OralHybridNet: A Deep Learning Framework for Multi-Label Classification of Dental Restorations and Prostheses in Panoramic Radiographs.
Automated prediction of dental conditions in Orthopantomogram (OPG) panoramic radiographs faces significant challenges due to class imbalance, rare pathologies, and complex anatomical structures. This study proposes OralHybridNet, a novel hybrid deep learning framework integrating hierarchical convo...
| Publicado en: | Inquiry (00469580) Vol. 63; pp. 1 - 19 |
|---|---|
| Autores principales: | , , , , , |
| Formato: | Artículo |
| Publicado: |
Sage Publications Inc.
4/13/2026
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=192982122&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 192982122 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 00469580 INQ jtl: Inquiry (00469580) issn: 00469580 maglogo: Y pubinfo: dt: 4/13/2026 vid: 63 pid: 344 pub: Sage Publications Inc. artinfo: ui: 192982122 10.1177/00469580261439986 ppf: 1 ppct: 18 formats: tig: atl: OralHybridNet: A Deep Learning Framework for Multi-Label Classification of Dental Restorations and Prostheses in Panoramic Radiographs. aug: au: Khurshid, Zohaib Moustafa Ali, Ramy Moustafa Alharbi, Ali Sulaiman Osathanon, Thanaphum Noom Alfaraj, Amal Porntaveetus, Thantrira affil: Department of Prosthodontics and Dental Implantology, College of Dentistry, King Faisal University, Al-Ahsa, Saudi Arabia Department of Anatomy, Faculty of Dentistry, Center of Artificial Intelligence and Innovation (CAII), Center of Excellence for Dental Stem Cell Biology, Chulalongkorn University, Bangkok, Thailand Center of Excellence in Precision Medicine and Digital Health, Department of Physiology, Faculty of Dentistry, Chulalongkorn University, Bangkok, Thailand University of Zurich, Switzerland su: Dental fillings Dental implants Statistical power analysis Bridges (Dentistry) Research funding Data analysis Receiver operating characteristic curves Dentures Research evaluation Dental crowns Retrospective studies Descriptive statistics Deep learning Panoramic radiography Root canal treatment Medical records Acquisition of data Research Conceptual structures Artificial neural networks Dental caries Confidence intervals Digital image processing Machine learning Algorithms Sensitivity & specificity (Statistics) Thailand sug: subj: Thailand Dental fillings Dental implants Statistical power analysis Bridges (Dentistry) Research funding Data analysis Receiver operating characteristic curves Dentures Research evaluation Dental crowns Retrospective studies Descriptive statistics Deep learning Panoramic radiography Root canal treatment Medical records Acquisition of data Research Conceptual structures Artificial neural networks Dental caries Confidence intervals Digital image processing Machine learning Algorithms Sensitivity & specificity (Statistics) keyword: artificial intelligence deep learning dental radiology feature fusion multi-label classification panoramic radiography ab: Automated prediction of dental conditions in Orthopantomogram (OPG) panoramic radiographs faces significant challenges due to class imbalance, rare pathologies, and complex anatomical structures. This study proposes OralHybridNet, a novel hybrid deep learning framework integrating hierarchical convolutional neural networks, such as CustomDentalNet integrate dual-attention mechanisms and OralNetXPlus. A multinational dataset comprising 2047 clinician-annotated panoramic radiographs spanning 7 diagnostic labels was used. An adaptive augmentation protocol combining Elastic Transformations and gamma correction mitigated class imbalance. A Hybrid Feature Selection (HFS) algorithm condensed 1208-dimensional embeddings into a discriminative 300-feature subset. Evaluated against ResNet50 baselines, OralHybridNet achieved 96.0% accuracy, 97.6% precision, and 0.993 AUC-ROC. The KNN Fine classifier on fused features yielded the highest performance, with real-time capability (9 ms inference time) using a neural network classifier. The proposed framework demonstrates a promising proof-of-concept framework for automated multi-label dental restoration classification. pubtype: Academic Journal doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2026 holdings: @attributes: islocal: N |
|---|