TECHNOLOGY AND CONTRAST MEDIA / ORIGINAL PAPER
AI-enhanced PET/CT image synthesis using CycleGAN for improved ovarian cancer imaging
1
Namazi Hospital, Shiraz University of Medical Sciences, Shiraz Iran
2
Radiation Research Center, School of Mechanical Engineering, Shiraz University, Shiraz, Iran
3
Nuclear Engineering Department, School of Mechanical Engineering, Shiraz University, Shiraz, Iran
4
Ionizing and Non-Ionizing Radiation protection research center, School of Paramedical Sciences, Shiraz University of Medical Sciences, Shiraz, Iran
5
PET/CT Center, Nuclear Medicine Department, Kowsar Hospital, Shiraz, Iran
Submission date: 2024-08-23
Final revision date: 2024-11-07
Acceptance date: 2024-12-03
Publication date: 2025-01-17
Corresponding author
Sedigheh Sina
Radiation Research Center, School of Mechanical Engineering, Shiraz University, Mollasadra Street, Shiraz, Iran
Radiation Research Center, School of Mechanical Engineering, Shiraz University, Mollasadra Street, Shiraz, Iran
Pol J Radiol, 2025; 90: 26-35
KEYWORDS
TOPICS
ABSTRACT
Purpose:
Ovarian cancer is the fifth fatal cancer among women. Positron emission tomography (PET), which offers detailed metabolic data, can be effectively used for early cancer screening. However, proper attenuation correction is essential for interpreting the data obtained by this imaging modality. Computed tomography (CT) imaging is commonly performed alongside PET imaging for attenuation correction. This approach may introduce some issues in spatial alignment and registration of the images obtained by the two modalities. This study aims to perform PET image attenuation correction by using generative adversarial networks (GANs), without additional CT imaging.
Material and methods:
The PET/CT data from 55 ovarian cancer patients were used in this study. Three GAN architectures: Conditional GAN, Wasserstein GAN, and CycleGAN, were evaluated for attenuation correction. The statistical performance of each model was assessed by calculating the mean squared error (MSE) and mean absolute error (MAE). The radiological performance assessments of the models were performed by comparing the standardised uptake value and the Hounsfield unit values of the whole body and selected organs, in the synthetic and real PET and CT images.
Results:
Based on the results, CycleGAN demonstrated effective attenuation correction and pseudo-CT generation, with high accuracy. The MAE and MSE for all images were 2.15 ± 0.34 and 3.14 ± 0.56, respectively. For CT reconstruction, such values were found to be 4.17 ± 0.96 and 5.66 ± 1.01, respectively.
Conclusions:
The results showed the potential of deep learning in reducing radiation exposure and improving the quality of PET imaging. Further refinement and clinical validation are needed for full clinical applicability.
Ovarian cancer is the fifth fatal cancer among women. Positron emission tomography (PET), which offers detailed metabolic data, can be effectively used for early cancer screening. However, proper attenuation correction is essential for interpreting the data obtained by this imaging modality. Computed tomography (CT) imaging is commonly performed alongside PET imaging for attenuation correction. This approach may introduce some issues in spatial alignment and registration of the images obtained by the two modalities. This study aims to perform PET image attenuation correction by using generative adversarial networks (GANs), without additional CT imaging.
Material and methods:
The PET/CT data from 55 ovarian cancer patients were used in this study. Three GAN architectures: Conditional GAN, Wasserstein GAN, and CycleGAN, were evaluated for attenuation correction. The statistical performance of each model was assessed by calculating the mean squared error (MSE) and mean absolute error (MAE). The radiological performance assessments of the models were performed by comparing the standardised uptake value and the Hounsfield unit values of the whole body and selected organs, in the synthetic and real PET and CT images.
Results:
Based on the results, CycleGAN demonstrated effective attenuation correction and pseudo-CT generation, with high accuracy. The MAE and MSE for all images were 2.15 ± 0.34 and 3.14 ± 0.56, respectively. For CT reconstruction, such values were found to be 4.17 ± 0.96 and 5.66 ± 1.01, respectively.
Conclusions:
The results showed the potential of deep learning in reducing radiation exposure and improving the quality of PET imaging. Further refinement and clinical validation are needed for full clinical applicability.
REFERENCES (34)
2.
Breen J, Allen K, Zucker K, Adusumilli P, Scarsbrook A, Hall G, et al. Artificial intelligence in ovarian cancer histopathology: a systematic review. NPJ Precis Oncol 2023; 7: 83. DOI: 10.1038/s41698-023-00432-6.
3.
Pianykh OS, Langs G, Dewey M, Enzmann DR, Herold CJ, Schoenberg SO, Brink JA. Continuous learning AI in radiology: implementation principles and early applications. Radiology 2020; 297: 6-14.
4.
Basu S, Hess S, Braad PEN, Olsen BB, Inglev S, Høilund-Carlsen PF. The basic principles of FDG-PET/CT imaging. PET Clin 2014; 9: 355-370.
5.
Gillen R, Firbank MJ, Lloyd J, T O’Brien J. CT-based attenuation and scatter correction compared with uniform attenuation correction in brain perfusion SPECT imaging for dementia. Phys Med Biol 2015; 60: 6775-6787.
6.
Frane N, Bitterman A. Radiation safety and protection. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2025.
7.
Son H, Khan SM, Rahaman J, Cameron KL, Prasad-Hayes M, Chuang L, et al. Role of FDG PET/CT in staging of recurrent ovarian cancer. Radiographics 2011; 31: 569-583.
8.
Khiewvan B, Torigian DA, Emamzadehfard S, Paydary K, Salavati A, Houshmand S, et al. An update on the role of PET/CT and PET/MRI in ovarian cancer. Eur J Nucl Med Mol Imaging 2017; 44: 1079-1091.
9.
Engbersen MP, Van Driel W, Lambregts D, Lahaye M. The role of CT, PET-CT, and MRI in ovarian cancer. Br J Radiol 2021; 94: 20210117. DOI: doi: 10.1259/bjr.20210117.
10.
Sadeghi MH, Sina S, Alavi M, Feshani ZN, Farshchitabrizi AH. 323: A comparison of 3D CNN architectures for monitoring ovarian cancer patients using PET/CT scans. Radiother Oncol 2024; 194: S4966-S4967.
11.
Sadeghi MH, Sina S, Alavi M, Feshani ZN, Farshchitabrizi AH. PS01. 01 3D multiclass semantic segmentation of PET/CT images for ovarian cancer using 3D U-Net. Phys Med 2024; 125: 104003.
12.
Sadeghi MH, Omidi H, Sina S. A systematic review on the use of artificial intelligence techniques in the diagnosis of COVID-19 from chest X-ray images. Avicenna J Med Biochem 2020; 8: 120-127.
13.
Xu HL, Gong TT, Liu FH, Chen HY, Xiao Q, Hou Y, et al. Artificial intelligence performance in image-based ovarian cancer identification: a systematic review and meta-analysis. EClinicalMedicine 2022; 53: 101662. DOI: 10.1016/j.eclinm.2022.101662.
14.
McBee MP, Awan OA, Colucci AT, Ghobadi CW, Kadom N, Kansagra AP, et al. Deep learning in radiology. Acad Radiol 2018; 25: 1472-1480.
15.
Matsubara K, Ibaraki M, Nemoto M, Watabe H, Kimura Y. A review on AI in PET imaging. Ann Nucl Med 2022; 36: 133-143.
16.
Sadaghiani MS, Rowe SP, Sheikhbahaei S. Applications of artificial intelligence in oncologic 18F-FDG PET/CT imaging: a systematic review. Ann Transl Med 2021; 9: 823. DOI: 10.21037/atm-20-6162.
17.
Lee JS. A review of deep-learning-based approaches for attenuation correction in positron emission tomography. IEEE Trans Radiat Plasma Med Sci 2020; 5: 160-184.
18.
Yang J, Sohn JH, Behr SC, Gullberg GT, Seo Y. CT-less direct correction of attenuation and scatter in the image space using deep learning for whole-body FDG PET: potential benefits and pitfalls. Radiol Artif Intell 2020; 3: e200137. DOI: 10.1148/ryai.2020200137.
19.
Liu F, Jang H, Kijowski R, Zhao G, Bradshaw T, McMillan AB. A deep learning approach for 18 F-FDG PET attenuation correction. EJNMMI Phys 2018; 5: 24. DOI: 10.1186/s40658-018-0225-8.
20.
Rao F, Yang B, Chen YW, Li J, Wang H, Ye H, et al. A novel supervised learning method to generate CT images for attenuation correction in delayed pet scans. Comput Methods Programs Biomed 2020; 197: 105764. DOI: 10.1016/j.cmpb.2020.105764.
21.
Sorin V, Barash Y, Konen E, Klang E. Creating artificial images for radiology applications using generative adversarial networks (GANs) – a systematic review. Acad Radiol 2020; 27: 1175-1185.
22.
Poeppel TD, Krause BJ, Heusner TA, Boy C, Bockisch A, Antoch G. PET/CT for the staging and follow-up of patients with malignancies. Eur J Radiol 2009; 70: 382-392.
23.
Fukui R, Fujii S, Ninomiya H, Fujiwara Y, Ida T. Generation of the pseudo CT image based on the deep learning technique aimed for the attenuation correction of the PET image. Nihon Hoshasen Gijutsu Gakkai Zasshi 2020; 76: 1152-1162.
24.
Hu Z, Li Y, Zou S, Xue H, Sang Z, Liu X, et al. Obtaining PET/CT images from non-attenuation corrected PET images in a single PET system using Wasserstein generative adversarial networks. Phys Med Biol 2020; 65: 215010. DOI: 10.1088/1361-6560/aba5e9.
25.
Yang X, Lei Y, Dong X, Wang T, Higgins K, Liu T, et al. Attenuation and scatter correction for whole-body PET using 3D generative adversarial networks. J Nucl Med 2019; 60 (Suppl 1): 174.
26.
Cho J, Yoon K. Conditional activation GAN: improved auxiliary classifier GAN. IEEE Access 2020; 8: 216729-216740. DOI: 10.1109/ACCESS.2020.3041480.
27.
Zhang C, Feng Y, Qiang B, Shang J. Wasserstein generative recurrent adversarial networks for image generating. 24th International Conference on Pattern Recognition (ICPR), Beijing, China, 2018, pp. 242-247. DOI: 10.1109/ICPR.2018.8546264.
28.
Peng X, Li Q, Wu T, Yuan S. Cross-GAN: Unsupervised image-to-image translation. 2022 IEEE 6th Information Technology and Mechatronics Engineering Conference (ITOEC), Chongqing, China, 2022, pp. 1755-1759. DOI: 10.1109/ITOEC53115.2022.9734431.
29.
Rahnenführer J, De Bin R, Benner A, Ambrogi F, Lusa L, Boulesteix AL, et al. Statistical analysis of high-dimensional biomedical data: a gentle introduction to analytical goals, common approaches and challenges. BMC Med 2023; 21: 182. DOI: 10.1186/s12916-023-02858-y.
30.
Sadeghi MH, Sina S, Omidi H, Farshchitabrizi AH, Alavi M. Deep learning in ovarian cancer diagnosis: a comprehensive review of various imaging modalities. Polish J Radiol 2024; 89: e30-e48. DOI: 10.5114/pjr.2024.134817.
31.
Wang Y, Wang M, Cao P, Wong EMF, Ho G, Lam TPW, et al. CT-based deep learning segmentation of ovarian cancer and the stability of the extracted radiomics features. Quant Imaging Med Surg 2023; 13: 5218-5229.
32.
Zarei M, Wallsten E, Grefve J, Söderkvist K, Gunnlaugsson A, Sandgren K, et al. Accuracy of gross tumour volume delineation with [68Ga]-PSMA-PET compared to histopathology for high-risk prostate cancer. Acta Oncol 2024; 63: 503-510.
33.
Sadeghi MH, Sina S, Alavi M, Giammarile F. The OCDA-Net: a 3D convolutional neural network-based system for classification and staging of ovarian cancer patients using [18F]FDG PET/CT examinations. Ann Nucl Med 2023; 37: 645-654.
34.
Sadeghi MH, Sina S, Alavi M, Giammarile F, Yeong CH. PET/CT-based 3D multi-class semantic segmentation of ovarian cancer and the stability of the extracted radiomics features. Phys Eng Sci Med 2024; 47: 1739-1749.
We process personal data collected when visiting the website. The function of obtaining information about users and their behavior is carried out by voluntarily entered information in forms and saving cookies in end devices. Data, including cookies, are used to provide services, improve the user experience and to analyze the traffic in accordance with the Privacy policy. Data are also collected and processed by Google Analytics tool (more).
You can change cookies settings in your browser. Restricted use of cookies in the browser configuration may affect some functionalities of the website.
You can change cookies settings in your browser. Restricted use of cookies in the browser configuration may affect some functionalities of the website.
