Application of synthetic aperture imaging to non-invasive vascular elastography

Thesis (Ph. D.)--University of Rochester. Dept. of Electrical and Computer Engineering, 2012The health of the carotid artery is an important indicator of cardiovascular disease (CVD). The advent of CVD results in reduced elasticity and flexibility of the arterial wall. These changes in elasticity can be measured using non-invasive vascular elastography (NIVE). Elastography is performed by subjecting the tissue under investigation to a form of mechanical excitation. The resultant tissue displacement is measured and used to compute the spatial variation of strain within the tissue. These strain maps, known as elastograms, serve as surrogates for tissue elasticity. While elastography can be performed with any imaging modality, ultrasound is portable, inexpensive and has high frame rates, making it ideal for diagnostic screening purposes. However, linear array based elastography can accurately estimate displacement only in the axial direction. Consequently, NIVE cannot characterize strain across artery cross-sections, restricting the diagnostic value of this technique. The objective of this thesis was to investigate the feasibility of using synthetic aperture (SA) imaging to accurately estimate the 2D displacement vector, thereby enhancing the performance of NIVE. We demonstrated, through simulation and experiment, that SA imaging can accurately measure both axial and lateral displacements. These displacements generate high quality radial and circumferential strain elastograms of the arterial cross-section. Compared to conventional ultrasound elastography, SA elastography improved the error in lateral displacements and the resultant strain elastograms by an order of magnitude. However, the low frame rates and large data volumes required by SA imaging render this approach clinically inviable. To overcome this limitation, we developed a sparse array based elastography system. We demonstrated that as few as eight transmit elements can generate strain elastograms with a 16x improvement in frame rate and data volume, at a minimal loss of image quality. The efficacy of sparse array elastography was compared to that of a compounded plane wave imaging system. It was demonstrated that sparse array imaging displayed higher lateral sensitivity thereby producing strain elastograms with 20% improved image quality. From the success of this novel approach to elastography, it was concluded that further development and integration on a commercial ultrasound system would greatly enhance clinical use