Combined spatial compounding and adaptive imaging for clinical breast ultrasound
Ultrasonic images are affected by tissue inhomogeneities which cause degradation in image quality. The distortions associated with these inhomogeneities are caused by errors in the arrival time of the acoustical wavefront at the elements of the transducer array, and are known as phase aberrations. Although current research is focused on the complex distributed model of aberration, all real-time adaptive imaging systems use the more simple near-field correction techniques.
Spatial compounding is another technique for improving ultrasonic images. Spatial compounding, however, attempts to improve image quality by reducing the grainy appearance typical of ultrasonic images, which is caused by speckle. Although the theory and physics of spatial compounding have been around for over two decades, spatial compounding has not been implemented on commercially available ultrasound scanners until recently.
This dissertation focuses on combining the advantages of spatial compounding and adaptive imaging in ultrasonic images. Image quality is compared using standard and new image quality metrics. Experiments in simulation, tissue mimicking phantoms, and clinical breast data showed that combined spatial compounding and adaptive imaging yielded strong improvements in contrast-to-noise ratio of cysts and image spatial frequencies. Clinical breast experiments also showed that the best image quality was obtained with a large sub-aperture size, and enough sub-apertures to cover the entire array.
Further experiments are presented in this dissertation to characterize the stability of clinical aberrations and the relation of clinical aberrator stability to real-time adaptive imaging. Factors which effect aberrator stability are characterized by experiments in tissue mimicking phantoms and simulations. The desired amount of aberrator stability for real-time adaptive imaging results in a trade-off in image quality and computational time.
New adaptive imaging techniques are proposed to produce improved estimates of phase aberrations. Adaptive and non-adaptive spatial filtering techniques are presented in simulation, phantom, and clinical experiments and show that these filters are effective in reducing the estimation fitter associated with time delay estimation techniques. The results of these experiments have implications for real-time adaptive imaging because the spatial filters proposed in this work improve the spatial stability of aberration.