Welcome to Francis Academic Press

Academic Journal of Computing & Information Science, 2021, 4(1); doi: 10.25236/AJCIS.2021.040108.

Super resolution reconstruction of PET images based on deep learning


Mingdong Liu1,*, Yunye Feng2, Chang Liu1, Mingyu Gao1, Chenglin Yan1

Corresponding Author:
​Mingdong Liu

1College of Medical Biological Information Engineering, Northeastern University, Shenyang 110167, China

2College of Software, Northeastern University, Shenyang 110167, China

*Corresponding Author


PET-CT is a high-end medical imaging technology that can detect tumors in the early stage, and has been widely used in clinical practice.Obtaining high-quality images is a prerequisite for doctors to make a correct diagnosis.For PET itself, how to continuously improve its resolution is a challenging problem.In this study, Generative Adversarial Networks (GAN) combined with Wasserstein distance and gradient penalty technology was used to achieve PET image super-resolution reconstruction.The experimental results show that the structural similarity of the final image reaches more than 60%.


PET, super resolution, deep learning, GAN

Cite This Paper

Mingdong Liu, Yunye Feng, Chang Liu, Mingyu Gao, Chenglin Yan. Super resolution reconstruction of PET images based on deep learning. Academic Journal of Computing & Information Science (2021), Vol. 4, Issue 1: 48-52. https://doi.org/10.25236/AJCIS.2021.040108.


[1] C. Ledig, L. Theis, F. Huszar, J. Caballero, A. Cunningham, A.Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang, W. Shi(2017), Photo-realistic single image super-resolution using a generative adversarial network, in: CVPR, pp. 4681–4690. 

[2] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, Y. Bengio(2014), Generative adversarial nets, in: NIPS, pp. 2672–2680.

[3] S.-J. Park, H. Son, S. Cho, K.-S. Hong, S. Lee, Srfeat(2018): Single image super-resolution with feature discrimination, in: ECCV,, pp. 439-455. 

[4] John A Kennedy; Ora Israel; Alex Frenkel; Rachel Bar-Shalom; Haim Azhari(2007)Improved image fusion in PET/CT using hybrid image reconstruction and super-resolution. International journal of biomedical imaging.1687-4188;1687-4188.

[5] II Jun Ahn 1 ; Ji Hye Kim 1 ; Yongjin Chang 1 ; Woo Hyun Nam 1 ; Jong Beom Ra 1 (2007)Super-Resolution Reconstruction of 3D PET Images Using Two Respiratory-Phase Low-Dose CT Images. IEEE TRANSACTIONS ON RADIATION AND PLASMA MEDICAL SCIENCES Vol.1 No.1 49-55 

[6] Hu, Zhanli 1; Wang, Ying 1,2 ; Zhang, Xuezhu 3 ; Zhang, Mengxi 3 ; Yang, Yongfeng 1 ; Liu, Xin 1 ; Zheng, Hairong 1 ; Liang, Dong 1 (dong.liang@siat.ac.cn).(2019) Super-resolution of PET image based on dictionary learning and random forests. Nuclear Instruments & Methods in Physics Research Section A.Vol.927 320-329.

[7] Malczewski, Krzysztof 1(2020). Algorithms.2020 Vol.13 No.5 129.

[8] X. Wang, K. Yu, S. Wu, J. Gu, Y. Liu, C. Dong, C. C. Loy, Y. Qiao, X. Tang(2018), EsrGAN: Enhanced super-resolution generative adversarial net-works, in: ECCV Workshop

[9] I.Gulrajani,et al. “Improved training of Wasserstein GANs.” Advances in neural information processing systems, 2017, pp.5767-5777.

[10] Q. Lyu, H. Shan and G. Wang, "MRI Super-Resolution With Ensemble Learning and Complementary Priors," in IEEE Transactions on Computational Imaging, vol. 6, pp. 615-624, 2020, doi: 10.1109/TCI.2020.2964201.