Autoencoder-based conditional optimal transport generative adversarial network for medical image generation
Medical image generation has recently garnered significant interest among researchers. However, the primary generative models, such as Generative Adversarial Networks (GANs), often encounter challenges during training, including mode collapse. To address these issues, we proposed the AE-COT-GAN mode...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-03-01
|
| Series: | Visual Informatics |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2468502X23000529 |
| _version_ | 1827295923038846976 |
|---|---|
| author | Jun Wang Bohan Lei Liya Ding Xiaoyin Xu Xianfeng Gu Min Zhang |
| author_facet | Jun Wang Bohan Lei Liya Ding Xiaoyin Xu Xianfeng Gu Min Zhang |
| author_sort | Jun Wang |
| collection | DOAJ |
| description | Medical image generation has recently garnered significant interest among researchers. However, the primary generative models, such as Generative Adversarial Networks (GANs), often encounter challenges during training, including mode collapse. To address these issues, we proposed the AE-COT-GAN model (Autoencoder-based Conditional Optimal Transport Generative Adversarial Network) for the generation of medical images belonging to specific categories. The training process of our model comprises three fundamental components. The training process of our model encompasses three fundamental components. First, we employ an autoencoder model to obtain a low-dimensional manifold representation of real images. Second, we apply extended semi-discrete optimal transport to map Gaussian noise distribution to the latent space distribution and obtain corresponding labels effectively. This procedure leads to the generation of new latent codes with known labels. Finally, we integrate a GAN to train the decoder further to generate medical images. To evaluate the performance of the AE-COT-GAN model, we conducted experiments on two medical image datasets, namely DermaMNIST and BloodMNIST. The model’s performance was compared with state-of-the-art generative models. Results show that the AE-COT-GAN model had excellent performance in generating medical images. Moreover, it effectively addressed the common issues associated with traditional GANs. |
| first_indexed | 2024-04-24T14:28:24Z |
| format | Article |
| id | doaj.art-e3dbfe4ea80a409a8d56ed198930b192 |
| institution | Directory Open Access Journal |
| issn | 2468-502X |
| language | English |
| last_indexed | 2024-04-24T14:28:24Z |
| publishDate | 2024-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Visual Informatics |
| spelling | doaj.art-e3dbfe4ea80a409a8d56ed198930b1922024-04-03T04:27:09ZengElsevierVisual Informatics2468-502X2024-03-01811525Autoencoder-based conditional optimal transport generative adversarial network for medical image generationJun Wang0Bohan Lei1Liya Ding2Xiaoyin Xu3Xianfeng Gu4Min Zhang5School of Software Technology, Zhejiang University, Ningbo, Zhejiang, ChinaCollege of Computer Science & Technology, Zhejiang University, Hangzhou, Zhejiang, ChinaDepartment of Pathology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, Zhejiang, ChinaDepartment of Computer Science, State University of New York at Stony Brook, Stony Brook, NY, USACollege of Computer Science & Technology, Zhejiang University, Hangzhou, Zhejiang, China; Corresponding author.Medical image generation has recently garnered significant interest among researchers. However, the primary generative models, such as Generative Adversarial Networks (GANs), often encounter challenges during training, including mode collapse. To address these issues, we proposed the AE-COT-GAN model (Autoencoder-based Conditional Optimal Transport Generative Adversarial Network) for the generation of medical images belonging to specific categories. The training process of our model comprises three fundamental components. The training process of our model encompasses three fundamental components. First, we employ an autoencoder model to obtain a low-dimensional manifold representation of real images. Second, we apply extended semi-discrete optimal transport to map Gaussian noise distribution to the latent space distribution and obtain corresponding labels effectively. This procedure leads to the generation of new latent codes with known labels. Finally, we integrate a GAN to train the decoder further to generate medical images. To evaluate the performance of the AE-COT-GAN model, we conducted experiments on two medical image datasets, namely DermaMNIST and BloodMNIST. The model’s performance was compared with state-of-the-art generative models. Results show that the AE-COT-GAN model had excellent performance in generating medical images. Moreover, it effectively addressed the common issues associated with traditional GANs.http://www.sciencedirect.com/science/article/pii/S2468502X23000529Medical image generationMode collapseMode mixingOptimal transportGenerative adversarial networks |
| spellingShingle | Jun Wang Bohan Lei Liya Ding Xiaoyin Xu Xianfeng Gu Min Zhang Autoencoder-based conditional optimal transport generative adversarial network for medical image generation Visual Informatics Medical image generation Mode collapse Mode mixing Optimal transport Generative adversarial networks |
| title | Autoencoder-based conditional optimal transport generative adversarial network for medical image generation |
| title_full | Autoencoder-based conditional optimal transport generative adversarial network for medical image generation |
| title_fullStr | Autoencoder-based conditional optimal transport generative adversarial network for medical image generation |
| title_full_unstemmed | Autoencoder-based conditional optimal transport generative adversarial network for medical image generation |
| title_short | Autoencoder-based conditional optimal transport generative adversarial network for medical image generation |
| title_sort | autoencoder based conditional optimal transport generative adversarial network for medical image generation |
| topic | Medical image generation Mode collapse Mode mixing Optimal transport Generative adversarial networks |
| url | http://www.sciencedirect.com/science/article/pii/S2468502X23000529 |
| work_keys_str_mv | AT junwang autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration AT bohanlei autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration AT liyading autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration AT xiaoyinxu autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration AT xianfenggu autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration AT minzhang autoencoderbasedconditionaloptimaltransportgenerativeadversarialnetworkformedicalimagegeneration |