WANG Lei,LI Yuanqian.CT Image Generation based on Generative Adversarial Network[J].Journal of Chengdu University of Information Technology,2021,36(03):286-292.[doi:10.16836/j.cnki.jcuit.2021.03.008]
基于生成对抗网络的CT图像生成
- Title:
- CT Image Generation based on Generative Adversarial Network
- 文章编号:
- 2096-1618(2021)03-0286-07
- 关键词:
- 生成对抗网络; 图像生成; CT; MRI-onlyRT
- Keywords:
- generative adversarial networks; image generation; CT; MRI-only RT
- 分类号:
- TP391
- 文献标志码:
- A
- 摘要:
- 准确地预测生成CT图像,在仅磁共振图像引导(MRI-onlyRT)的放疗计划中有着极其重要的作用,使用MRI预测生成CT图像可以避免患者单独进行CT扫描,从而避免额外的辐射剂量。在医学图像跨模态合成中,生成对抗网络(generative adversarial networks, GAN)正成为一种有影响力的方法。研究利用GAN,结合U-Net网络建立鼻咽癌磁共振图像(MRI)与CT图像的映射模型,实现在仅磁共振图像引导的放疗过程中CT图像的预测生成。实验结果表明,建立的网络模型可以生成接近真实数据的CT图像,与改进的U-Net卷积网络生成模型相比,降低了生成图像的模糊度,减小了MAE,提高了PSNR,生成的图片能够更好地展示细节信息。
- Abstract:
- Accurate prediction of CT image generation plays an important role in MRI-only radiotherapy planning.Using MRI prediction to generate CT images can prevent patients from having separate CT scans, thus avoiding additional radiation doses.Generative Adversarial Networks(GAN)is becoming an influential method in cross-modal synthesis of medical images.In this paper, the mapping model of MRI and CT images of nasopharyngeal carcinoma was established by GAN combined with U-Net network, to realize the prediction generation of CT images in the radiotherapy guided by MRI images only. The experimental results show that the model established in this paper can generate CT images close to real data. Compared with the improved U-Net convolutional network generation model, the fuzzy degree of the image is reduced,MAE is reduced and PSNR is improved, and the generated images can better display the detailed information.
参考文献/References:
[1] Edmumd JM,Nyholm T.A review of substitute CT generation for MRI-only radiation therapy[J].Radiat Oncol,2017,26,(1):28.
[2] Kim J,Garbarino K,Schultz L,et al.Dosimetric evaluation of synthetic CT relative to bulk density assignment-based magnetic resonance-only approaches for prostate radiotherapy[J].Radiation Oncology,2015,10(1):1-9.
[3] ShuHui Hsu,Irene Zawisza,Kyle O’Grady,et al. Towards abdominal MRI-based treatment planning using population-based Hounsfield units for bulk density assignment[J].Physics in Medicine & Biology,2018; 63(15):1.
[4] Chin AL,Lin A,Anamalayil S,et al.Feasibility and limitations of bulk density assignment in MRI for head and neck IMRT treatment planning[J].J Appl Clin Med Phys, 2014; 15(5):4851.
[5] QIN Songbing,ZHOU Juying,GONG Wei,et al.Study of the feasibility and precision in dose calculation with the method of bulk density assignment[J].Chinese Journal of Radiation Oncology,2013,22(3):247-249.
[6] Chen Lili,Nguyen B,JonesE,et al.Magnetic resonance-based treatment planning forprostate intensity-modulated mdiothempy:creation of digitally reconstructed radiographs[J].Int J Radiation oncol Biol Phys,2007,68(3):903-911.
[7] D H Ye,D Zikic,B Glocker, et al.Modality propagation: coherent synthesis of subject-specific scans with data-driven regularization[C].International Conference on Medical Image Computing and Computer-Assisted Intervention,2013:606-613.
[8] Roy S,Caress A,Prince JL.Magnetic Resonance Image Example-Based Contrast Synthesis[J].IEEE Transactions on Medical Imaging.2013; 32(12):2348-2363.
[9] F Guerreiro,N Burgos,A Dunlop,et al.Evaluation of a multiatlas CT synthesis approach for MRI-only radiotherapy treatment planning[J].Physica Medica,2017,35:7-17.
[10] SchreibmannE,Nye J A,Schuster DM,et a1.MR-based attenuation correction for hybrid PET-MR brain imaging systems using deformable image registration[J].Med Phys,2010,37(5):2101-2109.
[11] Andreasen D,Van Leemput K,Hansen RH,et a1.Patch-based generation of apseudo CT from conventional MRI sequences forMRI-onlyradiotherapy ofthe brain[J].Med Physics,2015,42(4):1596-1605.
[12] K N D BrouBoni,L Vanquin,A Wagner,et al.Deep MR to CT synthesis using paired data in the pelvic area[J].Physica Medica,2019,68:29.
[13] Jin C B,Kim H,Liu M,et al.Deep CT to MR Synthesis Using Paired and Unpaired Data[J].Sensors(14248220),2019,19(10):2361.
[14] NieD,TrulloR,LianJ,et al.Medical image synthesis with context-aware generative adversarial networks[C].The International Conference on Medical Image Computing and Computer-Assisted Intervention,417-425.
[15] Zhang Z,Yang L,Zheng Y.Translating and segmenting multimodal medical volumes with cycle-and shape-consistency generative adversarial network[C].The IEEE Conference on Computer Vision and Pattern Recognition,2018:9242-9251.
[16] Edmund J M,Nyholm T.A review of substitute CT generation for MRI-only radiation therapy[J].Radiation Oncology,2017,12(1):28.
[17] Wang Y,Yu B,Wang L,et al.3D conditional generative adversarial networks for high-quality PET image estimation at low dose[J].Neuroimage,2018,174:550-562.
[18] Yang G,Yu S,Dong H,et al. DAGAN:deep de-aliasing generative adversarial networks for fast compressed sensing MRI reconstruction[J].IEEE Trans Med Imag,2018,37(6):1310-1321.
[19] I Goodfellow,J PougetAbadie,M Mirza,et al.Generative adversarial nets[C].The International Conference on Neural InformationProcessing Systems,Montréal.Montréal:NIPS,2014:2672-2680.
[20] Mardani M,Gong E,Cheng J Y,et al.Deep generative adversarial networks for compressed sensing automates MRI[J].IEEE Trans Med Imaging,2019,38(1):167-179.
[21] O Ronneberger,P Fischer, T.Brox.U-Net:convolutional networks for biomedical imagesegmentation[C].MICCAI,2015:234-2.
[22] HanXiao.MR-based synthetic CT generation using a deep convolutional neural network method[J].Medical Physics,2017,44(4):1408-1419.
[23] Nie D,Cao X,Gao Y,et al.Estimating CT image from MRI data using 3D fully convolutional networks[M].Deep Learning and Data Labeling for Medical Applications.Springer,Cham,2016:170-178.
相似文献/References:
[1]何 妍,何 嘉.Non-local注意力机制生成对抗网络的油画修复研究[J].成都信息工程大学学报,2022,37(01):34.[doi:10.16836/j.cnki.jcuit.2022.01.006]
HE Yan,HE Jia.Research on Oil Painting Restoration of Non-local Attention Mechanism Generative Adversarial Network[J].Journal of Chengdu University of Information Technology,2022,37(03):34.[doi:10.16836/j.cnki.jcuit.2022.01.006]
备注/Memo
收稿日期:2020-03-28
基金项目:四川省科技计划资助项目(2019YFG0399、2019YFH0085)