دورية أكاديمية
Training of a deep learning based digital subtraction angiography method using synthetic data.
| العنوان: | Training of a deep learning based digital subtraction angiography method using synthetic data. |
|---|---|
| المؤلفون: | Duan L; Division of X-Ray Imaging and Computed Tomography, German Cancer Research Center (DKFZ), Heidelberg, Germany.; School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences (UCAS), Beijing, China.; Key Laboratory of Optical Engineering, Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu, China., Eulig E; Division of X-Ray Imaging and Computed Tomography, German Cancer Research Center (DKFZ), Heidelberg, Germany.; Faculty of Physics and Astronomy, Heidelberg University, Heidelberg, Germany., Knaup M; Division of X-Ray Imaging and Computed Tomography, German Cancer Research Center (DKFZ), Heidelberg, Germany., Adamus R; Department of Radiology, Neuroradiology and Nuclear Medicine, Klinikum Nürnberg, Paracelsus Medical University, Nürnberg, Germany., Lell M; Department of Radiology, Neuroradiology and Nuclear Medicine, Klinikum Nürnberg, Paracelsus Medical University, Nürnberg, Germany., Kachelrieß M; Division of X-Ray Imaging and Computed Tomography, German Cancer Research Center (DKFZ), Heidelberg, Germany.; Medical Faculty Heidelberg, Heidelberg University, Heidelberg, Germany. |
| المصدر: | Medical physics [Med Phys] 2024 Jul; Vol. 51 (7), pp. 4793-4810. Date of Electronic Publication: 2024 Feb 14. |
| نوع المنشور: | Journal Article |
| اللغة: | English |
| بيانات الدورية: | Publisher: John Wiley and Sons, Inc Country of Publication: United States NLM ID: 0425746 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 2473-4209 (Electronic) Linking ISSN: 00942405 NLM ISO Abbreviation: Med Phys Subsets: MEDLINE |
| أسماء مطبوعة: | Publication: 2017- : Hoboken, NJ : John Wiley and Sons, Inc. Original Publication: Lancaster, Pa., Published for the American Assn. of Physicists in Medicine by the American Institute of Physics. |
| مواضيع طبية MeSH: | Angiography, Digital Subtraction*/methods , Deep Learning* , Image Processing, Computer-Assisted*/methods, Humans ; Tomography, X-Ray Computed |
| مستخلص: | Background: Digital subtraction angiography (DSA) is a fluoroscopy method primarily used for the diagnosis of cardiovascular diseases (CVDs). Deep learning-based DSA (DDSA) is developed to extract DSA-like images directly from fluoroscopic images, which helps in saving dose while improving image quality. It can also be applied where C-arm or patient motion is present and conventional DSA cannot be applied. However, due to the lack of clinical training data and unavoidable artifacts in DSA targets, current DDSA models still cannot satisfactorily display specific structures, nor can they predict noise-free images. Purpose: In this study, we propose a strategy for producing abundant synthetic DSA image pairs in which synthetic DSA targets are free of typical artifacts and noise commonly found in conventional DSA targets for DDSA model training. Methods: More than 7,000 forward-projected computed tomography (CT) images and more than 25,000 synthetic vascular projection images were employed to create contrast-enhanced fluoroscopic images and corresponding DSA images, which were utilized as DSA image pairs for training of the DDSA networks. The CT projection images and vascular projection images were generated from eight whole-body CT scans and 1,584 3D vascular skeletons, respectively. All vessel skeletons were generated with stochastic Lindenmayer systems. We trained DDSA models on this synthetic dataset and compared them to the trainings on a clinical DSA dataset, which contains nearly 4,000 fluoroscopic x-ray images obtained from different models of C-arms. Results: We evaluated DDSA models on clinical fluoroscopic data of different anatomies, including the leg, abdomen, and heart. The results on leg data showed for different methods that training on synthetic data performed similarly and sometimes outperformed training on clinical data. The results on abdomen and cardiac data demonstrated that models trained on synthetic data were able to extract clearer DSA-like images than conventional DSA and models trained on clinical data. The models trained on synthetic data consistently outperformed their clinical data counterparts, achieving higher scores in the quantitative evaluation of peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) metrics for DDSA images, as well as accuracy, precision, and Dice scores for segmentation of the DDSA images. Conclusions: We proposed an approach to train DDSA networks with synthetic DSA image pairs and extract DSA-like images from contrast-enhanced x-ray images directly. This is a potential tool to aid in diagnosis. (© 2024 The Authors. Medical Physics published by Wiley Periodicals LLC on behalf of American Association of Physicists in Medicine.) |
| References: | Chilcote WA, Modic MT, Pavlicek WA, et al. Digital subtraction angiography of the carotid arteries: a comparative study in 100 patients. Radiology. 1981;139:287‐295. Brody WR. Digital subtraction angiography. IEEE Trans Nucl Sci. 1982;29:1176‐1180. Hanafee W, Stout P. Subtraction technic. Radiology. 1962;79:658‐661. Kralj V, Brkić Biloš I. Morbidity and mortality from cardiovascular diseases. Cardiol Croat. 2013;8:373‐378. Zhang X, Zhang F, Li R. DSA image registration based on 3D space‐time detection. Procedia Eng. 2010;7:426‐431. Hipwell JH, Penney GP, McLaughlin RA, et al. Intensity‐based 2‐D‐3‐D registration of cerebral angiograms. IEEE Trans Med Imaging. 2003;22:1417‐1426. Sundarapandian M, Kalpathi R, Manason VD. DSA image registration using non‐uniform MRF model and pivotal control points. Comput Med Imaging Graph. 2013;37:323‐336. Ronneberger O, Fischer P, Brox T. U‐net: Convolutional networks for biomedical image segmentation. In: International Conference on Medical Image Computing and Computer‐assisted Intervention. Springer; 2015:234‐241. Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60‐88. Eulig E, Maier J, Knaup M, Koenig T, Hörndler K, Kachelrieß M. Deep DSA (DDSA): learning mask‐free digital subtraction angiography for static and dynamic acquisitions protocols using a deep convolutional neural network. In: ECR 2019 Book of Abstracts, Insights into Imaging 10 (Suppl. 1). 2019:S379. Gao Y, Song Y, Yin X, et al. Deep learning‐based digital subtraction angiography image generation. Int J Comput Assist Radiol Surg. 2019;14:1775‐1784. Eulig E, Maier J, Knaup M, Koenig T, Hörndler K, Kachelrieß M. Learned digital subtraction angiography (Deep DSA): method and application to lower extremities. In: 15th International Meeting on Fully Three‐Dimensional Image Reconstruction in Radiology and Nuclear Medicine. Vol 11072. SPIE; 2019:360‐363. Yonezawa H, Ueda D, Yamamoto A, et al. Maskless 2‐dimensional digital subtraction angiography generation model for abdominal vasculature using deep learning. J Vasc Interv Radiol. 2022;33:845‐851. Ueda D, Katayama Y, Yamamoto A, et al. Deep learning–based angiogram generation model for cerebral angiography without misregistration artifacts. Radiology. 2021;299:675‐681. Vepa A, Choi A, Nakhaei N, et al. Weakly‐supervised convolutional neural networks for vessel segmentation in cerebral angiography. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2022:585‐594. Wang Z, Bovik AC, Sheikh HR, Simoncelli EP. Image quality assessment: from error visibility to structural similarity. IEEE Trans Image Process. 2004;13:600‐612. Galarreta‐Valverde MA, Macedo MM, Mekkaoui C, Jackowski MP. Three‐dimensional synthetic blood vessel generation using stochastic L‐systems. In: Medical Imaging 2013: Image Processing. Vol 8669. SPIE; 2013:414‐419. Bresenham JE. Algorithm for computer control of a digital plotter. IBM Syst J. 1965;4:25‐30. Gui J, Sun Z, Wen Y, et al. A review on generative adversarial networks: Algorithms, theory, and applications. IEEE Trans Knowl Data Eng. 2021;35:3313‐3332. Arjovsky M, Chintala S, Bottou L. Wasserstein generative adversarial networks. In: International Conference on Machine Learning. PMLR; 2017:214‐223. Gulrajani I, Ahmed F, Arjovsky M, Dumoulin V, Courville AC. Improved training of Wasserstein Gans. Adv Neural Inf Process Syst. 2017;30:5769‐5779. Lamy J, Merveille O, Kerautret B, Passat N. A benchmark framework for multiregion analysis of vesselness filters. IEEE Trans Med Imaging. 2022;41:3649‐3662. Qin B, Mao H, Liu Y, et al. Robust PCA unrolling network for super‐resolution vessel extraction in x‐ray coronary angiography. IEEE Trans Med Imaging. 2022;41:3087‐3098. Huang L, Qin J, Zhou Y, et al. Normalization techniques in training dnns: Methodology, analysis and application. IEEE Trans Pattern Anal Mach Intell. 2023;45:10173‐10196. Kingma DP, Ba J. Adam: A Method for Stochastic Optimization. In: International Conference on Learning Representations, ICLR; 2015:1‐13. |
| معلومات مُعتمدة: | 202104910370 China Scholarship Council; Deutsches Krebsforschungszentrum; Helmholtz International Graduate School for Cancer Research |
| فهرسة مساهمة: | Keywords: deep learning; digital subtraction angiography; fluoroscopy; synthetic training data |
| تواريخ الأحداث: | Date Created: 20240214 Date Completed: 20240709 Latest Revision: 20240709 |
| رمز التحديث: | 20240709 |
| DOI: | 10.1002/mp.16973 |
| PMID: | 38353632 |
| قاعدة البيانات: | MEDLINE |
Full Text Finder كن أول من يترك تعليقا!