Keywords: MRI, DWI, Deep Learning, Deep Sets, Image Reconstruction, Liver Diffusion
TL;DR: This short paper proposes a generic approach for efficiently implementing joint reconstruction of image repetitions and evaluates it on the example of abdominal DWI
Abstract: Parallel imaging with multiple receiver coils has become a standard in many MRI applications. Methods based on Deep Learning (DL) were shown to allow higher acceleration factors than conventional methods. In the case of diffusion-weighted imaging (DWI) where multiple repetitions of a slice are acquired, a DL-based reconstruction method should ideally make use of available redundancies. Based on the concept of Deep Sets which outlines a generic approach for operating on set-structured data, this work investigates the benefits of joint reconstruction of image repetitions in DWI. Evaluations show that, compared to separate processing of repetitions, reconstructions can be improved both qualitatively and quantitatively by incorporating simple and computationally inexpensive operations into an existing DL architecture.
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.