In this work, we propose the development of an innovative processing approach based on Full Waveform Inversion (FWI). Known as a powerful tool for seismic imaging and subsurface exploration, FWI has also shown potential in medical imaging, as both fields share similar principles and techniques in signal processing. However, while historically constrained by computational burdens, advancements in technology and data science now enable its broader application.
This project aims to integrate FWI into time-harmonic shear wave elastography while also leveraging machine learning approaches, such as neural operator methods, to accelerate the computationally intensive simulations of nonlinear wavefield propagation and mode conversion. This will contribute to a significant advancement, enhancing both the efficiency and accuracy of the FWI imaging technique. An illustrative application could involve assessing cardiac health and diagnosing conditions by estimating myocardial stiffness. Validation against state-of-the-art techniques will be achieved through simulation studies, channel data recorded from elasticity phantoms, as well as in vivo data recorded using the research ultrasound scanners available in the University of Oslo laboratory.
?