Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement
Phase contrast computed tomography (PCCT) provides an effective non-destructive testing tool for weak absorption objects. Limited by the phase stepping principle and radiation dose requirement, sparse-view sampling is usually performed in PCCT, introducing severe artifacts in reconstruction. In this...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-05-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/13/10/6051 |
_version_ | 1797601284746903552 |
---|---|
author | Changsheng Zhang Jian Fu Gang Zhao |
author_facet | Changsheng Zhang Jian Fu Gang Zhao |
author_sort | Changsheng Zhang |
collection | DOAJ |
description | Phase contrast computed tomography (PCCT) provides an effective non-destructive testing tool for weak absorption objects. Limited by the phase stepping principle and radiation dose requirement, sparse-view sampling is usually performed in PCCT, introducing severe artifacts in reconstruction. In this paper, we report a dual-domain (i.e., the projection sinogram domain and image domain) enhancement framework based on deep learning (DL) for PCCT with sparse-view projections. It consists of two convolutional neural networks (CNN) in dual domains and the phase contrast Radon inversion layer (PCRIL) to connect them. PCRIL can achieve PCCT reconstruction, and it allows the gradients to backpropagate from the image domain to the projection sinogram domain while training. Therefore, parameters of CNNs in dual domains are updated simultaneously. It could overcome the limitations that the enhancement in the image domain causes blurred images and the enhancement in the projection sinogram domain introduces unpredictable artifacts. Considering the grating-based PCCT as an example, the proposed framework is validated and demonstrated with experiments of the simulated datasets and experimental datasets. This work can generate high-quality PCCT images with given incomplete projections and has the potential to push the applications of PCCT techniques in the field of composite imaging and biomedical imaging. |
first_indexed | 2024-03-11T03:58:59Z |
format | Article |
id | doaj.art-c4f7963961a940e7ba153f986a0396b0 |
institution | Directory Open Access Journal |
issn | 2076-3417 |
language | English |
last_indexed | 2024-03-11T03:58:59Z |
publishDate | 2023-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Applied Sciences |
spelling | doaj.art-c4f7963961a940e7ba153f986a0396b02023-11-18T00:19:48ZengMDPI AGApplied Sciences2076-34172023-05-011310605110.3390/app13106051Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain EnhancementChangsheng Zhang0Jian Fu1Gang Zhao2School of Mechanical Engineering and Automation, Beihang University, Beijing 100190, ChinaSchool of Mechanical Engineering and Automation, Beihang University, Beijing 100190, ChinaSchool of Mechanical Engineering and Automation, Beihang University, Beijing 100190, ChinaPhase contrast computed tomography (PCCT) provides an effective non-destructive testing tool for weak absorption objects. Limited by the phase stepping principle and radiation dose requirement, sparse-view sampling is usually performed in PCCT, introducing severe artifacts in reconstruction. In this paper, we report a dual-domain (i.e., the projection sinogram domain and image domain) enhancement framework based on deep learning (DL) for PCCT with sparse-view projections. It consists of two convolutional neural networks (CNN) in dual domains and the phase contrast Radon inversion layer (PCRIL) to connect them. PCRIL can achieve PCCT reconstruction, and it allows the gradients to backpropagate from the image domain to the projection sinogram domain while training. Therefore, parameters of CNNs in dual domains are updated simultaneously. It could overcome the limitations that the enhancement in the image domain causes blurred images and the enhancement in the projection sinogram domain introduces unpredictable artifacts. Considering the grating-based PCCT as an example, the proposed framework is validated and demonstrated with experiments of the simulated datasets and experimental datasets. This work can generate high-quality PCCT images with given incomplete projections and has the potential to push the applications of PCCT techniques in the field of composite imaging and biomedical imaging.https://www.mdpi.com/2076-3417/13/10/6051phase contrast computed tomographysparse-view samplingdual domainconvolutional neural networkradon inversion layer |
spellingShingle | Changsheng Zhang Jian Fu Gang Zhao Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement Applied Sciences phase contrast computed tomography sparse-view sampling dual domain convolutional neural network radon inversion layer |
title | Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement |
title_full | Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement |
title_fullStr | Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement |
title_full_unstemmed | Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement |
title_short | Learning from Projection to Reconstruction: A Deep Learning Reconstruction Framework for Sparse-View Phase Contrast Computed Tomography via Dual-Domain Enhancement |
title_sort | learning from projection to reconstruction a deep learning reconstruction framework for sparse view phase contrast computed tomography via dual domain enhancement |
topic | phase contrast computed tomography sparse-view sampling dual domain convolutional neural network radon inversion layer |
url | https://www.mdpi.com/2076-3417/13/10/6051 |
work_keys_str_mv | AT changshengzhang learningfromprojectiontoreconstructionadeeplearningreconstructionframeworkforsparseviewphasecontrastcomputedtomographyviadualdomainenhancement AT jianfu learningfromprojectiontoreconstructionadeeplearningreconstructionframeworkforsparseviewphasecontrastcomputedtomographyviadualdomainenhancement AT gangzhao learningfromprojectiontoreconstructionadeeplearningreconstructionframeworkforsparseviewphasecontrastcomputedtomographyviadualdomainenhancement |