Deep Learning Image Reconstruction for CT: Technical Principles and Clinical Prospects
Реконструкция изображений КТ с использованием глубокого обучения: технические принципы и клинические перспективы
2023-01-31
SCID: 54.1/b2m2t34c
Discuss with AI
computed tomographydeep learning reconstructionfiltered back projectionmetal artifact reductionmodel-based iterative reconstruction
Figures from the paper
Abstract (AI)
Filtered back projection (FBP) has been the standard CT image reconstruction method for 4 decades. A simple, fast, and reliable technique, FBP has delivered high-quality images in several clinical applications. However, with faster and more advanced CT scanners, FBP has become increasingly obsolete. Higher image noise and more artifacts are especially noticeable in lower-dose CT imaging using FBP. This performance gap was partly addressed by model-based iterative reconstruction (MBIR). Yet, its "plastic" image appearance and long reconstruction times have limited widespread application. Hybrid iterative reconstruction partially addressed these limitations by blending FBP with MBIR and is currently the state-of-the-art reconstruction technique. In the past 5 years, deep learning reconstruction (DLR) techniques have become increasingly popular. DLR uses artificial intelligence to reconstruct high-quality images from lower-dose CT faster than MBIR. However, the performance of DLR algorithms relies on the quality of data used for model training. Higher-quality training data will become available with photon-counting CT scanners. At the same time, spectral data would greatly benefit from the computational abilities of DLR. This review presents an overview of the principles, technical approaches, and clinical applications of DLR, including metal artifact reduction algorithms. In addition, emerging applications and prospects are discussed.
Key Findings
1
Deep learning reconstruction can generate high-quality lower-dose CT images faster than model-based iterative reconstruction, although performance depends strongly on training-data quality.
2
Filtered back projection remains fast and reliable but increasingly produces excessive noise and artifacts as CT imaging advances, particularly at lower radiation doses.
3
Hybrid iterative reconstruction, combining filtered back projection and model-based iterative reconstruction, is described as the current state-of-the-art approach.
4
Model-based iterative reconstruction reduces some limitations of filtered back projection, but plastic-looking images and long reconstruction times restrict its clinical adoption.
5
Photon-counting CT may provide higher-quality training data, while deep learning reconstruction could enhance spectral CT processing and support metal-artifact reduction and other emerging applications.
Research Object
deep learning reconstruction (DLR) techniques for computed tomography (CT) images
Research Subject
the technical principles, performance advantages and limitations, and clinical applications and prospects of DLR, including low-dose and spectral CT reconstruction and metal artifact reduction
Publication Details
Publication Date
2023-01-31
Journal
Publisher
ISSN
Cited by
469
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest