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
. 2024 Apr;8(4):333-347.
doi: 10.1109/trpms.2023.3349194. Epub 2024 Jan 2.

A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches

Alexandre Bousse  1 , Venkata Sai Sundar Kandarpa  1 , Kuangyu Shi  2 , Kuang Gong  3 , Jae Sung Lee  4 , Chi Liu  5 , Dimitris Visvikis  1
Affiliations

A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches

Alexandre Bousse et al. IEEE Trans Radiat Plasma Med Sci. 2024 Apr.

Abstract

Low-dose emission tomography (ET) plays a crucial role in medical imaging, enabling the acquisition of functional information for various biological processes while minimizing the patient dose. However, the inherent randomness in the photon counting process is a source of noise which is amplified low-dose ET. This review article provides an overview of existing post-processing techniques, with an emphasis on deep neural network (NN) approaches. Furthermore, we explore future directions in the field of NN-based low-dose ET. This comprehensive examination sheds light on the potential of deep learning in enhancing the quality and resolution of low-dose ET images, ultimately advancing the field of medical imaging.

Keywords: Deep Learning; Low-Dose; PET; SPECT.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.
Main steps in clinical ET. The patient is initially administered with a radioactive tracer. The emission is captured by an imaging system. The mapping of the raw detector data to an image is performed by an image reconstruction method. Finally, after post-processing, the image is used for diagnosis.
Fig. 2.
Fig. 2.
Representation of supervised deep learning-based method from Xu et al. [30].
Fig. 3.
Fig. 3.
Structure of PT-WGAN from Gong et al. [54].
Fig. 4.
Fig. 4.
(a) The architecture of the Swin transformer network and (b) consecutive blocks of the Swin transformer. Reprint from Liu et al. [80].
Fig. 5.
Fig. 5.
Schematic of N2N network model improved by the incorporation of WTs. Reprint from Kang et al. [107].
Fig. 6.
Fig. 6.
Full count image, noisy input, Gaussian filtered and denoised images using the N2N without and with incorporating the trainable WT for clinical data. Reprint from Kang et al. [107].
See this image and copyright information in PMC

References

    1. Mattsson S and Söderberg M, “Radiation dose management in CT, SPECT/CT and PET/CT techniques,” Radiation protection dosimetry, vol. 147, no. 1-2, pp. 13–21, 2011. - PubMed
    1. Gimelli A, Achenbach S, Buechel RR, Edvardsen T, Francone M, Gaemperli O, Hacker M, Hyafil F, Kaufmann PA, Lancellotti P, et al. “Strategies for radiation dose reduction in nuclear cardiology and cardiac computed tomography imaging: A report from the European Association of Cardiovascular Imaging (EACVI), the Cardiovascular Committee of European Association of Nuclear Medicine (EANM), and the European Society of Cardiovascular Radiology (ESCR),” European heart journal, vol. 39, no. 4, pp. 286–296, 2018. - PubMed
    1. Anger HO, “Scintillation camera,” Review of scientific instruments, vol. 29, no. 1, pp. 27–33, 1958.
    1. Natterer F, The Mathematics of Computerized Tomography (Classics in Applied Mathematics). SIAM, 2001.
    1. Shepp LA and Vardi Y, “Maximum likelihood reconstruction for emission tomography,” IEEE transactions on medical imaging, vol. 1, no. 2, pp. 113–122, 1982 - PubMed

LinkOut - more resources