Alshamlan, Nayef2023-05-122023-05-122023-01https://hdl.handle.net/11299/254132University of Minnesota Ph.D. dissertation. January 2023. Major: Electrical Engineering. Advisor: Emad Ebbini. 1 computer file (PDF); v, 97 pages.Nonlinearity is inherent to ultrasound wave propagation and has been utilized in modernscanners to enhance contrast, lateral resolution and improve imaging depth. However, the use of nonlinear imaging methods is mostly limited to second harmonic imaging (SHI), which suffers from low SNR. The Volterra Filter (VF) has been shown to provide a new approach to pulse-echo ultrasound, which separates the linear and nonlinear components of the echo data. For example, the second-order VF (SoVF) separates the echo into linear and quadratic components. Compared to SHI, the quadratic component of the SoVF preserves the signal bandwidth and dynamic range while improving the SNR due to its inherent rejection of additive Gaussian noise. Quadratic B-mode (QB-mode) ultrasound offers the promise of higher contrast imaging than the linear B-mode (LB-mode), which could lead to a new imaging modality with different spatial and contrast resolutions. However, despite improved contrast and/or sensitivity to nonlinear echo components compared with conventional B-mode ultrasound, QB-mode still suffers from the fundamental limitations imposed by speckle and clutter phenomena. We propose a quadratic kernel design approach for improving the spatial and contrast resolutions of QB-mode imaging. Adaptive SoVF is used to estimate the quadratic kernel and characterize its bi-frequency response, which reveals the frequency interactions underlying the observed 1D quadratic signal spectrum. Preliminary results show that the kernel design approach leads to improved imaging performance in terms of spatial and contrast resolution tradeoffs. In addition to the standard resolution criteria, the proposed thesis research investigates the use of mutual information (MI) in characterizing the discrepancy between the LB-mode and QB-mode image representations of different tissue types in vivo. Finally, we investigate the use of the SoVF in imaging the tissue nonlinearity parameter, so-called B/A, which could provide quantitative and functional anatomical imaging of tissue water content.enThesis or Dissertation