Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Mar;52(2):286-292.
doi: 10.12182/20210360506.

[Noise Reduction Effect of Deep-learning-based Image Reconstruction Algorithms in Thin-section Chest CT]

[Article in Chinese]
Wen Zeng  1 , Ling-Ming Zeng  1 , Xu Xu  1 , Si-Xian Hu  1 , Ke-Ling Liu  1 , Jin-Ge Zhang  1 , Wan-Lin Peng  1 , Chun-Chao Xia  1 , Zhen-Lin Li  1
Affiliations

[Noise Reduction Effect of Deep-learning-based Image Reconstruction Algorithms in Thin-section Chest CT]

[Article in Chinese]
Wen Zeng et al. Sichuan Da Xue Xue Bao Yi Xue Ban. 2021 Mar.

Abstract
in English, Chinese

Objective: To evaluate the noise reduction effect of deep learning-based reconstruction algorithms in thin-section chest CT images by analyzing images reconstructed with filtered back projection (FBP), adaptive statistical iterative reconstruction (ASIR), and deep learning image reconstruction (DLIR) algorithms.

Methods: The chest CT scan raw data of 47 patients were included in this study. Images of 0.625 mm were reconstructed using six reconstruction methods, including FBP, ASIR hybrid reconstruction (ASIR50%, ASIR70%), and deep learning low, medium and high modes (DL-L, DL-M, and DL-H). After the regions of interest were outlined in the aorta, skeletal muscle and lung tissue of each group of images, the CT values, SD values and signal-to-noise ratio (SNR) of the regions of interest were measured, and two radiologists evaluated the image quality.

Results: CT values, SD values and SNR of the images obtained by the six reconstruction methods showed statistically significant difference ( P<0.001). There were statistically significant differences in the image quality scores of the six reconstruction methods ( P<0.001). Images reconstruced with DL-H have the lowest noise and the highest overall quality score.

Conclusion: The model based on deep learning can effectively reduce the noise of thin-section chest CT images and improve the image quality. Among the three deep-learning models, DL-H showed the best noise reduction effect.

目的: 为了评估基于深度学习的重建算法在胸部薄层计算机断层扫描（computed tomography，CT）图像中的降噪效果，对滤波反投影重建（filtered back projection，FBP）、自适应统计迭代重建（adaptive statistical iterative reconstruction，ASIR）与深度学习图像重建（deep learning image reconstruction，DLIR）图像进行分析。

方法: 回顾性纳入47例患者胸部CT平扫原始数据，利用FBP，ASIR混合重建（ASIR50%、ASIR70%），深度学习低、中、高3种模式（DL-L、DL-M、DL-H）共6种，重建出0.625 mm的图像。在每组图像的主动脉内、骨骼肌以及肺组织内分别勾画感兴趣区，测量感兴趣区内的CT值、SD值和信噪比（signal-to-noise ratio，SNR）进行客观评价，并对图像进行主观评价。

结果: 6种重建图像CT、SD和SNR值的差异有统计学意义（P<0.001）。6种重建图像主观评分差异有统计学意义（P<0.001）。DLIR在主动脉和骨骼肌处的图像噪声明显低于传统的FBP和ASIR，图像质量能够满足临床需求。而且呈现出DL-H降噪效果最佳、噪声最低，ASIR70%、DL-M、ASIR50%、DL-L、FBP 图像噪声依次增加。通过主观评分的比较发现，DL-H的图像整体质量有明显的提升，但不能使肺纹理重建更清晰。

结论: 基于深度学习的模型能够有效减少胸部薄层CT图像的噪声，提高图像的质量。而在3种深度学习模型中，DL-H的降噪效能最佳。

Keywords: Computed tomography; Deep learning; The noise reduction algorithm.

PubMed Disclaimer

Figures

图 1
图 1
The algorithm training process flowchart 算法训练流程
图 2
图 2
Flowchart of the study 研究流程图
图 3
图 3
A 62-year-old man underwent thin-section CT chest scan, and six images were obtained with FBP, ASIR50%, ASIR70%, DL-L, DL-M, and DL-H reconstruction 62岁男性行胸部薄层CT扫描，FBP、ASIR50%、ASIR70%、DL-L、DL-M、DL-H重建图像
图 4
图 4
A 62-year-old man underwent thin-section CHEST CT scan. With six reconstruction methods, including FBP, ASIR50%, ASIR70%, DL-L, DL-M, and DL-H, a magnified view of local musculoskeletal was obtained 62岁男性行胸部薄层CT扫描，经过FBP、ASIR50%、ASIR70%、DL-L、DL-M、DL-H重建后骨骼肌局部放大图
图 5
图 5
A thin section CT scan of the chest of a 65-year-old woman revealed a 1 cm ground glass nodule in the upper lobe of the right lung. The scan data were reconstructed in six ways, and a local magnified view of the lung nodule was shown 65岁女性胸部薄层CT扫描发现右肺上叶1 cm磨玻璃结节，扫描数据经过6种方式重建后，肺结节的局部放大图
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. ABERLE D R, BERG C D, BLACK W C, et al The National Lung Screening Trial: overview and study design. Radiology. 2011;258(1):243–253. doi: 10.1148/radiol.10091808. - DOI - PMC - PubMed
    1. ABERLE D R, ADAMS A M, BERG C D, et al Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365(5):395–409. doi: 10.1056/NEJMoa1102873. - DOI - PMC - PubMed
    1. KIM M J, PARK C H, CHOI S J, et al Multidetector computed tomography chest examinations with Low-Kilovoltage protocols in adults. J Comput Assist Tomogr. 2009;33(3):416–421. doi: 10.1097/RCT.0b013e318181fab5. - DOI - PubMed
    1. GREFFIER J, MACRI F, LARBI A, et al Dose reduction with iterative reconstruction: optimization of CT protocols in clinical practice. Diagn Interv Imaging. 2015;96(5):477–486. doi: 10.1016/j.diii.2015.02.007. - DOI - PubMed
    1. MACRI F, GREFFIER J, PEREIRA F, et al Value of ultra-low-dose chest CT with iterative reconstruction for selected emergency room patients with acute dyspnea. Eur J Radiol. 2016;85(9):1637–1644. doi: 10.1016/j.ejrad.2016.06.024. - DOI - PubMed

LinkOut - more resources