RetinaRegen: A Hybrid Model for Readability and Detail Restoration in Fundus Images
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916632745476096 |
|---|---|
| author | Tang, Yuhan Wang, Yudian Li, Weizhen Yue, Ye Pan, Chengchang Qi, Honggang |
| author_facet | Tang, Yuhan Wang, Yudian Li, Weizhen Yue, Ye Pan, Chengchang Qi, Honggang |
| contents | Fundus image quality is crucial for diagnosing eye diseases, but real-world conditions often result in blurred or unreadable images, increasing diagnostic uncertainty. To address these challenges, this study proposes RetinaRegen, a hybrid model for retinal image restoration that integrates a readability classifi-cation model, a Diffusion Model, and a Variational Autoencoder (VAE). Ex-periments on the SynFundus-1M dataset show that the proposed method achieves a PSNR of 27.4521, an SSIM of 0.9556, and an LPIPS of 0.1911 for the readability labels of the optic disc (RO) region. These results demonstrate superior performance in restoring key regions, offering an effective solution to enhance fundus image quality and support clinical diagnosis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_19153 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RetinaRegen: A Hybrid Model for Readability and Detail Restoration in Fundus Images Tang, Yuhan Wang, Yudian Li, Weizhen Yue, Ye Pan, Chengchang Qi, Honggang Image and Video Processing Computer Vision and Pattern Recognition Machine Learning Fundus image quality is crucial for diagnosing eye diseases, but real-world conditions often result in blurred or unreadable images, increasing diagnostic uncertainty. To address these challenges, this study proposes RetinaRegen, a hybrid model for retinal image restoration that integrates a readability classifi-cation model, a Diffusion Model, and a Variational Autoencoder (VAE). Ex-periments on the SynFundus-1M dataset show that the proposed method achieves a PSNR of 27.4521, an SSIM of 0.9556, and an LPIPS of 0.1911 for the readability labels of the optic disc (RO) region. These results demonstrate superior performance in restoring key regions, offering an effective solution to enhance fundus image quality and support clinical diagnosis. |
| title | RetinaRegen: A Hybrid Model for Readability and Detail Restoration in Fundus Images |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2502.19153 |