Comparison of deep learning models to detect crossbites on 2D intraoral photographs.
Artificial intelligence
Crossbite
Deep learning
Neural networks
Orthodontic diagnosis
Journal
Head & face medicine
ISSN: 1746-160X
Titre abrégé: Head Face Med
Pays: England
ID NLM: 101245792
Informations de publication
Date de publication:
02 Sep 2024
02 Sep 2024
Historique:
received:
30
05
2024
accepted:
26
08
2024
medline:
3
9
2024
pubmed:
3
9
2024
entrez:
2
9
2024
Statut:
epublish
Résumé
To support dentists with limited experience, this study trained and compared six convolutional neural networks to detect crossbites and classify non-crossbite, frontal, and lateral crossbites using 2D intraoral photographs. Based on 676 photographs from 311 orthodontic patients, six convolutional neural network models were trained and compared to classify (1) non-crossbite vs. crossbite and (2) non-crossbite vs. lateral crossbite vs. frontal crossbite. The trained models comprised DenseNet, EfficientNet, MobileNet, ResNet18, ResNet50, and Xception. Among the models, Xception showed the highest accuracy (98.57%) in the test dataset for classifying non-crossbite vs. crossbite images. When additionally distinguishing between lateral and frontal crossbites, average accuracy decreased with the DenseNet architecture achieving the highest accuracy among the models with 91.43% in the test dataset. Convolutional neural networks show high potential in processing clinical photographs and detecting crossbites. This study provides initial insights into how deep learning models can be used for orthodontic diagnosis of malocclusions based on intraoral 2D photographs.
Sections du résumé
BACKGROUND
BACKGROUND
To support dentists with limited experience, this study trained and compared six convolutional neural networks to detect crossbites and classify non-crossbite, frontal, and lateral crossbites using 2D intraoral photographs.
METHODS
METHODS
Based on 676 photographs from 311 orthodontic patients, six convolutional neural network models were trained and compared to classify (1) non-crossbite vs. crossbite and (2) non-crossbite vs. lateral crossbite vs. frontal crossbite. The trained models comprised DenseNet, EfficientNet, MobileNet, ResNet18, ResNet50, and Xception.
FINDINGS
RESULTS
Among the models, Xception showed the highest accuracy (98.57%) in the test dataset for classifying non-crossbite vs. crossbite images. When additionally distinguishing between lateral and frontal crossbites, average accuracy decreased with the DenseNet architecture achieving the highest accuracy among the models with 91.43% in the test dataset.
CONCLUSIONS
CONCLUSIONS
Convolutional neural networks show high potential in processing clinical photographs and detecting crossbites. This study provides initial insights into how deep learning models can be used for orthodontic diagnosis of malocclusions based on intraoral 2D photographs.
Identifiants
pubmed: 39223562
doi: 10.1186/s13005-024-00448-8
pii: 10.1186/s13005-024-00448-8
doi:
Types de publication
Journal Article
Comparative Study
Langues
eng
Sous-ensembles de citation
IM
Pagination
45Informations de copyright
© 2024. The Author(s).
Références
Kunz F, Stellzig-Eisenhauer A, Boldt J. Applications of Artificial Intelligence in Orthodontics—An overview and perspective based on the current state of the art. Appl Sci. 2023;13(6).
Di Blasio M, Vaienti B, Pedrazzi G, Cassi D, Magnifico M, Meneghello S, Di Blasio A. Are the reasons why patients are referred for an Orthodontic visit correct? Int J Environ Res Public Health. 2021;18(10).
O’Brien K, Wright J, Conboy F, Bagley L, Lewis D, Read M, et al. The effect of orthodontic referral guidelines: a randomised controlled trial. Br Dent J. 2000;188(7):392–7.
doi: 10.1038/sj.bdj.4800492a
pubmed: 10816930
Furhmann RAW. Genehmigungsfähigkeit bei unklarem KIG-Befund Obergutachten zur Überprüfung von GKV-Gutachten. 2019.
Gesch D, Kirbschus A, Schröder W, Bernhardt O, Proff P, Bayerlein T, et al. Influence of examiner differences on KIG-classification when assessing malocclusions. J Orofac Orthop = Fortschr Der Kieferorthopadie: Organ/official J Dtsch Gesellschaft fur Kieferorthop. 2006;67(2):81–91.
doi: 10.1007/s00056-006-5037-0
Stolze A, Goldbecher H. Der optimale Behandlungsbeginn | Quintessenz Verlags-GmbH. 2012. pp. 271 – 84.
Reddy S, Derringer KA, Rennie L. Orthodontic referrals: why do GDPs get it wrong? Br Dent J. 2016;221(9):583–7.
doi: 10.1038/sj.bdj.2016.826
pubmed: 27811869
Weide E. Die Macht der künstlichen Intelligenz. Ein gelungener Einstieg ins nächste Jahrtausend. Ullstein Taschenbuchvlg; 1993.
Auconi P, Gili T, Capuani S, Saccucci M, Caldarelli G, Polimeni A, Di Carlo G. The validity of machine learning procedures in Orthodontics: what is still missing? J Pers Med. 2022;12(6).
Ryu J, Kim YH, Kim TW, Jung SK. Evaluation of artificial intelligence model for crowding categorization and extraction diagnosis using intraoral photographs. Sci Rep. 2023;13(1):5177.
doi: 10.1038/s41598-023-32514-7
pubmed: 36997621
pmcid: 10063582
Schwendicke F, Chaurasia A, Arsiwala L, Lee JH, Elhennawy K, Jost-Brinkmann PG, et al. Deep learning for cephalometric landmark detection: systematic review and meta-analysis. Clin Oral Investig. 2021;25(7):4299–309.
doi: 10.1007/s00784-021-03990-w
pubmed: 34046742
pmcid: 8310492
Taraji S, Atici SF, Viana G, Kusnoto B, Allareddy VS, Miloro M, Elnagar MH. Novel machine learning algorithms for prediction of treatment decisions in adult patients with Class III Malocclusion. J Oral Maxillofac Surg. 2023.
Lee K-S, Ryu J-J, Jang HS, Lee D-Y, Jung S-K. Deep Convolutional Neural Networks Based Analysis of Cephalometric Radiographs for Differential diagnosis of orthognathic surgery indications. Appl Sci. 2020;10(6).
He K, Zhang X, Ren S, Sun J, editors. Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Recognition P. (CVPR); 2016 27–30 June 2016.
Howard A, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T et al. MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. 2017.
Chollet F, Xception. Deep Learning with Depthwise Separable Convolutions2017. 1800-7 p.
Huang G, Liu Z, Van Der Maaten L, Weinberger KQ, editors. Densely connected convolutional networks. Proceedings of the IEEE conference on computer vision and pattern recognition; 2017.
Tan M, Le Q, editors. Efficientnet: Rethinking model scaling for convolutional neural networks. International conference on machine learning; 2019: PMLR.
Deng J, Dong W, Socher R, Li LJ, Kai L, Li F-F, editors. ImageNet: A large-scale hierarchical image database. 2009 IEEE Conference on Computer Vision and Pattern Recognition; 2009 20–25 June 2009.
Lee K-S, Ryu J-J, Jang HS, Lee D-Y, Jung S-K. Deep Convolutional Neural Networks Based Analysis of Cephalometric Radiographs for Differential diagnosis of orthognathic surgery indications. Appl Sci [Internet]. 2020; 10(6).
Lee C, Jeon KJ, Han SS, Kim YH, Choi YJ, Lee A, Choi JH. Ct-like MRi using the zero-te technique for osseous changes of the tMJ. Dentomaxillofacial Radiol. 2020;49(3).
Choi H-I, Jung S-K, Baek S-H, Lim WH, Ahn S-J, Yang I-H, Kim T-W. Artificial Intelligent Model with neural network machine learning for the diagnosis of orthognathic surgery. J Craniofac Surg. 2019;30(7).
Shin W, Yeom H-G, Lee GH, Yun JP, Jeong SH, Lee JH, et al. Deep learning based prediction of necessity for orthognathic surgery of skeletal malocclusion using cephalogram in Korean individuals. BMC Oral Health. 2021;21(1):130.
doi: 10.1186/s12903-021-01513-3
pubmed: 33736627
pmcid: 7977585
Alwosheel A, van Cranenburgh S, Chorus CG. Is your dataset big enough? Sample size requirements when using artificial neural networks for discrete choice analysis. J Choice Modelling. 2018;28:167–82.
doi: 10.1016/j.jocm.2018.07.002
Schopf P. Kieferorthopädische Abrechnung (BEMA und GOZ GOÄ) mit Erläuterung der ab 1.1.2002 gültigen Kieferorthopädischen Indikationsgruppen (KIG) sowie der ab 1.1.2004 geltenden Fassung des BEMA und der Richtlinien des Bundesausschusses der Zahnärzte und Krankenkassen.
Ryu J, Kim YH, Kim TW, Jung SK. Evaluation of artificial intelligence model for crowding categorization and extraction diagnosis using intraoral photographs. Sci Rep. 2023;13(1).
Jackson T, Kirk C, Phillips C, Koroluk L. Diagnostic accuracy of intraoral photographic orthodontic records. J Esthetic Restor Dentistry. 2018;31.