Computed tomography (CT) is a beneficial imaging tool for diagnostic purposes. CT scans provide detailed information concerning the internal anatomic structures of a patient, but present higher radiation dose and costs compared to X-ray imaging. In this paper, we build on previous research to convert orthogonal X-ray images into simulated CT volumes by exploring larger datasets and various model structures. Significant model variations include UNet architectures, custom connections, activation functions, loss functions, optimizers, and a novel back projection approach.
@article{arxiv.2403.00771,
title = {XProspeCT: CT Volume Generation from Paired X-Rays},
author = {Benjamin Paulson and Joshua Goldshteyn and Sydney Balboni and John Cisler and Andrew Crisler and Natalia Bukowski and Julia Kalish and Theodore Colwell},
journal= {arXiv preprint arXiv:2403.00771},
year = {2024}
}
Comments
Originally submitted as part of the MICS 2023 Undergraduate Paper Competition