Single-step prediction of inferior alveolar nerve injury after mandibular third-molar extraction using contrastive learning and Bayesian auto-tuned deep learning model
The proposed single-step deep learning approach effectively predicts IAN injury, with contrastive learning addressing data imbalance and Bayesian optimization enhancing model performance. While artificial intelligence surpasses human performance on CBCT images, panoramic radiograph analysis still benefits from expert interpretation. Future work should focus on multi-center validation and explainable artificial intelligence for broader clinical adoption.
Authors
- Jaeyeon Kim (ORCID: https://orcid.org/0000-0002-0432-6847)
- Jongeun Choi (ORCID: https://orcid.org/0000-0002-7532-5315)
- Wonse Park (ORCID: https://orcid.org/0000-0002-2081-1156)
- Yiseul Choi (ORCID: https://orcid.org/0000-0001-6283-9686)
- Myoungho Lee (ORCID: https://orcid.org/0000-0002-8214-2189)
- Kyubaek Yoon
- Jin‐Woo Kim (ORCID: https://orcid.org/0000-0002-1672-5730)
- Jun‐Young Kim (ORCID: https://orcid.org/0000-0002-6596-6135)
Institutions
- Ewha Womans University (KR)
- Yonsei University (KR)
- University Health System (US)
- Yonsei University Health System (KR)
- Artificial Intelligence in Medicine (Canada) (CA)
- Yonsei University College of Dentistry
Publication Details
- Journal
- Dentomaxillofacial Radiology
- Published
- 2025-09-27
- DOI
- https://doi.org/10.1093/dmfr/twaf069
- Primary Topic
- Dental Radiography and Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00