Artificial intelligence (AI) involves machines exhibiting their own type of intelligence, with the aim of enabling them to learn from data and solve problems autonomously. Although AI and machine learning (ML) are frequently used interchangeably in research, they represent distinct concepts. Arthur Samuel first used the term “machine learning” in 1959 [1]. It is a subset of artificial intelligence. ML uses algorithms, such as artificial neural networks (ANN), to predict outcomes based on provided datasets. ANNs are made up of interconnected artificial neurons that mimic the structure of the central nervous system. Deep learning (DL) is an extension of ML that possess multiple computational layers to automatically detect patterns and enhance feature recognition. By extracting features from abstracted layers of filters, deep learning is particularly effective in analyzing large and complex images [2].

The term artificial intelligence (AI) now encompasses two distinct subconcepts: strong AI and weak AI [3]. Strong AI, often known as Artificial General Intelligence (AGI), is capable of performing every task that an individual can do. In essence, these computing systems can independently reach a level of self-awareness, allowing them to recognize and "understand" objects in an autonomous and proactive manner. Strong AI possesses cognitive abilities similar to those of humans [4]. Weak AI, often referred to as artificial narrow intelligence (ANI), lacks cognitive abilities and is intended to carry out particular tasks or concentrate on a single task at a time. The concept of weak AI acknowledges that fully replicating human intelligence is fundamentally impossible, as human cognition extends beyond mathematical processes. In practice, human thoughts and actions often occur spontaneously and arbitrarily, without relying solely on calculations. The frequent use of weak AI is the implementation of machine learning through bio-inspired artificial neural networks [5].

Many oral diseases, especially oral cancer and potentially malignant disorders (PMDs), present subtle signs that are easily missed during routine clinical examinations. Manual interpretation of dental radiographs is time-consuming and susceptible to oversights or misinterpretation, especially in complex cases such as periodontal bone loss, cysts, or root fractures. Artificial intelligence (AI) has demonstrated potential in modern healthcare for disease diagnosis, patient outcome prediction, and individualized treatment regimens. Specifically, AI allows dentists make vital, time-sensitive judgments (Patil et al., 2022)2

A large volume of datasets is essential for applying machine learning. These datasets can include different types of inputs, such as photographs, radiographs, patient records, and symptom details. In dentistry, a large number of images are produced annually. For instance, in many European countries, dental radiography-constitutes the majority of all radiographic imaging. With the inclusion of additional imaging methods, such as photographs and 3D scans, there is significant potential for applying convolutional neural networks (CNNs) in dentistry [6].

A search was conducted to identify relevant articles for this literature review. The databases used for the search included PubMed, Scopus, and Google Scholar to ensure comprehensive coverage of the existing literature. The search was performed separately for each oral disease included in this review to obtain disease-specific articles.

Appropriate keywords were used to refine the search for each disease. For example, keywords such as ‘Artificial intelligence”, “oral cancer,” “periodontitis,” “dental caries,” and “Orthodontic diagnosis”, “Temporomandibular joint disorder” were used for search.

Inclusion criteria

Articles published in English

Studies relevant to the topic of interest

Original research articles and review articles

Articles with clear methodology and results

Exclusion criteria

Non-English publications

Case reports and letters to the editor

Studies lacking clear methodology or with insufficient data

The majority of the articles (approximately 60%) were selected from the last five years (2020 to 2025) to ensure up-to-date evidence while including relevant older studies where necessary for background and foundational concepts.

AI devices can be roughly categorized into two types. The first group comprises ML approaches for analyzing structured data, such as imaging, genetic, and electrophysiology data. The second group includes natural language processing (NLP) approaches that extract insights from unstructured data, thereby supplementing and improving structured medical information [7].

Figure 1: The Pathway From Clinical Data Collection To Natural Language Processing For Data Enhancement, Followed By Machine Learning-Based Data Analysis, Ultimately Leading To Informed Clinical Decision-Making.

Application of Artificial Intelligence in Diagnosis and Prognosis of Diverse

Oral Diseases

AI and Oral cancer

Detecting oral cancer at an early stage greatly enhances prognosis and increases survival rates. Artificial intelligence (AI) has the capability to enable early detection, helping to lower the rates of mortality and morbidity linked to this disease. Nayak et al. used artificial neural networks (ANN) to analyze laser-induced autofluorescence spectra and distinguish between normal and precancerous situations [8]. To distinguish between premalignant and malignant lesions using autofluorescence and white-light photos, Uthoff et al. also used convolutional neural networks. Their investigation showed that CNN performed better than experts in identifying these lesions, indicating its potential for autonomous image-based classification and detection of oral cancer [9]. Furthermore, Ariji et al. investigated the effectiveness of deep learning (DL) to identify cervical lymph node metastases by employing CT imaging. The DL model demonstrated competence equivalent to that of expert radiologists. Furthermore, the researchers pointed out that DL could accurately detect extra-nodal extension in metastatic cervical lymph nodes [10]. The DL system prevailed over radiologists in this challenge, indicating its potential in the diagnosis of extra-nodal metastases.

Figure 2: Artificial Intelligence in the Diagnosis of Oral Cancer [11-17].

AI And Temporomandibular Joint Disorders

Diagnosing temporomandibular joint disorders can be challenging, especially for less experienced dentists. Artificial neural networks (ANN) offer a means to simplify and enhance this diagnostic process. In a study using clinical symptoms of TMJ problems, Bas et al. found that the ANN performed satisfactorily in identifying internal TMJ derangements. The diagnosis accuracy of these systems could be further improved by adding more imaging, radiographs, clinical findings, and patient data [18]. Similarly, Orhan et al. utilized magnetic resonance imaging (MRI) to diagnose TMJ pathologies, including condylar changes and disk derangements, through the application of an AI-based model [19]. The precision and dependability of the diagnostic procedure depend heavily on the data collected for automated TMD diagnosis. The success of the system in recognizing and categorizing TMD symptoms is largely dependent on the kind and caliber of data utilized to train and evaluate ML or DL models. The size of the training dataset is particularly crucial, as larger datasets typically produce more accurate and generalizable models. In contrast, smaller datasets can lead to overfitting, where the model performs well on the training data but struggles to generalize to new, unseen cases [20].

Precise image segmentation enhances a machine's ability to interpret visual information, facilitating the creation of advanced and intelligent systems. Deep learning methods, particularly CNNs, have greatly advanced image segmentation by automatically extracting meaningful features from data [21]. Vinayagalingam et al. employed a deep learning method using a 3D U-Net architecture for TMJ segmentation and found that AI-based segmentation offers excellent accuracy and speed [22]. For patients who had segmental mandibular resection and reconstruction, the AI models were used for the evaluation of asymmetry that worked well as the main aesthetic outcome indicator [23].

Figure 3: Artificial Intelligence in the Diagnosis of Temporomandibular Joint Disorders.

To securely store data, implement algorithms, and carry out operations in a web-based environment, the Data Storage for Computation and Integration (DSCI) provided novel management features. An automated image processing method for TMJ segmentation utilizing high quality images was included of the software [24].

The swift advancement of technology has led to a significant surge in the amount of recorded data. Effective data management and analysis enable the extraction of valuable insights that enhance public health and overall well-being. A clinical decision support system (CDSS) is a computer-based tool that helps people evaluate different possibilities, speeding up problem-solving and decision-making processes [25].

However, widespread use of CDSS in dentistry clinics is still impracticable and requires additional development in key areas [26].

TMJ Clinical Decision Support System

The TMJ CDSS, which is integrated into the DSCI, allows for more accurate identification of TMJ osteoarthritis and observation of surface changes in the afflicted condyles [25].

Figure 4: Workflow of Clinical Decision Support System for Osteoarthritis..

Changes in the surfaces of the condyles are recognized by the Shape Variation Analyzer, a machine learning model that is part of the DSCI. Based on the degree of TMJ degeneration, this tool aids in the diagnosis and categorization of TMJ osteoarthritis [24].

Using orthopantomogram (OPG) pictures, Choi et al. Created an AI model that can detect osteoarthritis in TMJs [27].

Figure 5: Examples of AI Models Employed In the Diagnosis of TMD.

These developments highlight AI's potential in enhancing the detection and diagnosis of TMJ disorders.

Dental caries is the most widespread disease worldwide, with diagnosis primarily relying on visual assessments and radiographic images. These images can serve as input data for machine learning (ML) applications. In a study conducted by Devito et al., researchers investigated whether applying an artificial intelligence model could improve the radiographic detection of proximal caries. Their findings revealed a diagnostic improvement of 39.4% [28].

Lee et al. utilized a U-shaped deep convolutional neural network (U-Net) to detect caries in bitewing radiographs. The model enhanced clinicians' overall diagnostic performance, with notable improvements observed in the detection of initial and moderate caries [29].

A study by Cantu et al. found that a deep neural network (DNN) performed better than dentists in identifying caries lesions on bitewing radiographs with an accuracy of 0.80, which was significantly greater than the dentists [30]. Applied various supervised machine learning techniques to predict root caries in individuals. Among the models developed, the support vector machine (SVM) exhibited the highest performance in identifying root caries. Utilizing machine learning for root caries prediction gives a tremendous potential for early intervention and improved dental health, particularly among the aging population. This method could be used as a screening tool in a variety of locations, including dental clinics, medical offices, social service organizations, and even online, recommending dental exams to high-risk individuals [31]. Using near-infrared transillumination (TI) imaging, Casalegno et al. utilized a deep learning CNN model to identify caries early. They stated that the speed and accuracy of caries diagnosis were improved by this deep learning technique [32]. A comparison between machine learning-based prediction models and traditional logistic regression models for detecting early childhood caries revealed that both approaches demonstrated favorable performance [33]. The success of endodontic treatment largely relies on the precise working length determination. Saghiri et al. utilized an ANN system for this purpose and demonstrated remarkable accuracy of 96%, surpassing the accuracy achieved by professional endodontists [34]. Similarly, Fukuda et al. conducted a study using a CNN model for detecting vertical root fractures, demonstrating highly promising precision [35].

Table 1: Examples of Machine Learning Algorithms for Diagnosis of Dental Caries AI and Periodontal Diseases.

Investigated the effectiveness of artificial neural networks, decision trees (DT), and SVM for the diSagnosis and categorization of periodontal disorders. In comparison to ANN, their results showed that SVM and DT were more accurate diagnostic support tools [36]. Oral malodour primarily results from volatile sulphur compounds produced by bacterial activity and interactions. A study investigated the use of a DL model to predict oral malodour based on salivary microbiota. The model demonstrated a predictive accuracy of 97% and is anticipated to be a valuable tool for saliva-based screening of oral malodour [37]. Using a deep CNN algorithm, Lee et al. developed a computer-assisted detection method that uses periapical radiographs to identify and predict teeth with poor periodontal health. The system achieved prediction accuracies of 82.8% for premolars and 73.4% for molars regarding the need for extraction [38]. Krois et al. applied CNNs to detect periodontal bone loss (PBL) on orthopantamograms. The study's results closely matched expert opinions, and this system has the potential to assist in reducing the diagnostic workload for dentists [39].

Jungdaeng et al. utilized the YOLOv8 model for tooth segmentation in panoramic radiographs, achieving an accuracy of 97% [40]. This indicates that AI models are capable of accurately detecting periodontal bone loss from panoramic radiographs, surpassing existing diagnostic techniques. The classification of periodontal disease has also been studied using SVM and DT models; SVM has demonstrated better performance in distinguishing between healthy and diseased tissues using radiographic characteristics [41]. Furthermore, numerous studies in the field have demonstrated the potential for improving the accuracy of periodontal disease identification by hybrid techniques that combine CNNs with SVM or DT [42].

Figure 6: Application of AI in Periodontology AI in Orthodontic Diagnosis.

Orthodontic practice often needs extensive time for numerous investigations, relying heavily on the clinical competence of orthodontists. This workload influences the efficiency of clinical practice and limits the accessibility of orthodontic treatment for non-specialists due to the experience necessary. However, as AI algorithms evolve, computing power increases, and datasets become more widely available, the applications of AI in orthodontics expand, as do performance gains [43].

Cephalometric Analysis

The most widely researched field of AI applications in orthodontics is now cephalometric analysis. Compared to conventional ML techniques, DL-in particular, convolutional neural networks, has demonstrated better performance and drawn more study interest. Numerous research have shown that the popular CNN algorithm You-Only-Look-Once version 3 (YOLOv3) has achieved remarkable accuracy in automatic landmark detection [44-46].

Kim et al. collected 3,150 lateral cephalograms from various locations throughout the nation using nine distinct cephalography machines to reduce the risk of overfitting and improve the data's generalizability. By applying a cascade CNN model, the researchers achieved an average automated detection error of 1.36 ± 0.98 mm [47].

Simultaneously, studies have documented the use of CNNs for automated landmark detection on posteroanterior cephalograms, for the facial symmetry evaluation [48].

Dental and Facial Analysis

In orthodontic practice, intraoral photographs and orthodontic study models are essential for dental analysis. AI has the potential to automate this process, reducing the need for human involvement. Using the YOLO technique, Talaat et al. were able to detect malocclusion in intraoral photos with a remarkable 99.99% accuracy rate [49].

Figure 7: Examples of Algorithms Used for Cephalometric Analysis.

CNN - Convolutional Neural Network; GCN- Graph Convolutional Networks; YOLOv3- You-Only-Look- Once version 3

Table 2: List of Few Commercially Available Softwares for Cephalometric Analysis.

Companies like Invisalign have successfully leveraged 3D oral scan data and digital models for automated dental measurement and analysis. Software tools such as Dynamic-Graph Convolutional Neural Network (DGCNN), OrthoAnalyzer, and Autolign are commonly used for automated tooth segmentation. Among these, DGCNN demonstrated the highest accuracy and faster computational performance [50].

 For evaluating facial proportions and asymmetry, facial photos are essential. However, future study is needed to improve the accuracy and expand the applications of automated facial analysis, which is still in its infancy [43].

Skeletal Maturation Determination

When receiving orthodontic treatment, evaluating a patient’s growth spurt is crucial, especially in situations when functional or orthopedic adjustments are needed. Hand-wrist radiographs or lateral cephalograms, which measure cervical vertebral maturation (CVM), can be used to identify skeletal maturation. ANN were shown to be the most reliable machine learning technique when Kök et al. Used seven different algorithms to detect CVM stages, with differing degrees of accuracy [51]. A study by Amasya et al. Found similar results [52]. One of the first groups to compare the effectiveness of six unsupervised CNN models was Seo et al. According to their research, all of the algorithms performed better than Inception-ResNet-v2, with an accuracy rate of above 90% [53].

In summary, while artificial neural networks (ANNs) initially garnered significant attention, convolutional neural networks (CNNs) have increasingly gained prominence in recent years, particularly for image-related tasks.

Treatment Outcome Predictions

By forecasting treatment results, orthodontists can take a more scientific approach to malocclusion examination and treatment, reducing risks and complications during and after treatment. Particularly in the clear aligner industry, automated virtual setups have grown in popularity because to continuous developments in digital orthodontics and artificial intelligence [54]. Numerous studies have looked into using AI to forecast skeletal and facial changes after orthodontic treatment, in addition to tooth changes. Park et al. Predicted cephalometric changes in Class II patients using a CNN model [55].

Tanikawa et al. Used deep learning and geometric morphometric approaches to forecast changes in 3D facial morphology after orthognathic surgery or orthodontic therapy (including the extraction of four premolars) [56].

AI And Salivary Gland Diseases

Salivary gland illnesses can be difficult to diagnose for inexperienced dentists due to their complicated and comparable physical characteristics. Artificial intelligence (AI) is increasingly being integrated into the diagnosis and prognosis of salivary gland tumors, enhancing accuracy and aiding clinical decision-making. Deep learning (DL) models have, in certain cases, demonstrated performance superior to that of radiologists. DL was used in an early Japanese study to detect fatty degeneration on CT scans, which is a hallmark of Sjögren’s syndrome.The study revealed that DL models performed on par with experienced radiologists and significantly outperformed those with limited experience [57].

The review synthesizes findings from a range of studies assessing the role of AI in diagnosing and predicting various oral diseases. Most of the literature demonstrates that AI-based models improve diagnostic accuracy, reduce clinical subjectivity, and offer faster processing of clinical and radiological data.

AI in Dental Caries Detection

Several models including CNN, ANN, and SVM have been successfully implemented for the detection of dental caries, as outlined in Table 1. CNNs consistently outperform traditional algorithms in image-based diagnostics due to their superior feature extraction capabilities. The models exhibit high sensitivity and specificity in detecting early-stage lesions, particularly when applied to bitewing and periapical radiographs.

Comparative Performance of AI Models

Table 2: presents different AI models across various oral diagnostic tasks. CNNs are particularly effective in radiograph interpretation and lesion classification, while ANNs and decision trees are applied more frequently for structured data such as clinical parameters and symptom sets. SVMs, although useful, show slightly lower performance due to limitations in learning from image-heavy datasets.

Application across Disease Categories

AI has demonstrated broad applicability across caries detection, periodontal disease staging, oral cancer diagnosis, and TMJ analysis. AI-driven Clinical Decision Support Systems (CDSS) and diagnostic tools-illustrated in Figures 1 through 5-have shown the ability to process complex datasets, segment lesion areas, and suggest diagnostic outcomes. Figure 4 demonstrates AI integration in TMJ CDSS workflows, while Figure 3 highlights CNN architecture in oral cancer image classification.

Key Observations

CNNs provide the highest diagnostic accuracy for image-based tasks (oral cancer, caries).

ANNs are efficient for prediction tasks using structured clinical data.

NLP applications are growing in relevance for text-based record analysis and documentation.

Integration of AI into clinical workflows is feasible, but limited by dataset quality, lack of standardization, and ethical concerns.      

Limitations of using Artificial intelligence in disease diagnosis

The limited availability and Current research is less trustworthy due to training data’s weak generalizability [58].

Some AI models used for decision support have not incorporated a broad variety of representative case types in their training data, raising doubts about their predictions for rare deformity cases [58].

Obtaining large amounts of high-quality data remains a big difficulty.

The lack of established evaluation frameworks and specialized data may hinder the continued use of few-shot learning [58].

Overfitting is a common problem in AI, when a model excels on training data but falls short on testing [59].

Artificial Intelligence (AI) has emerged as a transformative force in modern dentistry, enabling enhanced diagnostic accuracy, real-time decision support, and predictive analytics. The integration of AI-based systems-especially machine learning (ML), deep learning (DL), and neural networks-has been particularly impactful in the diagnosis and prognosis of dental caries, periodontal disease, oral cancer, and temporomandibular joint disorders. AI models like Convolutional Neural Networks (CNNs), Artificial Neural Networks (ANNs), and Natural Language Processing (NLP) frameworks have redefined diagnostic workflows by reducing subjectivity and increasing efficiency [6,29].

AI systems excel in pattern recognition and decision support, particularly in interpreting radiographs, segmenting images, and analyzing large volumes of data. CNNs have shown diagnostic accuracies comparable to or exceeding those of experienced clinicians in detecting dental caries [29] and oral cancer lesions [9]. ANN-based models have also proven useful in classifying lesion types and assessing prognosis [8]. In addition, NLP tools can automate the extraction of relevant clinical data from electronic records, streamlining documentation and enabling faster diagnostics [60].

Despite its promise, AI has limitations. Most models require large, high-quality datasets for training, which are often unavailable in dentistry. AI models may perform poorly when applied to unseen or heterogeneous clinical data. Additionally, their “black box” nature often lacks transparency, making it difficult for clinicians to interpret or trust the outcomes. Key challenges include the lack of standardized datasets, inconsistencies in data annotation, ethical concerns surrounding patient privacy, and limited regulatory guidance for clinical deployment. Furthermore, the integration of AI tools into existing dental software and workflows remains technically and logistically complex. Many studies focus on model development in research environments, with minimal translation into real-world clinical settings [61,62].

Collectively, existing literature indicates that AI models are well-suited for image-based diagnosis and structured clinical prediction tasks. CNNs, in particular, dominate radiographic and histopathological analysis, while ANNs and SVMs are better suited for structured data tasks. The current landscape suggests that AI is poised to support-but not yet replace-clinical judgment. Future research should focus on multicenter collaborations to generate large annotated datasets, the development of explainable AI models for improved transparency, and regulatory frameworks for ethical deployment. Emerging techniques like federated learning could enable training across decentralized data without compromising patient confidentiality. Transformer models such as BERT and GPT also offer potential for real-time documentation and triage support [61,62]

AI is already being applied in clinical decision support for caries detection, orthodontic analysis, periodontal disease staging, and oral cancer screening. Tools integrating AI into radiology workstations, CBCT interpretation, and automated charting systems have shown promising preliminary results. With further refinement and validation, AI can significantly augment the diagnostic capabilities of general dentists and specialists alike.

Artificial Intelligence has shown promising applications in the diagnosis and prognosis of various oral diseases, including dental caries, periodontal conditions, oral cancer, and temporomandibular joint disorders. AI models such as Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs) have demonstrated high accuracy in interpreting clinical and radiographic data, contributing to faster and more reliable diagnostic workflows.

Despite these advancements, the implementation of AI in routine dental practice is still evolving. Challenges such as the need for high-quality annotated datasets, standardization of model training, and ethical considerations related to patient data remain significant. Moreover, many AI models have not yet undergone large-scale clinical validation. Overall, while AI cannot yet replace clinical expertise, it serves as a valuable adjunct that enhances diagnostic precision, aids in treatment planning, and supports research and education. Future research should focus on multicenter data sharing, development of interpretable AI systems, and seamless integration into clinical environments to fully harness the potential of AI in dentistry.

1. Khanagar SB, Al-Ehaideb A, Maganur PC, Vishwanathaiah S, Patil S, Baeshen HA, et al. Developments, application, and performance of artificial intelligence in dentistry - A systematic review. J Dent Sci. 2021; 16: 508-522.
2. Patil S, Albogami S, Hosmani J, Mujoo S, Kamil MA, Mansour MA, et al. Artificial Intelligence in the Diagnosis of Oral Diseases: Applications and Pitfalls. Diagnostics (Basel, Switzerland). 2022; 12: 1029.
3. Park WJ, Park JB. History and application of artificial neural networks in dentistry. Eur J Dent. 2018; 12: 594-601.
4. Albus JS. Outline for a theory of intelligence. IEEE Transactions on Systems, Man, and Cybernetics. 1991; 21: 473-509.
5. Weinbaum D, Veitas V. Open ended intelligence: the individuation of intelligent agents. Journal of Experimental & Theoretical Artificial Intelligence. 2016; 29: 371-396.
6. Schwendicke F, Golla T, Dreher M, Krois J. Convolutional neural networks for dental image diagnostics: A scoping review. Journal of dentistry. 2019; 91: 103226.
7. Jiang F, Jiang Y, Zhi H, Dong Y, Li H, Ma S, et al. Artificial intelligence in healthcare: past, present and future. Stroke and vascular neurology. 2017; 2: 230-243.
8. Nayak GS, Kamath S, Pai KM, Sarkar A, Ray S, Kurien J, et al. Principal component analysis and artificial neural network analysis of oral tissue fluorescence spectra: classification of normal premalignant and malignant pathological conditions. Biopolymers. 2006; 82: 152-166.
9. Uthoff RD, Song B, Sunny S, Patrick S, Suresh A, Kolur T, et al. Point-of-care, smartphone-based, dual-modality, dual-view, oral cancer screening device with neural network classification for low-resource communities. PloS one. 2018; 13: e0207493.
10. Ariji Y, Sugita Y, Nagao T, Nakayama A, Fukuda M, Kise Y, et al. CT evaluation of extranodal extension of cervical lymph node metastases in patients with oral squamous cell carcinoma using deep learning classification. Oral radiology. 2020; 36: 148-155.
11. Nguyen PTH, Sakamoto K, Ikeda T. Deep-learning application for identifying histological features of epithelial dysplasia of tongue. J Oral Maxillofac Surg Med Pathol. 2022; 34: 514-522.
12. Talwar V, Singh P, Mukhia N, Shetty A, Birur P, Desai KM, et al. AI-Assisted Screening of Oral Potentially Malignant Disorders Using Smartphone-Based Photographic Images. Cancers. 2023; 15: 4120.
13. Warin K, Limprasert W, Suebnukarn S, Jinaporntham S, Jantana P. Automatic classification and detection of oral cancer in photographic images using deep learning algorithms. J Oral Pathol Med: official publication of the International Association of Oral Pathologists and the American Academy of Oral Pathology. 2021; 50: 911-918.
14. Warin K, Limprasert W, Suebnukarn S, Jinaporntham S, Jantana P. Performance of deep convolutional neural network for classification and detection of oral potentially malignant disorders in photographic images. Int J Oral Maxillofac Surg. 2022; 51: 699-704.
15. Marzouk R, Alabdulkreem E, Dhahbi SK, Nour M, Al Duhayyim M, Othman M, et al. Deep Transfer Learning Driven Oral Cancer Detection and Classification Model. Computers, Materials & Continua. 2022; 73: 3905-3920.
16. Jubair F, Al Karadsheh O, Malamos D, Al Mahdi S, Saad Y, Hassona Y. A novel lightweight deep convolutional neural network for early detection of oral cancer. Oral Dis. 2022; 28: 1123-1130.
17. Adami GR, Zhou Y, Kolokythas A. When will RNA-based tests similar to Oncotype DX be used for oral cancer? World Journal of Stomatology. 2015; 4: 121.
18. Bas B, Ozgonenel O, Ozden B, Bekcioglu B, Bulut E, Kurt M. Use of artificial neural network in differentiation of subgroups of temporomandibular internal derangements: a preliminary study. J Oral Maxillofac Surg: official journal of the American Association of Oral and Maxillofacial Surgeons.2012; 70: 51-59.
19. Orhan K, Driesen L, Shujaat S, Jacobs R, Chai X. Development and Validation of a Magnetic Resonance Imaging-Based Machine Learning Model for TMJ Pathologies. Biomed Res Int. 2021; 6656773.
20. Ozsari S, Guzel MS, Y?lmaz D, Kamburoglu K. A Comprehensive Review of Artificial Intelligence Based Algorithms Regarding Temporomandibular Joint Related Diseases. Diagnostics (Basel, Switzerland). 2023; 13: 2700.
21. Ito S, Mine Y, Yoshimi Y, Takeda S, Tanaka A, Onishi A, et al. Automated segmentation of articular disc of the temporomandibular joint on magnetic resonance images using deep learning. Sci Rep 2022; 12: 221.
22. Vinayahalingam S, Berends B, Baan F, Moin DA, van Luijn R, Berge S, et al. Deep learning for automated segmentation of the temporomandibular joint. Journal of dentistry. 2023; 132: 104475.
23. Hidaka T, Tanaka K, Mori H. An Artificial Intelligence-Based Cosmesis Evaluation for Temporomandibular Joint Reconstruction. Laryngoscope. 2023; 133: 841-848.
24. Brosset S, Dumont M, Cevidanes L, Soroushmehr R, Bianchi J, Gurgel M, et al. Web Infrastructure for Data Management, Storage and Computation. Proceedings of SPIE-the International Society for Optical Engineering. 2021; 11600: 116001N.
25. Al Turkestani N, Bianchi J, Deleat-Besson R, Le C, Tengfei L, Prieto JC, et al. Clinical decision support systems in orthodontics: A narrative review of data science approaches. Orthodontics & craniofacial research, 24 Suppl 2 (Suppl 2). 2021; 26-36.
26. Machoy ME, Szyszka-Sommerfeld L, Vegh A, Gedrange T, Wozniak K. The ways of using machine learning in dentistry. Advances in clinical and experimental medicine: official organ Wroclaw Medical University. 2020; 29: 375-384.
27. Choi E, Kim D, Lee JY, Park HK. Artificial intelligence in detecting temporomandibular joint osteoarthritis on orthopantomogram. Sci Rep. 2021; 11: 10246.
28. Devito KL, de Souza Barbosa F, Felippe Filho WN. An artificial multilayer perceptron neural network for diagnosis of proximal dental caries. Oral surgery, oral medicine, oral pathology, oral radiology, and endodontics. 2008; 106: 879-884.
29. Lee JH, Kim DH, Jeong SN, Choi SH. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm. Journal of dentistry. 2018; 77: 106-111.
30. Cantu AG, Gehrung S, Krois J, Chaurasia A, Rossi JG, Gaudin R, et al. Detecting caries lesions of different radiographic extension on bitewings using deep learning. Journal of dentistry. 2020; 100: 103425.
31. Hung M, Voss MW, Rosales MN, Li W, Su W, Xu J, et al. Application of machine learning for diagnostic prediction of root caries. Gerodontology. 2019; 36: 395-404.
32. Casalegno F, Newton T, Daher R, Abdelaziz M, Lodi-Rizzini A, Schurmann F, et al. Caries Detection with Near-Infrared Transillumination Using Deep Learning. J Dent Res. 2019; 98: 1227-1233.
33. Park YH, Kim SH, Choi YY. Prediction Models of Early Childhood Caries Based on Machine Learning Algorithms. Int J Environ Res Public Health. 2021; 18: 8613.
34. Saghiri MA, Asgar K, Boukani KK, Lotfi M, Aghili H, Delvarani A, et al. A new approach for locating the minor apical foramen using an artificial neural network. Int Endod J. 2012; 45: 257-265.
35. Fukuda M, Inamoto K, Shibata N, Ariji Y, Yanashita Y, Kutsuna S, et al. Evaluation of an artificial intelligence system for detecting vertical root fracture on panoramic radiography. Oral Radiol. 2020; 36: 337-343.
36. Ozden FO, Ozgonenel O, Ozden B, Aydogdu A. Diagnosis of periodontal diseases using different classification algorithms: a preliminary study. Niger J Clin Pract. 2015; 18: 416-421.
37. Nakano Y, Suzuki N, Kuwata F. Predicting oral malodour based on the microbiota in saliva samples using a deep learning approach. BMC oral Health. 2018; 18: 128.
38. Lee JH, Kim DH, Jeong SN, Choi SH. Diagnosis and prediction of periodontally compromised teeth using a deep learning-based convolutional neural network algorithm. J Periodontal Implant Sci. 2018; 48: 114-123.
39. Krois J, Ekert T, Meinhold L, Golla T, Kharbot B, Wittemeier A, et al. Deep Learning for the Radiographic Detection of Periodontal Bone Loss. Sci Rep. 2019; 9: 8495.
40. Jundaeng J, Chamchong R, Nithikathkul C. Artificial intelligence-powered innovations in periodontal diagnosis: a new era in dental healthcare. Front Med Technol. 2025; 6: 1469852.
41. Li X, Zhao D, Xie J, Wen H, Liu C, Li Y, et al. Deep learning for classifying the stages of periodontitis on dental images: a systematic review and meta-analysis. BMC oral health. 2023; 23: 1017.
42. Li H, Zhou J, Zhou Y, Chen Q, She Y, Gao F, et al. An Interpretable Computer-Aided Diagnosis Method for Periodontitis from Panoramic Radiographs. Front Physiol. 2021; 12: 655556.
43. Liu J, Zhang C, Shan Z. Application of Artificial Intelligence in Orthodontics: Current State and Future Perspectives. Healthcare (Basel, Switzerland). 2023; 11: 2760.
44. Moon JH, Hwang HW, Yu Y, Kim MG, Donatelli RE, Lee SJ. How much deep learning is enough for automatic identification to be reliable? The Angle orthodontist. 2020; 90: 823-830.
45. Hwang HW, Park JH, Moon JH, Yu Y, Kim H, Her SB, et al. Automated identification of cephalometric landmarks: Part 2-Might it be better than human? Angle Orthod. 2020; 90: 69-76.
46. Bulatova G, Kusnoto B, Grace V, Tsay TP, Avenetti DM, Sanchez FJC. Assessment of automatic cephalometric landmark identification using artificial intelligence. Orthodontics & craniofacial research. 2021; 24: 37-42.
47. Kim J, Kim I, Kim YJ, Kim M, Cho JH, Hong M, et al. Accuracy of automated identification of lateral cephalometric landmarks using cascade convolutional neural networks on lateral cephalograms from nationwide multi-centres. Orthodontics & craniofacial research. 2021; 24: 59-67.
48. Takeda S, Mine Y, Yoshimi Y, Ito S, Tanimoto K, Murayama T. Landmark annotation and mandibular lateral deviation analysis of posteroanterior cephalograms using a convolutional neural network. J Dent Sci. 2021; 16: 957-963.
49. Talaat S, Kaboudan A, Talaat W, Kusnoto B, Sanchez F, Elnagar MH, et al. The validity of an artificial intelligence application for assessment of orthodontic treatment need from clinical images. Semin Orthod. 2021; 27: 164-171.
50. Im J, Kim JY, Yu HS, Lee KJ, Choi SH, Kim JH, et al. Accuracy and efficiency of automatic tooth segmentation in digital dental models using deep learning. Sci Rep. 2022; 12: 9429.
51. Kok H, Acilar AM, Izgi MS. Usage and comparison of artificial intelligence algorithms for determination of growth and development by cervical vertebrae stages in orthodontics. Prog Orthod. 2019; 20: 41.
52. Amasya H, Yildirim D, Aydogan T, Kemaloglu N, Orhan K. Cervical vertebral maturation assessment on lateral cephalometric radiographs using artificial intelligence: comparison of machine learning classifier models. Dentomaxillofac Radiol. 2020; 49: 20190441.
53. Seo H, Hwang J, Jeong T, Shin J. Comparison of Deep Learning Models for Cervical Vertebral Maturation Stage Classification on Lateral Cephalometric Radiographs. Journal of clinical medicine. 2021; 10: 3591.
54. Woo H, Jha N, Kim YJ, Sung Sj. Evaluating the accuracy of automated orthodontic digital setup models. Semin Orthod. 2022.
55. Jae Hyun P, Kim I, Jae Hyun K, Kim J, Kim IH, Kim N, et al. (2021). Use of artificial intelligence to predict outcomes of nonextraction treatment of Class II malocclusions. Seminars in Orthodontics, 27(2), 87–95.
56. Tanikawa C, Yamashiro T. Development of novel artificial intelligence systems to predict facial morphology after orthognathic surgery and orthodontic treatment in Japanese patients. Sci Rep. 2021; 11: 15853.
57. Kise Y, Ikeda H, Fujii T, Fukuda M, Ariji Y, Fujita H, et al. Preliminary study on the application of deep learning system to diagnosis of Sjogren’s syndrome on CT images. Dentomaxillofac Radiol. 2019; 48: 20190019.
58. Ge Y, Guo Y, Das S, Al-Garadi MA, Sarker A. Few-shot learning for medical text: A review of advances, trends, and opportunities. J Biomed Inform. 2023; 144: 104458.
59. Wang K, Yang B, Li Q, Liu S. Systematic Evaluation of Genomic Prediction Algorithms for Genomic Prediction and Breeding of Aquatic Animals. Genes. 2022; 13: 2247.
60. Ferrucci D, Levas A, Bagchi S, Gondek D, Mueller ET. Watson: Beyond Jeopardy! Artificial Intelligence. 2013; 93-105.
61. Koroteev M. BERT: A Review of Applications in Natural Language Processing and Understanding. 2021.
62. Brown T, Mann BF, Ryder N, Subbiah M, Kaplan J, Dhariwal P, et al. Language Models Are Few-Shot Learners. ArXiv (Cornell University). 2020.