Scale-Agnostic Super-Resolution in MRI using Feature-Based Coordinate Networks

Dave Van Veen, Rogier Van der Sluijs, Batu Ozturkler, Arjun D Desai, Christian Bluethgen, Robert D. Boutin, Marc H. Willis, Gordon Wetzstein, David B. Lindell, Shreyas Vasanawala, John M. Pauly, Akshay Chaudhari

Keywords: Coordinate networks, super-resolution, MRI

TL;DR: We present a framework for scale-agnostic MRI super-resolution by using the continuous representation of a coordinate network to decode the latent space.

Abstract: We propose using a coordinate network as a decoder for MRI super-resolution. The continuous signal representation of coordinate networks enables this approach to be scale-agnostic, i.e. training over a continuous range of scales and querying at arbitrary resolutions. We evaluate the benefits of denoising for coordinate networks and also compare our method to a convolutional decoder using image quality metrics and a radiologist study.

Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: novel methodological ideas without extensive validation

Primary Subject Area: Image Acquisition and Reconstruction

Secondary Subject Area: Application: Radiology

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.