EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

A Novel Multimodal Deep Image Analysis Model for Predicting Extraction/Non‐Extraction Decision

Sunna Imtiaz Ahmad, Jakub Olczyk, Adriel S Araújo, João Pedro de Moura Medeiros, Vinicius C Teixeira, Carlos F A Gomes, Maurício Cecílio Magnaguagno, Quinn Roederer, Vinicius Dutra iD, R Scott Conley iD, Dalvan Griebler, George Eckert iD and 2 more

Orthodontics & Craniofacial Research · 2026-02 · 29 (1) · 101-111

DOI: 10.1111/ocr.70057 ↗

At a glance

Journal article in Orthodontics & Craniofacial Research, published online on 2025-11-06. Written by Sunna Imtiaz Ahmad and 13 co-authors. Institution: School of Dentistry, Indiana University, Indianapolis, Indiana, USA. Major MEDLINE topics: Cephalometry, Deep Learning, Tooth Extraction, Image Processing, Computer-Assisted. Cited 5 times so far (Crossref); 40 references. The full text is freely available (open access, CC BY 4.0). Funded by: National Institute of Science and Technology in Medicine Assisted by Scientific Computing, Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)-Brasil, CNPq Productivity Scholarships.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Objective

This study aimed to develop a deep learning model classifier capable of predicting the extraction/non‐extraction binary decision using lateral cephalometric radiographs (LCRs) and intraoral scans (IOS) to serve as an additional decision‐support tool for orthodontists. Materials and Methods The dataset was composed of LCRs and IOS from 617 patients (mean age: 18.2, 63.5% female) treated at the Indiana University School of Dentistry. Subjects were categorised into two groups: extraction (192) and non‐extraction (425). Two sets of features were extracted from IOS: traditional arch measurements and novel tooth spatial features. For LCRs, features were derived using CephNet‐based landmark detection (Land), a convolutional autoencoder (AE), and the dimensionality was reduced using Principal Component Analysis (PCA). Models were evaluated using accuracy, sensitivity, specificity, positive predictive value (PPV or precision), negative predictive value (NPV), positive likelihood ratio (LR + ), negative likelihood ratio (LR − ), and F1 score.

Results

IOS + Land model achieved the highest overall accuracy (77%) and F1 score (0.62), with strong specificity (83%) and PPV (62%). In contrast, the Land model yielded the highest sensitivity (82%), but at the cost of lower specificity (57%). McNemar's test revealed that the AE model was significantly less accurate than IOS + AE ( p = 0.048), IOS + Land ( p = 0.006), and IOS + AE + Land ( p = 0.005).

Conclusion

Deep learning models can predict the extraction/non‐extraction decision using IOS and LCRs with high accuracy and diagnostic performance. Multimodal approaches, particularly those integrating IOS with cephalometric landmarks, demonstrate superior accuracy, sensitivity, and specificity compared to single‐modality models.

Subjects

MeSH terms (MEDLINE)

CephalometryDeep LearningTooth ExtractionImage Processing, Computer-AssistedHumansFemaleAdolescentMaleSensitivity and SpecificityYoung AdultAnatomic LandmarksPredictive Value of TestsPrincipal Component Analysis

Author keywords

artificial intelligenceclinical decision‐makingdeep learningorthodonticstooth extraction

Authors and institutions

  • School of Dentistry, Indiana University, Indianapolis, Indiana, USA.
  • School of Technology, Pontifical Catholic University of Rio Grande do Sul, Porto Alegre, Brazil.

ORCID profiles: Vinicius Dutra, R Scott Conley, George Eckert, Hakan Turkkahraman

Funding

  • National Institute of Science and Technology in Medicine Assisted by Scientific Computing
  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)-Brasil
  • CNPq Productivity Scholarships
  • FABERGS RITE CIARS Inteligência Artificial Aplicada à Saúde
  • Instituto Nacional de Ciência e Tecnologia em Medicina Assistida por Computação Científica
  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Cite

APA

Ahmad, S. I., Olczyk, J., Araújo, A. S., de Moura Medeiros, J. P., Teixeira, V. C., Gomes, C. F. A., Magnaguagno, M. C., Roederer, Q., Dutra, V., Conley, R. S., Griebler, D., Eckert, G., Pinho, M. S., & Turkkahraman, H. (2026). A Novel Multimodal Deep Image Analysis Model for Predicting Extraction/Non‐Extraction Decision. Orthodontics & Craniofacial Research, 29(1), 101-111. https://doi.org/10.1111/ocr.70057

Vancouver

Ahmad SI, Olczyk J, Araújo AS, de Moura Medeiros JP, Teixeira VC, Gomes CFA, et al. A Novel Multimodal Deep Image Analysis Model for Predicting Extraction/Non‐Extraction Decision. Orthodontics & Craniofacial Research. 2026;29(1):101-111. doi:10.1111/ocr.70057

BibTeX
@article{Ahmad2026,
  author = {Ahmad, Sunna Imtiaz and Olczyk, Jakub and Araújo, Adriel S and de Moura Medeiros, João Pedro and Teixeira, Vinicius C and Gomes, Carlos F A and Magnaguagno, Maurício Cecílio and Roederer, Quinn and Dutra, Vinicius and Conley, R Scott and Griebler, Dalvan and Eckert, George and Pinho, Márcio Sarroglia and Turkkahraman, Hakan},
  title = {A Novel Multimodal Deep Image Analysis Model for Predicting Extraction/Non‐Extraction Decision},
  journal = {Orthodontics & Craniofacial Research},
  year = {2026},
  volume = {29},
  number = {1},
  pages = {101--111},
  doi = {10.1111/ocr.70057}
}
Start for free →