Improving the Image Quality of Synthetic Transmit Aperture Ultrasound Images - Achieving Real-Time In-Vivo Imaging

Synthetic transmit aperture (STA) imaging has the potential to increase the image quality of medical ultrasound images beyond the levels obtained by conventional imaging techniques (linear, phased, and convex array imaging). Currently, however, in-vivo applications of STA imaging is limited by a low signal-to-noise ratio (SNR), due to the application of a single transducer element at each emission, and higher susceptibility to tissue motion, produced by the summation of sequentially acquired low resolution images. In order to make real-time STA imaging feasible for in-vivo applications, these issues need to solved. The goal of this PhD study has been to find methods that can be used to overcome the above mentioned limitations, and hereby improve the image quality of STA imaging to a clinically desirable level, enabling real-time in-vivo STA imaging. The thesis investigates a new method to increase the SNR, which employs multi-element subapertures and linearly frequency modulated (FM) signals at each emission. The subaperture is applied to emulate a high power spherical wave transmitted by a virtual point source positioned behind the subaperture, and the linear FM signal replaces the conventional short excitation signal to increase the transmitted temporal energy. This approach, named Temporally encoded Multi-element Synthetic transmit aperture (TMS) imaging, is evaluated in detail for linear array and convex array imaging applications using simulations, and phantom and in-vivo experiments. The thesis contains summaries of four journal articles and four corresponding conference publications, which comprise the primary contributions of the PhD. The first two papers give elaborated evaluations of TMS imaging for linear array and convex array imaging, respectively. The results, including initial in-vivo experiments, showed, that TMS imaging can increase the SNR by as much as 17 dB compared to the traditional imaging techniques, which improves the in-vivo image quality to a highly competitive level. An in-vivo evaluation of convex array TMS imaging for abdominal imaging applications is presented in the third paper, based on a clinical trial with 7 healthy male volunteers. Real-time movie sequences of 3 seconds duration were acquired and analyzed by experienced medical doctors using blinded clinical evaluation. The results showed a statistically significant improvement in image quality of convex array TMS imaging compared to conventional convex array imaging. Only minor motion artifacts causing subtle image brightness fluctuations were reported in TMS imaging, which did not depreciate the diagnostic value of the images. The influence of tissue motion and a method for two-dimensional motion compensation is investigated in the fourth and final paper. The method estimates the tissue velocity and motion vii Abstract direction at each image point by correlating image lines beamformed along a set of motion directions and selects the direction and velocity corresponding to the highest correlation. Using these estimates, motion compensation is obtained by tracking the location of each pixel, when reconstructing the low resolution images. The presented phantom and in-vivo results showed, that severe tissue motion has a negative influence on the image quality of STA imaging as expected, but, most importantly, that the proposed method successfully compensates for the motion, thus, retaining the image quality of TMS imaging, when scanning moving tissue. In conclusion, the results of the research presented in this thesis have demonstrated, that TMS imaging is feasible for real-time in-vivo imaging, and that the obtained image quality is highly competitive with the techniques applied in current medical ultrasound scanners. Hereby, the goals of the PhD have been successfully achieved.