertain malignant potential. It is known that breast lesions have different mechanical properties with respect to its surrounding tissue. Consequently, imaging the biomechanics of a breast might improve diagnosis in this disease. First, we improved the image quality by taking full advantage of the ultrafast acquisition and by applying dedicated signal processing techniques. Next, by imaging biomechanics we increased the specificity since compared to benign lesions, malignant lesions are often stiffer, and more grown into the surrounding tissue (firmly bonded). We extended conventional 2-D strain imaging to full 3-D (shear) strain imaging by combining automated breast volume scanning with ultrafast plane wave imaging (10 000 frames/s). Validation studies in breast phantoms revealed that lesions could be differentiated based on strain and shear strain imaging. Decreased strain values were found in stiff lesions and firmly bonded lesions demonstrated low shear strain values. Initial acquisitions in patients demonstrated decreasing strain values with increasing severity of the disease. Additionally, we developed an ultrasensitive 3-D Doppler method to detect neo-vascularization that might be present around malignant lesions. In conclusion, imaging the biomechanical properties in 3-D improves diagnosis breast cancer.With an automated breast volume scanner (ABVS), a volumetric breast image can be constructed. Clinical studies show high sensitivity but also high recall-rates due to the detection of many lesions of uncertain malignant potential. It is known that breast lesions have different mechanical properties with respect to its surrounding tissue. Consequently, imaging the biomechanics of a breast might improve diagnosis in this disease. First, we improved the image quality by taking full advantage of the ultrafast acquisition and by applying dedicated signal processing techniques. Next, by imaging biomechanics we increased the specificity since compared to benign lesions, malignant lesions are often stiffer, and more grown into the surrounding tissue (firmly bonded). We extended conventional 2-D strain imaging to full 3-D (shear) strain imaging by combining automated breast volume scanning with ultrafast plane wave imaging (10 000 frames/s). Validation studies in breast phantoms revealed that lesions could be diff...