A Tomographic Reconstruction Method using Coordinate-based Neural Network with Spatial Regularization

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Tomographic reconstruction is concerned with computing the cross-sections of an object from a finite number of projections. Many conventional methods represent the cross-sections as images on a regular grid. In this paper, we study a recent coordinatebased neural network for tomographic reconstruction, where the network inputs a spatial coordinate and outputs the attenuation coefficient on the coordinate. This coordinate-based network allows the continuous representation of an object. Based on this network, we propose a spatial regularization term, to obtain a high-quality reconstruction. Experimental results on synthetic data show that the regularization term improves the reconstruction quality significantly, compared to the baseline. We also provide an ablation study for different architecture configurations and hyper-parameters.
Original languageEnglish
Title of host publicationProceedings of Northern Lights Deep Learning Workshop 2021
Number of pages7
PublisherPresses Universitaires du Septentrion
Publication statusAccepted/In press - 2021
EventNorthern Lights Deep Learning Workshop 2021
- Virtual event, Tromsø , Norway
Duration: 18 Jan 202120 Jan 2021
http://www.nldl.org

Workshop

WorkshopNorthern Lights Deep Learning Workshop 2021
LocationVirtual event
CountryNorway
CityTromsø
Period18/01/202120/01/2021
Internet address

Fingerprint Dive into the research topics of 'A Tomographic Reconstruction Method using Coordinate-based Neural Network with Spatial Regularization'. Together they form a unique fingerprint.

Cite this