Literature DB >> 31131901

Ultra-low-dose PET reconstruction using generative adversarial network with feature matching and task-specific perceptual loss.

Jiahong Ouyang1, Kevin T Chen1, Enhao Gong2, John Pauly3, Greg Zaharchuk1,2.   

Abstract

PURPOSE: Our goal was to use a generative adversarial network (GAN) with feature matching and task-specific perceptual loss to synthesize standard-dose amyloid Positron emission tomography (PET) images of high quality and including accurate pathological features from ultra-low-dose PET images only.
METHODS: Forty PET datasets from 39 participants were acquired with a simultaneous PET/MRI scanner following injection of 330 ± 30 MBq of the amyloid radiotracer 18F-florbetaben. The raw list-mode PET data were reconstructed as the standard-dose ground truth and were randomly undersampled by a factor of 100 to reconstruct 1% low-dose PET scans. A 2D encoder-decoder network was implemented as the generator to synthesize a standard-dose image and a discriminator was used to evaluate them. The two networks contested with each other to achieve high-visual quality PET from the ultra-low-dose PET. Multi-slice inputs were used to reduce noise by providing the network with 2.5D information. Feature matching was applied to reduce hallucinated structures. Task-specific perceptual loss was designed to maintain the correct pathological features. The image quality was evaluated by peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and root mean square error (RMSE) metrics with and without each of these modules. Two expert radiologists were asked to score image quality on a 5-point scale and identified the amyloid status (positive or negative).
RESULTS: With only low-dose PET as input, the proposed method significantly outperformed Chen et al.'s method (Chen et al. Radiology. 2018;290:649-656) (which shows the best performance in this task) with the same input (PET-only model) by 1.87 dB in PSNR, 2.04% in SSIM, and 24.75% in RMSE. It also achieved comparable results to Chen et al.'s method which used additional magnetic resonance imaging (MRI) inputs (PET-MR model). Experts' reading results showed that the proposed method could achieve better overall image quality and maintain better pathological features indicating amyloid status than both PET-only and PET-MR models proposed by Chen et al.
CONCLUSION: Standard-dose amyloid PET images can be synthesized from ultra-low-dose images using GAN. Applying adversarial learning, feature matching, and task-specific perceptual loss are essential to ensure image quality and the preservation of pathological features.
© 2019 American Association of Physicists in Medicine.

Entities:  

Keywords:  GAN; deep learning; ultra-low-dose PET reconstruction

Year:  2019        PMID: 31131901      PMCID: PMC6692211          DOI: 10.1002/mp.13626

Source DB:  PubMed          Journal:  Med Phys        ISSN: 0094-2405            Impact factor:   4.071


  37 in total

Review 1.  Applications of artificial intelligence in nuclear medicine image generation.

Authors:  Zhibiao Cheng; Junhai Wen; Gang Huang; Jianhua Yan
Journal:  Quant Imaging Med Surg       Date:  2021-06

2.  Artificial Intelligence in Neuroradiology: Current Status and Future Directions.

Authors:  Y W Lui; P D Chang; G Zaharchuk; D P Barboriak; A E Flanders; M Wintermark; C P Hess; C G Filippi
Journal:  AJNR Am J Neuroradiol       Date:  2020-07-30       Impact factor: 3.825

3.  Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.

Authors:  Yan-Ran Joyce Wang; Lucia Baratto; K Elizabeth Hawk; Ashok J Theruvath; Allison Pribnow; Avnesh S Thakor; Sergios Gatidis; Rong Lu; Santosh E Gummidipundi; Jordi Garcia-Diaz; Daniel Rubin; Heike E Daldrup-Link
Journal:  Eur J Nucl Med Mol Imaging       Date:  2021-02-01       Impact factor: 9.236

4.  Next generation research applications for hybrid PET/MR and PET/CT imaging using deep learning.

Authors:  Greg Zaharchuk
Journal:  Eur J Nucl Med Mol Imaging       Date:  2019-06-29       Impact factor: 9.236

Review 5.  3D/4D Reconstruction and Quantitative Total Body Imaging.

Authors:  Jinyi Qi; Samuel Matej; Guobao Wang; Xuezhu Zhang
Journal:  PET Clin       Date:  2021-01

Review 6.  Artificial intelligence for nuclear medicine in oncology.

Authors:  Kenji Hirata; Hiroyuki Sugimori; Noriyuki Fujima; Takuya Toyonaga; Kohsuke Kudo
Journal:  Ann Nucl Med       Date:  2022-01-14       Impact factor: 2.668

Review 7.  A review on AI in PET imaging.

Authors:  Keisuke Matsubara; Masanobu Ibaraki; Mitsutaka Nemoto; Hiroshi Watabe; Yuichi Kimura
Journal:  Ann Nucl Med       Date:  2022-01-14       Impact factor: 2.668

Review 8.  Machine learning in quantitative PET: A review of attenuation correction and low-count image reconstruction methods.

Authors:  Tonghe Wang; Yang Lei; Yabo Fu; Walter J Curran; Tian Liu; Jonathon A Nye; Xiaofeng Yang
Journal:  Phys Med       Date:  2020-07-29       Impact factor: 2.685

9.  Generative adversarial network based regularized image reconstruction for PET.

Authors:  Zhaoheng Xie; Reheman Baikejiang; Tiantian Li; Xuezhu Zhang; Kuang Gong; Mengxi Zhang; Wenyuan Qi; Evren Asma; Jinyi Qi
Journal:  Phys Med Biol       Date:  2020-06-23       Impact factor: 3.609

10.  True ultra-low-dose amyloid PET/MRI enhanced with deep learning for clinical interpretation.

Authors:  Kevin T Chen; Tyler N Toueg; Mary Ellen Irene Koran; Guido Davidzon; Michael Zeineh; Dawn Holley; Harsh Gandhi; Kim Halbert; Athanasia Boumis; Gabriel Kennedy; Elizabeth Mormino; Mehdi Khalighi; Greg Zaharchuk
Journal:  Eur J Nucl Med Mol Imaging       Date:  2021-01-08       Impact factor: 9.236

View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.