Accuracy of transfer learning in predicting peri-implantitis onset from intraoral radiograph: A multi-center longitudinal retrospective study.
6 affiliations
- 1Department of Periodontics and Oral Medicine, University of Michigan School of Dentistry, Ann Arbor, MI, USA. [email protected].
- 2Centre for Host Microbiome Interactions, Faculty of Dentistry, Oral and Craniofacial Sciences, King's College London, London, UK.
- 3Department of Periodontics and Oral Medicine, University of Michigan School of Dentistry, Ann Arbor, MI, USA.
- 4Department of Oral Medicine, Oral Pathology, Oral and Maxillofacial Surgery, University of Tripoli School of Dentistry, Tripoli, Libya.
- 5Department of Clinical and Experimental Medicine, University of Foggia, Foggia, 71122, Italy.
- 6Department of Medicine and Surgery, LUM University, Casamassima, 70010, Italy. [email protected].
Abstract
Background
The growing application of artificial intelligence in dentistry has opened new opportunities for predicting peri-implantitis (PI) using convolutional neural networks (CNNs). Transfer learning, where knowledge from a model pre-trained on a large dataset is adapted to a new, domain-specific task, may enhance performance, particularly when medical datasets are limited. This study evaluated the ability of several pre-trained CNN architectures applied to baseline 2D intraoral radiographs to predict peri-implantitis onset.Methods
A retrospective multi-center study was conducted at the University of Michigan and King's College London. A total of 195 implants (146 from Michigan and 49 from London) placed in 108 maintained patients were included. For each implant, a baseline periapical radiograph taken up to one year after prosthetic restoration was analyzed. Five pre-trained CNN architectures (ResNet50, DenseNet, VGG16, InceptionV3, EfficientNet) were combined with four transfer learning strategies: feature extraction (hybrid and non-hybrid), partial fine-tuning (50% of layers), and full fine-tuning. Models were trained using patient-wise splits to avoid leakage, and performance was evaluated on an independent test set using accuracy, area under the ROC curve (AUC), sensitivity, and specificity.Results
All CNN architectures showed variable performance across transfer learning strategies. The non-hybrid feature extraction approach provided the most consistent discrimination, with the best models achieving AUC values between 0.74 and 0.83. InceptionV3 reached the highest AUC (0.83). Fine-tuning strategies yielded modest or low performance, with AUCs generally ranging from 0.42 to 0.69 across architectures.Conclusion
Non-hybrid feature extraction applied to pre-trained CNNs showed the most promising, though moderate, performance for predicting PI onset from baseline intraoral radiographs. These findings indicate that transfer learning may assist early risk stratification in implant dentistry, but larger prospective studies are needed before clinical adoption.Clinical relevance
Deep learning models may serve as adjunctive tools to identify implants at higher risk for future PI based on routine baseline radiographs, supporting more individualized maintenance protocols and earlier preventive interventions.
Related
- CT-Referenced Diagnostic Performance and Complementary Error Patterns of Artificial Intelligence Versus a Single Emergency Physician in Chest Radiograph Interpretation: A Retrospective Paired Diagnostic Accuracy Study
- AI-Powered Preoperative Chest Radiograph-Derived Cardiovascular Border Indices and Risk of Postoperative Major Adverse Cardiovascular Events: A Large-Scale Retrospective Cohort Study.
- Predicting subsequent-year physical fitness assessment failure in university students using longitudinal data and explainable machine learning
- Multivariate time series dataset for predicting surface roughness and dimensional error in longitudinal turning.
- Interpretable machine learning for predicting intraoperative oxygenation impairment during one-lung ventilation in thoracic surgery: a retrospective cohort study
BibTeX
@article{saleh2026accuracy,
title = {{Accuracy of transfer learning in predicting peri-implantitis onset from intraoral radiograph: A multi-center longitudinal retrospective study.}},
author = {Muhammad H A Saleh and Luigi Nibali and Abdusalam E Alrmali and Vito Carlo Alberto Caponio and Pasquale Santamaria and Francesco Fanelli and Khrystyna Zhurakivska and Hom-Lay Wang and Giuseppe Troiano},
journal = {Clinical oral investigations},
year = {2026},
month = oct,
doi = {10.1007/s00784-026-07181-3},
url = {https://europepmc.org/article/MED/42894037},
note = {CC BY},
}
RIS
TY - JOUR
TI - Accuracy of transfer learning in predicting peri-implantitis onset from intraoral radiograph: A multi-center longitudinal retrospective study.
AU - Muhammad H A Saleh
AU - Luigi Nibali
AU - Abdusalam E Alrmali
AU - Vito Carlo Alberto Caponio
AU - Pasquale Santamaria
AU - Francesco Fanelli
AU - Khrystyna Zhurakivska
AU - Hom-Lay Wang
AU - Giuseppe Troiano
PY - 2026
DA - 2026/10/19
JO - Clinical oral investigations
DO - 10.1007/s00784-026-07181-3
UR - https://europepmc.org/article/MED/42894037
AB - Background The growing application of artificial intelligence in dentistry has opened new opportunities for predicting peri-implantitis (PI) using convolutional neural networks (CNNs). Transfer learning, where knowledge from a model pre-trained on a large dataset is adapted to a new, domain-specific task, may enhance performance, particularly when medical datasets are limited. This study evaluated the ability of several pre-trained CNN architectures applied to baseline 2D intraoral radiographs to predict peri-implantitis onset.Methods A retrospective multi-center study was conducted at the University of Michigan and King's College London. A total of 195 implants (146 from Michigan and 49 from London) placed in 108 maintained patients were included. For each implant, a baseline periapical radiograph taken up to one year after prosthetic restoration was analyzed. Five pre-trained CNN architectures (ResNet50, DenseNet, VGG16, InceptionV3, EfficientNet) were combined with four transfer learning strategies: feature extraction (hybrid and non-hybrid), partial fine-tuning (50% of layers), and full fine-tuning. Models were trained using patient-wise splits to avoid leakage, and performance was evaluated on an independent test set using accuracy, area under the ROC curve (AUC), sensitivity, and specificity.Results All CNN architectures showed variable performance across transfer learning strategies. The non-hybrid feature extraction approach provided the most consistent discrimination, with the best models achieving AUC values between 0.74 and 0.83. InceptionV3 reached the highest AUC (0.83). Fine-tuning strategies yielded modest or low performance, with AUCs generally ranging from 0.42 to 0.69 across architectures.Conclusion Non-hybrid feature extraction applied to pre-trained CNNs showed the most promising, though moderate, performance for predicting PI onset from baseline intraoral radiographs. These findings indicate that transfer learning may assist early risk stratification in implant dentistry, but larger prospective studies are needed before clinical adoption.Clinical relevance Deep learning models may serve as adjunctive tools to identify implants at higher risk for future PI based on routine baseline radiographs, supporting more individualized maintenance protocols and earlier preventive interventions.
SN - 1432-6981
SN - 1436-3771
ER -