ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks
Multimodal medical imaging of human pathological tissues provides comprehensive information to assist in clinical diagnosis. However, due to the high cost of imaging, physiological incompatibility, and the harmfulness of radioactive tracers, multimodal medical image data remains scarce. Currently, c...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tsinghua University Press
2024-12-01
|
| Series: | CAAI Artificial Intelligence Research |
| Subjects: | |
| Online Access: | https://www.sciopen.com/article/10.26599/AIR.2024.9150035 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841550180074127360 |
|---|---|
| author | Liming Xu Yanrong Lei Bochuan Zheng Jiancheng Lv Weisheng Li |
| author_facet | Liming Xu Yanrong Lei Bochuan Zheng Jiancheng Lv Weisheng Li |
| author_sort | Liming Xu |
| collection | DOAJ |
| description | Multimodal medical imaging of human pathological tissues provides comprehensive information to assist in clinical diagnosis. However, due to the high cost of imaging, physiological incompatibility, and the harmfulness of radioactive tracers, multimodal medical image data remains scarce. Currently, cross-modal medical synthesis methods can generate desired modal images from existing modal images. However, most existing methods are limited to specific domains. This paper proposes an Adaptive Domain Medical Image Synthesis Method based on Generative Adversarial Networks (ADGAN) to address this issue. ADGAN achieves multidirectional medical image synthesis and ensures pathological consistency by constructing a single generator to learn the latent shared representation of multiple domains. The generator employs dense connections in shallow layers to preserve edge details and incorporates auxiliary information in deep layers to retain pathological features. Additionally, spectral normalization is introduced into the discriminator to control discriminative performance and indirectly enhance the image synthesis ability of the generator. Theoretically, it can be proved that the proposed method can be trained quickly, and spectral normalization contributes to adaptive and multidirectional synthesis. In practice, comparing with recent state-of-the-art methods, ADGAN achieves average increments of 4.7% SSIM, 6.7% MSIM, 7.3% PSNR, and 9.2% VIF. |
| format | Article |
| id | doaj-art-49d05580422941faa285248cff2e3bb6 |
| institution | Kabale University |
| issn | 2097-194X 2097-3691 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Tsinghua University Press |
| record_format | Article |
| series | CAAI Artificial Intelligence Research |
| spelling | doaj-art-49d05580422941faa285248cff2e3bb62025-01-10T06:44:32ZengTsinghua University PressCAAI Artificial Intelligence Research2097-194X2097-36912024-12-013915003510.26599/AIR.2024.9150035ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial NetworksLiming Xu0Yanrong Lei1Bochuan Zheng2Jiancheng Lv3Weisheng Li4School of Computer Science, China West Normal University, Nanchong 637002, ChinaSchool of Computer Science, China West Normal University, Nanchong 637002, ChinaSchool of Computer Science, China West Normal University, Nanchong 637002, ChinaCollege of Computer Science and Technology, Sichuan University, Chengdu 130012, ChinaCollege of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaMultimodal medical imaging of human pathological tissues provides comprehensive information to assist in clinical diagnosis. However, due to the high cost of imaging, physiological incompatibility, and the harmfulness of radioactive tracers, multimodal medical image data remains scarce. Currently, cross-modal medical synthesis methods can generate desired modal images from existing modal images. However, most existing methods are limited to specific domains. This paper proposes an Adaptive Domain Medical Image Synthesis Method based on Generative Adversarial Networks (ADGAN) to address this issue. ADGAN achieves multidirectional medical image synthesis and ensures pathological consistency by constructing a single generator to learn the latent shared representation of multiple domains. The generator employs dense connections in shallow layers to preserve edge details and incorporates auxiliary information in deep layers to retain pathological features. Additionally, spectral normalization is introduced into the discriminator to control discriminative performance and indirectly enhance the image synthesis ability of the generator. Theoretically, it can be proved that the proposed method can be trained quickly, and spectral normalization contributes to adaptive and multidirectional synthesis. In practice, comparing with recent state-of-the-art methods, ADGAN achieves average increments of 4.7% SSIM, 6.7% MSIM, 7.3% PSNR, and 9.2% VIF.https://www.sciopen.com/article/10.26599/AIR.2024.9150035generative adversarial networksmultimodal medical imagesadaptive domainlatent shared representation |
| spellingShingle | Liming Xu Yanrong Lei Bochuan Zheng Jiancheng Lv Weisheng Li ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks CAAI Artificial Intelligence Research generative adversarial networks multimodal medical images adaptive domain latent shared representation |
| title | ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks |
| title_full | ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks |
| title_fullStr | ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks |
| title_full_unstemmed | ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks |
| title_short | ADGAN: Adaptive Domain Medical Image Synthesis Based on Generative Adversarial Networks |
| title_sort | adgan adaptive domain medical image synthesis based on generative adversarial networks |
| topic | generative adversarial networks multimodal medical images adaptive domain latent shared representation |
| url | https://www.sciopen.com/article/10.26599/AIR.2024.9150035 |
| work_keys_str_mv | AT limingxu adganadaptivedomainmedicalimagesynthesisbasedongenerativeadversarialnetworks AT yanronglei adganadaptivedomainmedicalimagesynthesisbasedongenerativeadversarialnetworks AT bochuanzheng adganadaptivedomainmedicalimagesynthesisbasedongenerativeadversarialnetworks AT jianchenglv adganadaptivedomainmedicalimagesynthesisbasedongenerativeadversarialnetworks AT weishengli adganadaptivedomainmedicalimagesynthesisbasedongenerativeadversarialnetworks |