Expert-guided StyleGAN2 image generation elevates AI diagnostic accuracy for maxillary sinus lesions
The StyleGAN2-based image generation tool effectively addressed data scarcity and imbalance through high-quality MSL image synthesis, consequently boosting diagnostic model performance. This work not only facilitates AI-assisted preoperative assessment for maxillary sinus lift procedures but also establishes a methodological framework for overcoming data limitations in medical image analysis.
Authors
- Yixiong Lin
- Rihui Song (ORCID: https://orcid.org/0000-0001-9466-2408)
- Mengru Shi (ORCID: https://orcid.org/0000-0003-4100-3577)
- Zetao Chen
- K. X. Huang (ORCID: https://orcid.org/0000-0003-3347-204X)
- Haopeng Li
- Peisheng Zeng (ORCID: https://orcid.org/0000-0003-4542-9014)
- Yue Chen (ORCID: https://orcid.org/0000-0002-5619-6917)
- Xiaohang Li (ORCID: https://orcid.org/0000-0002-4434-365X)
- Zhuohong Gong (ORCID: https://orcid.org/0009-0001-3169-7651)
- Gengbin Cai
- Shijie Chen (ORCID: https://orcid.org/0009-0007-3127-6930)
Institutions
- Guangdong University of Technology (CN)
- Sun Yat-sen University (CN)
- Shunde Polytechnic (CN)
- Guangdong Polytechnic Normal University (CN)
- Stomatology Hospital (CN)
- Wuyi University (CN)
Publication Details
- Journal
- Communications Medicine
- Published
- 2025-05-20
- DOI
- https://doi.org/10.1038/s43856-025-00907-6
- Citations
- 5
- Primary Topic
- Medical Imaging and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 3.37
Funders
- National Natural Science Foundation of China
- Sun Yat-sen University
- Guangzhou Science and Technology Program key projects