University of Oulu

S. Bussod et al., "Convolutional Neural Network for Material Decomposition in Spectral CT Scans," 2020 28th European Signal Processing Conference (EUSIPCO), Amsterdam, 2021, pp. 1259-1263,

Convolutional neural network for material decomposition in spectral CT scans

Saved in:
Author: Bussod, Suzanne1; Abascal, Juan F.P.J.1; Arridge, Simon2;
Organizations: 1CNRS UMR 5220, U1206 Univ. Lyon, INSA-Lyon Lyon, France
2Department of Computer Science, University College London, London, United Kingdom
3Research Unit of Mathematical Sciences, University of Oulu, Oulu, Finland
4Department of Computer Science, University College London
5B3OA, UMR CNRS 7052, Inserm U1271 University Paris Diderot Paris, France
6The European Synchrotron Radiation Facility, Grenoble, France
Format: article
Version: accepted version
Access: open
Online Access: PDF Full Text (PDF, 0.5 MB)
Persistent link:
Language: English
Published: Institute of Electrical and Electronics Engineers, 2021
Publish Date: 2021-01-26


Spectral computed tomography acquires energy-resolved data that allows recovery of densities of constituents of an object. This can be achieved by decomposing the measured spectral projection into material projections, and passing these decomposed projections through a tomographic reconstruction algorithm, to get the volumetric mass density of each material. Material decomposition is a nonlinear inverse problem that has been traditionally solved using model-based material decomposition algorithms. However, the forward model is difficult to estimate in real prototypes. Moreover, the traditional regularizers used to stabilized inversions are not fully relevant in the projection domain. In this study, we propose a deep-learning method for material decomposition in the projection domain. We validate our methodology with numerical phantoms of human knees that are created from synchrotron CT scans. We consider four different scans for training, and one for validation. The measurements are corrupted by Poisson noise, assuming that at most 10⁵ photons hit the detector. Compared to a regularized Gauss-Newton algorithm, the proposed deep-learning approach provides a compromise between noise and resolution, which reduces the computation time by a factor of 100.

see all

Series: European Signal Processing Conference
ISSN: 2219-5491
ISSN-E: 2076-1465
ISSN-L: 2219-5491
ISBN: 978-9-0827-9705-3
ISBN Print: 978-1-7281-5001-7
Pages: 1259 - 1263
DOI: 10.23919/Eusipco47968.2020.9287781
Host publication: 2020 28th European Signal Processing Conference (EUSIPCO)
Conference: European Signal Processing Conference
Type of Publication: A4 Article in conference proceedings
Field of Science: 113 Computer and information sciences
217 Medical engineering
Funding: We acknowledge the support of the ANR project SALTO (ANR-17-CE19-0011-01). This study was also performed within the framework of LabExPRIMES (ANR-11-LABX-0063) of the Université de Lyon. The project also received funding from the European Union Horizon 2020 Research and Innovation Programme under both grant agreement N◦668142 and Marie Sklodowska-Curie grant agreement N◦701915. This study was also carried out in the context of France Life Imaging (grant ANR-11-INBS-0006). A PhD is funded by the ED EEA of Université de Lyon.
Copyright information: © 2021 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.