Prediction of midpalatal suture maturation stage based on transfer learning and enhanced vision transformer.
Cone beam computed tomography images
Midpalatal suture maturation stages
Self-attention
Vision transformer
Journal
BMC medical informatics and decision making
ISSN: 1472-6947
Titre abrégé: BMC Med Inform Decis Mak
Pays: England
ID NLM: 101088682
Informations de publication
Date de publication:
22 Aug 2024
22 Aug 2024
Historique:
received:
06
01
2024
accepted:
02
07
2024
medline:
23
8
2024
pubmed:
23
8
2024
entrez:
22
8
2024
Statut:
epublish
Résumé
Maxillary expansion is an important treatment method for maxillary transverse hypoplasia. Different methods of maxillary expansion should be carried out depending on the midpalatal suture maturation levels, and the diagnosis was validated by palatal plane cone beam computed tomography (CBCT) images by orthodontists, while such a method suffered from low efficiency and strong subjectivity. This study develops and evaluates an enhanced vision transformer (ViT) to automatically classify CBCT images of midpalatal sutures with different maturation stages. In recent years, the use of convolutional neural network (CNN) to classify images of midpalatal suture with different maturation stages has brought positive significance to the decision of the clinical maxillary expansion method. However, CNN cannot adequately learn the long-distance dependencies between images and features, which are also required for global recognition of midpalatal suture CBCT images. The Self-Attention of ViT has the function of capturing the relationship between long-distance pixels of the image. However, it lacks the inductive bias of CNN and needs more data training. To solve this problem, a CNN-enhanced ViT model based on transfer learning is proposed to classify midpalatal suture CBCT images. In this study, 2518 CBCT images of the palate plane are collected, and the images are divided into 1259 images as the training set, 506 images as the verification set, and 753 images as the test set. After the training set image preprocessing, the CNN-enhanced ViT model is trained and adjusted, and the generalization ability of the model is tested on the test set. The classification accuracy of our proposed ViT model is 95.75%, and its Macro-averaging Area under the receiver operating characteristic Curve (AUC) and Micro-averaging AUC are 97.89% and 98.36% respectively on our data test set. The classification accuracy of the best performing CNN model EfficientnetV2_S was 93.76% on our data test set. The classification accuracy of the clinician is 89.10% on our data test set. The experimental results show that this method can effectively complete CBCT images classification of midpalatal suture maturation stages, and the performance is better than a clinician. Therefore, the model can provide a valuable reference for orthodontists and assist them in making correct a diagnosis.
Sections du résumé
BACKGROUND
BACKGROUND
Maxillary expansion is an important treatment method for maxillary transverse hypoplasia. Different methods of maxillary expansion should be carried out depending on the midpalatal suture maturation levels, and the diagnosis was validated by palatal plane cone beam computed tomography (CBCT) images by orthodontists, while such a method suffered from low efficiency and strong subjectivity. This study develops and evaluates an enhanced vision transformer (ViT) to automatically classify CBCT images of midpalatal sutures with different maturation stages.
METHODS
METHODS
In recent years, the use of convolutional neural network (CNN) to classify images of midpalatal suture with different maturation stages has brought positive significance to the decision of the clinical maxillary expansion method. However, CNN cannot adequately learn the long-distance dependencies between images and features, which are also required for global recognition of midpalatal suture CBCT images. The Self-Attention of ViT has the function of capturing the relationship between long-distance pixels of the image. However, it lacks the inductive bias of CNN and needs more data training. To solve this problem, a CNN-enhanced ViT model based on transfer learning is proposed to classify midpalatal suture CBCT images. In this study, 2518 CBCT images of the palate plane are collected, and the images are divided into 1259 images as the training set, 506 images as the verification set, and 753 images as the test set. After the training set image preprocessing, the CNN-enhanced ViT model is trained and adjusted, and the generalization ability of the model is tested on the test set.
RESULTS
RESULTS
The classification accuracy of our proposed ViT model is 95.75%, and its Macro-averaging Area under the receiver operating characteristic Curve (AUC) and Micro-averaging AUC are 97.89% and 98.36% respectively on our data test set. The classification accuracy of the best performing CNN model EfficientnetV2_S was 93.76% on our data test set. The classification accuracy of the clinician is 89.10% on our data test set.
CONCLUSIONS
CONCLUSIONS
The experimental results show that this method can effectively complete CBCT images classification of midpalatal suture maturation stages, and the performance is better than a clinician. Therefore, the model can provide a valuable reference for orthodontists and assist them in making correct a diagnosis.
Identifiants
pubmed: 39174951
doi: 10.1186/s12911-024-02598-w
pii: 10.1186/s12911-024-02598-w
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
232Informations de copyright
© 2024. The Author(s).
Références
Reyneke JP, Conley RS. Surgical/Orthodontic correction of transverse maxillary discrepancies. Oral Maxillofac Surg Clin N Am. 2020;32(1):53–69.
doi: 10.1016/j.coms.2019.08.007
Yoon A, Guilleminault C, Zaghi S, Liu SY. Distraction Osteogenesis Maxillary expansion (DOME) for adult obstructive sleep apnea patients with narrow maxilla and nasal floor. Sleep Med. 2020;65:172–6.
doi: 10.1016/j.sleep.2019.06.002
pubmed: 31606311
Sitzia E, Santarsiero S, Tucci FM, De Vincentiis G, Galeotti A, Festa P. Balloon dilation and rapid maxillary expansion: a novel combination treatment for congenital nasal pyriform aperture stenosis in an infant. Ital J Pediatr. 2021;47(1):189.
doi: 10.1186/s13052-021-01124-2
pubmed: 34530869
pmcid: 8447711
Patil GV, Lakhe P, Niranjane P. Maxillary expansion and its effects on Circummaxillary structures: a review. Cureus. 2023;15(1):e33755.
pubmed: 36793826
pmcid: 9922614
Gao L, Sun J, Zhou X, Yu G. In vivo methods for evaluating human midpalatal suture maturation and ossification: an updated review. Int Orthod. 2022;20(2):100634.
doi: 10.1016/j.ortho.2022.100634
pubmed: 35589538
Shayani A, Sandoval Vidal P, Garay Carrasco I, Merino Gerlach M. Midpalatal Suture Maturation Method for the Assessment of Maturation before Maxillary Expansion: a systematic review. Diagnostics (Basel Switzerland) 2022, 12(11).
Liu S, Xu T, Zou W. Effects of rapid maxillary expansion on the midpalatal suture: a systematic review. Eur J Orthod. 2015;37(6):651–5.
doi: 10.1093/ejo/cju100
pubmed: 25700989
Luyten J, De Roo NMC, Christiaens J, Van Overberghe L, Temmerman L, De Pauw GAM. Rapid maxillary expansion vs slow maxillary expansion in patients with cleft lip and/or palate: a systematic review and meta-analysis. Angle Orthod. 2023;93(1):95–103.
doi: 10.2319/030122-188.1
pubmed: 36240430
Ok UPD, Kaya TU. Fractal Perspective on the Rapid Maxillary Expansion Treatment; evaluation of the Relationship between Midpalatal suture opening and Dental effects. J Stomatology oral Maxillofacial Surg. 2022;123(4):422–8.
doi: 10.1016/j.jormas.2021.09.002
Samra DA, Hadad R. Skeletal age-related changes of Midpalatal suture densities in skeletal Maxillary Constriction patients: CBCT Study. J Contemp Dent Pract. 2018;19(10):1260–6.
doi: 10.5005/jp-journals-10024-2414
pubmed: 30498183
Rachmiel A, Turgeman S, Shilo D, Emodi O, Aizenbud D. Surgically assisted Rapid Palatal expansion to correct maxillary transverse Deficiency. Annals Maxillofacial Surg. 2020;10(1):136–41.
doi: 10.4103/ams.ams_163_19
de Oliveira CB, Ayub P, Ledra IM, Murata WH, Suzuki SS, Ravelli DB, Santos-Pinto A. Microimplant assisted rapid palatal expansion vs surgically assisted rapid palatal expansion for maxillary transverse discrepancy treatment. Am J Orthod Dentofac Orthopedics: Official Publication Am Association Orthodontists its Constituent Soc Am Board Orthod. 2021;159(6):733–42.
doi: 10.1016/j.ajodo.2020.03.024
Chamberland S. Maxillary expansion in nongrowing patients. Conventional, surgical, or miniscrew-assisted, an update. J World Federation Orthodontists. 2023;12(4):173–83.
doi: 10.1016/j.ejwf.2023.04.005
Ventura V, Botelho J, Machado V, Mascarenhas P, Pereira FD, Mendes JJ, Delgado AS, Pereira PM. Miniscrew-assisted Rapid Palatal Expansion (MARPE): an Umbrella Review. J Clin Med 2022, 11(5).
Colonna A, Cenedese S, Sartorato F, Spedicato GA, Siciliani G, Lombardo L. Association of the mid-palatal suture morphology to the age and to its density: a CBCT retrospective comparative observational study. Int Orthod. 2021;19(2):235–42.
doi: 10.1016/j.ortho.2021.03.002
pubmed: 33785290
Angelieri F, Cevidanes LH, Franchi L, Gonçalves JR, Benavides E, McNamara JA Jr. Midpalatal suture maturation: classification method for individual assessment before rapid maxillary expansion. Am J Orthod Dentofac Orthopedics: Official Publication Am Association Orthodontists its Constituent Soc Am Board Orthod. 2013;144(5):759–69.
doi: 10.1016/j.ajodo.2013.04.022
Chun JH, de Castro ACR, Oh S, Kim KH, Choi SH, Nojima LI, Nojima M, Lee KJ. Skeletal and alveolar changes in conventional rapid palatal expansion (RPE) and miniscrew-assisted RPE (MARPE): a prospective randomized clinical trial using low-dose CBCT. BMC Oral Health. 2022;22(1):114.
doi: 10.1186/s12903-022-02138-w
pubmed: 35395801
pmcid: 8994336
Gao L, Chen Z, Zang L, Sun Z, Wang Q, Yu G. Midpalatal Suture CBCT Image Quantitive Characteristics Analysis Based on Machine Learning Algorithm Construction and optimization. Bioeng (Basel Switzerland) 2022, 9(7).
Hung K, Yeung AWK, Tanaka R, Bornstein MM. Current applications, opportunities, and limitations of AI for 3D imaging in Dental Research and Practice. Int J Environ Res Public Health 2020, 17(12).
Wang SF, Xie XJ, Zhang L, Chang S, Zuo FF, Wang YJ, Bai YX. [Research on multi-class orthodontic image recognition system based on deep learning network model]. Zhonghua Kou Qiang Yi Xue Za Zhi = Zhonghua Kouqiang Yixue Zazhi = Chin J Stomatology. 2023;58(6):561–8.
Hung KF, Ai QYH, Wong LM, Yeung AWK, Li DTS, Leung YY. Current applications of Deep Learning and Radiomics on CT and CBCT for Maxillofacial diseases. Diagnostics (Basel Switzerland) 2022, 13(1).
Duman ŞB, Syed AZ, Celik Ozen D, Bayrakdar İ, Salehi HS, Abdelkarim A, Celik Ö, Eser G, Altun O, Orhan K. Convolutional Neural Network Performance for Sella Turcica Segmentation and classification using CBCT images. Diagnostics (Basel Switzerland) 2022, 12(9).
Zhu M, Yang P, Bian C, Zuo F, Guo Z, Wang Y, Wang Y, Bai Y, Zhang N. Convolutional neural network-assisted diagnosis of midpalatal suture maturation stage in cone-beam computed tomography. J Dent. 2024;141:104808.
doi: 10.1016/j.jdent.2023.104808
pubmed: 38101505
Liu Y, Yu J, Han Y. Understanding the effective receptive field in semantic image segmentation. MULTIMEDIA TOOLS Appl. 2018;77:22159–71.
doi: 10.1007/s11042-018-5704-3
Gongbo T, Mathias M, Annette R, Rico S. Why self-attention? A targeted evaluation of neural machine translation architectures. arXiv 2018:arXiv:180808946.
Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, Dehghani M, Minderer M, Heigold G, Gelly S. An image is worth 16x16 words: transformers for image recognition at scale. arXiv 2020:arXiv:2010.11929.
Battaglia PW, Hamrick JB, Bapst V, Sanchez-Gonzalez A, Zambaldi V, Malinowski M, Tacchetti A, Raposo D, Santoro A, Faulkner R. Relational inductive biases, deep learning, and graph networks. arXiv 2018:arXiv:180601261.
Shorten C, Khoshgoftaar, TMJJobd. A survey on image data augmentation for deep learning. J Big Data. 2019;6(1):1–48.
doi: 10.1186/s40537-019-0197-0
Sun L, Jiang Z, Chang Y, Ren L. Building a patient-specific model using transfer learning for four-dimensional cone beam computed tomography augmentation. Quant Imaging Med Surg. 2021;11(2):540–55.
doi: 10.21037/qims-20-655
pubmed: 33532255
pmcid: 7779907
Sun L, Dong J, Tang J, Pan J. Spatially-adaptive feature modulation for efficient image Super-resolution. arXiv 2023:arXiv:2302.13800.
d’Ascoli S, Touvron H, Leavitt ML, Morcos AS, Biroli G, Sagun L. Convit: Improving vision transformers with soft convolutional inductive biases. arXiv 2021.
Okolo GI, Katsigiannis S, Ramzan N. IEViT: an enhanced vision transformer architecture for chest X-ray image classification. Comput Methods Programs Biomed. 2022;226:107141.
doi: 10.1016/j.cmpb.2022.107141
pubmed: 36162246
Dai Z, Liu H, Le QV, Tan M. Coatnet: Marrying convolution and attention for all data sizes. 2021, 34:3965–3977.
Wang W-Y, Tang Y-C, Du W-W, Peng W-C. NYCU_TWD@ LT-EDI-ACL2022: Ensemble models with VADER and contrastive learning for detecting signs of depression from social media.pp.137.
Yadav SP, Yadav S. Image fusion using hybrid methods in multimodality medical images. Med Biol Eng Comput. 2020;58(4):669–87.
doi: 10.1007/s11517-020-02136-6
pubmed: 31993885
Singh D, Singh B. Investigating the impact of data normalization on classification performance. Appl Soft Comput 2020, 97.
Mutasa S, Sun S, Ha R. Understanding artificial intelligence based radiology studies: what is overfitting? Clin Imaging. 2020;65:96–9.
doi: 10.1016/j.clinimag.2020.04.025
pubmed: 32387803
pmcid: 8150901
Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A. Going deeper with convolutions. arXiv 2015:1–9.arXiv:1409.4842.
He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2016:770–778.
Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser Ł. Polosukhin IJAinips: Attention is all you need. Annual Conference on Neural Information Processing Systems 2017, 30.
Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, Huang Z, Karpathy A, Khosla A. Imagenet large scale visual recognition challenge. Int J Comput Vision 2015:211–52.
Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z. Rethinking the inception architecture for computer vision. Proceedings of the IEEE conference on computer vision and pattern recognition 2016:2818–2826.
Lever J, Krzywinski M, Altman N. Classification evaluation: it is important to understand both what a classification metric expresses and what it hides. 2016, 13(8):603–5.
Liu Z, Lin Y, Cao Y, Hu H, Wei Y, Zhang Z, Lin S, Guo B. Swin transformer: Hierarchical vision transformer using shifted windows. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021: pp. 9992–10002.
Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition. arXiv 2014.
Omiotek Z, Kotyra A. Flame image Processing and classification using a pre-trained VGG16 model in Combustion diagnosis. Sensors 2021, 21(2).
Tan M, Le Q. Efficientnet: Rethinking model scaling for convolutional neural networks. International conference on machine learning 2019:6105–6114.
Dou S, Wang L, Fan D, Miao L, Yan J, He H. Classification of Citrus Huanglongbing Degree based on CBAM-MobileNetV2 and transfer learning. Sensors 2023, 23(12).
Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D. Grad-cam: Visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE international conference on computer vision 2017:pp.618–626.