2018/07/03 by Hani Almansouri, Singanallur Venkatakrishnan, Almansouri, Hani +7
Earth and Planetary Sciences · Engineering · #Electrical and Bioimpedance Tomography #FOS: Electrical engineering #Geophysical Methods and Applications #Image and Video Processing (eess.IV) #Seismic Imaging and Inversion Techniques #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1807.01224
openalex publication_date 2018/07/03 · openalex created_date 2022/08/04 · openalex updated_date 2026/07/28
Ultrasound reflection tomography is widely used to image large complex\nspecimens that are only accessible from a single side, such as well systems and\nnuclear power plant containment walls. Typical methods for inverting the\nmeasurement rely on delay-and-sum algorithms that rapidly produce\nreconstructions but with significant artifacts. Recently, model-based\nreconstruction approaches using a linear forward model have been shown to\nsignificantly improve image quality compared to the conventional approach.\nHowever, even these techniques result in artifacts for complex objects because\nof the inherent non-linearity of the ultrasound forward model.\n In this paper, we propose a non-iterative model-based reconstruction method\nfor inverting measurements that are based on non-linear forward models for\nultrasound imaging. Our approach involves obtaining an approximate estimate of\nthe reconstruction using a simple linear back-projection and training a deep\nneural network to refine this to the actual reconstruction. We apply our method\nto simulated ultrasound data and demonstrate dramatic improvements in image\nquality compared to the delay-and-sum approach and the linear model-based\nreconstruction approach.\n