A deep-learning based algorithm for image reconstruction in Compton tomography

Ishak Ayad; Cécilia Tarpau; Javier Cebeiro; Mai K. Nguyen

doi:10.1109/IPTA59101.2023.10320071

A deep-learning based algorithm for image reconstruction in Compton tomography

Ishak Ayad^*, Cécilia Tarpau, Javier Cebeiro, Mai K. Nguyen

^*Corresponding author for this work

School of Mathematical & Computer Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In this paper we study the applications of deep-learning to the problem of image reconstruction in Compton scatter tomography, a field where deep-learning techniques are still unexplored. Particularly, we focus on a new design with uncollimated detectors that simplifies some previous configurations of Compton scanners. The system inherits attractive advantages such as non-moving components and the ability to combine with other imaging modes. Since there is no an analytic inverse formula for image reconstruction, we developed a GAN based algorithm that provides an efficient mapping between data and image domains. We compare our method against several algorithmic approaches and show that high quality image reconstruction is feasible. Results encourage further research in the application of deep-learning reconstruction techniques in Compton scatter tomography, particularly when inverse reconstruction formulas are unknown.

Original language	English
Title of host publication	12th International Conference on Image Processing Theory, Tools and Applications 2023
Publisher	IEEE
ISBN (Electronic)	9798350325416
DOIs	https://doi.org/10.1109/IPTA59101.2023.10320071
Publication status	Published - 21 Nov 2023
Event	12th International Conference on Image Processing Theory, Tools and Application 2023 - Paris, France Duration: 16 Oct 2023 → 19 Oct 2023

Conference

Conference	12th International Conference on Image Processing Theory, Tools and Application 2023
Abbreviated title	IPTA 2023
Country/Territory	France
City	Paris
Period	16/10/23 → 19/10/23

Keywords

circular Radon transform
Compton tomography
deep-learning image reconstruction
GANs

ASJC Scopus subject areas

Artificial Intelligence
Computer Science Applications
Computer Vision and Pattern Recognition
Signal Processing
Radiology Nuclear Medicine and imaging

Access to Document

10.1109/IPTA59101.2023.10320071

Cite this

@inproceedings{d225faff6f4c4c9e84df53ab6dc101bc,

title = "A deep-learning based algorithm for image reconstruction in Compton tomography",

abstract = "In this paper we study the applications of deep-learning to the problem of image reconstruction in Compton scatter tomography, a field where deep-learning techniques are still unexplored. Particularly, we focus on a new design with uncollimated detectors that simplifies some previous configurations of Compton scanners. The system inherits attractive advantages such as non-moving components and the ability to combine with other imaging modes. Since there is no an analytic inverse formula for image reconstruction, we developed a GAN based algorithm that provides an efficient mapping between data and image domains. We compare our method against several algorithmic approaches and show that high quality image reconstruction is feasible. Results encourage further research in the application of deep-learning reconstruction techniques in Compton scatter tomography, particularly when inverse reconstruction formulas are unknown.",

keywords = "circular Radon transform, Compton tomography, deep-learning image reconstruction, GANs",

author = "Ishak Ayad and C{\'e}cilia Tarpau and Javier Cebeiro and Nguyen, {Mai K.}",

note = "Funding Information: This research work is supported by grants from R{\'e}gion Ile-de-France (in Mathematics and Innovation) and the resources of IDRIS (Institut du d{\'e}veloppement et des ressources en informatique scientifique). Publisher Copyright: {\textcopyright} 2023 IEEE.; 12th International Conference on Image Processing Theory, Tools and Application 2023, IPTA 2023 ; Conference date: 16-10-2023 Through 19-10-2023",

year = "2023",

month = nov,

day = "21",

doi = "10.1109/IPTA59101.2023.10320071",

language = "English",

booktitle = "12th International Conference on Image Processing Theory, Tools and Applications 2023",

publisher = "IEEE",

address = "United States",

}

Ayad, I, Tarpau, C, Cebeiro, J & Nguyen, MK 2023, A deep-learning based algorithm for image reconstruction in Compton tomography. in 12th International Conference on Image Processing Theory, Tools and Applications 2023., 10320071, IEEE, 12th International Conference on Image Processing Theory, Tools and Application 2023, Paris, France, 16/10/23. https://doi.org/10.1109/IPTA59101.2023.10320071

TY - GEN

T1 - A deep-learning based algorithm for image reconstruction in Compton tomography

AU - Ayad, Ishak

AU - Tarpau, Cécilia

AU - Cebeiro, Javier

AU - Nguyen, Mai K.

N1 - Funding Information: This research work is supported by grants from Région Ile-de-France (in Mathematics and Innovation) and the resources of IDRIS (Institut du développement et des ressources en informatique scientifique). Publisher Copyright: © 2023 IEEE.

PY - 2023/11/21

Y1 - 2023/11/21

N2 - In this paper we study the applications of deep-learning to the problem of image reconstruction in Compton scatter tomography, a field where deep-learning techniques are still unexplored. Particularly, we focus on a new design with uncollimated detectors that simplifies some previous configurations of Compton scanners. The system inherits attractive advantages such as non-moving components and the ability to combine with other imaging modes. Since there is no an analytic inverse formula for image reconstruction, we developed a GAN based algorithm that provides an efficient mapping between data and image domains. We compare our method against several algorithmic approaches and show that high quality image reconstruction is feasible. Results encourage further research in the application of deep-learning reconstruction techniques in Compton scatter tomography, particularly when inverse reconstruction formulas are unknown.

AB - In this paper we study the applications of deep-learning to the problem of image reconstruction in Compton scatter tomography, a field where deep-learning techniques are still unexplored. Particularly, we focus on a new design with uncollimated detectors that simplifies some previous configurations of Compton scanners. The system inherits attractive advantages such as non-moving components and the ability to combine with other imaging modes. Since there is no an analytic inverse formula for image reconstruction, we developed a GAN based algorithm that provides an efficient mapping between data and image domains. We compare our method against several algorithmic approaches and show that high quality image reconstruction is feasible. Results encourage further research in the application of deep-learning reconstruction techniques in Compton scatter tomography, particularly when inverse reconstruction formulas are unknown.

KW - circular Radon transform

KW - Compton tomography

KW - deep-learning image reconstruction

KW - GANs

UR - http://www.scopus.com/inward/record.url?scp=85179552513&partnerID=8YFLogxK

U2 - 10.1109/IPTA59101.2023.10320071

DO - 10.1109/IPTA59101.2023.10320071

M3 - Conference contribution

AN - SCOPUS:85179552513

BT - 12th International Conference on Image Processing Theory, Tools and Applications 2023

PB - IEEE

T2 - 12th International Conference on Image Processing Theory, Tools and Application 2023

Y2 - 16 October 2023 through 19 October 2023

ER -

A deep-learning based algorithm for image reconstruction in Compton tomography

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this