@article{nokey,

title = {A semi-automated assessment tool for craniofacial landmarks in CBCT: InVivo7 software},

author = {Luísa Schubach da Costa Barreto and Bruno Moreira das Neves and Jonas Bianchi and Heesoo Oh and Klaus Barretto dos Santos Lopes Batista and Jose Augusto Mendes Miguel},

url = {https://www.sciencedirect.com/science/article/pii/S0300571225007353?pes=vor&utm_source=scopus&getft_integrator=scopus},

doi = {https://doi.org/10.1016/j.jdent.2025.106292},

year  = {2026},

date = {2026-02-01},

urldate = {2026-02-01},

journal = {Journal of Dentistry},

volume = {165},

abstract = {Objectives: This study describes and evaluates the functionality of the InVivo7 3D imaging software as a semiautomated tool for identifying craniofacial landmarks in CBCT scans. Methods: AI-assisted landmark tracing in InVivo7 was used to automatically identify anatomical points in CBCT images. Each landmark was manually verified by a skilled evaluator to ensure accurate and reliable results, particularly for soft tissue markers and dental measurements, which often presented challenges for AI detection. The study utilized a standardized cephalometric analysis to compare the software’s performance. The evaluation included assessing the software’s ability to recognize skeletal, dental, and soft tissue structures accurately. Results: The semi-automated AI-assisted algorithm showed high precision in landmark identification. Manual verification confirmed its reliability and allowed the creation of a customized automated configuration for orthodontic diagnosis and treatment outcome evaluation. Specific clinical measures, such as the facial plane angle and molar relationships, were calculated using established formulas, allowing the software to categorize molar relationship classes (Angle Class I, II, III). Conclusions: InVivo7 presents a reliable and efficient tool for craniofacial landmark analysis, enhancing diagnostic accuracy while reducing manual labor. However, ongoing validation and software updates are essential to fully optimize its clinical applicability and ensure consistent performance across diverse patient populations. Future developments should focus on refining AI algorithms to improve soft tissue landmark detection and expanding datasets to enhance the robustness of automated analyses. Clinical Relevance: Rule-based automated algorithm CBCT craniofacial landmark detection using InVivo7 provides accurate, reproducible measurements, reducing manual workload and enhancing orthodontic diagnostic efficiency. Its integration into clinical practice supports standardized assessments, streamlining treatment planning.},

keywords = {3d, anatomic landmarks, artificial intelligence, CBCT, orthodontics},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {A comparative analysis of artificial intelligence generated 3d facial models from 2D photos: A procrustes-based validation of landmark accuracy in Orthodontics},

author = {Jonas Bianchi and Lorena Wilka and Gabriel Bravo Vallejo and Felicia Miranda and Camila Massaro and Lucia Cevidanes and Heesoo Oh},

url = {https://www.sciencedirect.com/science/article/pii/S1073874625001409?pes=vor&utm_source=scopus&getft_integrator=scopus},

doi = {https://doi.org/10.1053/j.sodo.2025.10.014},

issn = {1073-8746},

year  = {2025},

date = {2025-10-29},

journal = {Seminars in Orthodontics},

pages = {1-9},

abstract = {Objectives: The primary aim of this study was to evaluate the accuracy and precision of a deep learning artificial intelligence (AI) model in reconstructing 3D facial geometry from a single 2D photograph. The study specifically measured the coordinate differences of human-placed facial landmarks on the AI-generated models against those placed on 3D models from a gold-standard scanner (Bellus 3D). Secondary objectives included identifying dimensional variations (X, Y, Z) and systematic biases in the AI reconstructions. Methods: This was a retrospective comparative study using a sample of 47 anonymized patient records from the

University of the Pacific database. For each patient, a 3D facial model was generated from a 2D photograph using the 3DDFA-V2 deep learning model. These AI-generated models were compared to reference models captured on the same day using a Bellus 3D Scanner. Two trained observers identified a set of facial landmarks on both the AI generated and Bellus 3D models. The coordinate data were normalized using Procrustes analysis to compare shape and landmark position independent of scale and orientation. The differences in landmark coordinates between the two model types were then statistically analyzed. Results: The analysis revealed that human landmark identification on the AI-generated 3D models reconstructed from 2D photos showed consistently lower variability compared to the gold-standard scans, suggesting that those models may lead to more repeatable measurements. The greatest overall deviations (Euclidean error) were found in landmarks associated with the lateral and inferior borders of the face, such as Right Gonion (RGo), Trichion (Tr), and Left Gonion (LGo). Quantitatively, the mean Procrustes-normalized error across all landmarks was 0.19 ± 0.08 units (range: 0.08−0.45), corresponding approximately to 0.2 mm of proportional deviation when referenced to the 1:1 scale Bellus3D models. While the AI-generated model demonstrated high fidelity, systematic biases were observed mainly as inferior displacements of key landmarks, including Menton (−0.35 units) and Pogonion (−0.30 units), reflecting a consistent downward shift in AI-based reconstruction. Conclusions: The 3DDFA-V2 deep learning application demonstrated in this study is a promising low-cost tool for

generating 3D facial models from 2D photographs with a high degree of precision in landmark placement. While the AI models facilitated more repeatable landmarking, they also exhibited some systematic biases. These findings have clinical implications for the use AI-generated 3D models in diagnosis and treatment planning, highlighting areas where further refinement of the AI algorithm is needed.},

keywords = {3D facial imaging, artificial intelligence, facial landmarks, facial scanner, soft tissue morphology},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyi,

title = {Factors influencing the predictive performance of artificial intelligence for craniofacial growth},

author = {Naeun Kwona and Jong-Hak Kima and Heeyeon Suh and Heesoo Oh and Shin-Jae Lee},

url = {https://angle-orthodontist.kglmeridian.com/view/journals/angl/96/1/article-p106.xml?isSearch=true},

doi = {10.2319/031025-197.1},

year  = {2025},

date = {2025-09-29},

urldate = {2025-09-29},

journal = {Angle Orthodontist},

volume = {96},

issue = {1},

pages = {106-113},

abstract = {Objectives: To evaluate factors influencing the prediction error of artificial intelligence (AI) that predict craniofacial growth and to identify an optimal AI training condition to improve the predictive performance of the AI model. Materials and Methods: Original growth data were collected from the Mathews longitudinal serial growth study. From the original data consisting of 1257 datasets from 33 growing children of northern European descent, 60 data subsets were generated using random resampling procedures to include 12, 18, and 24 subjects, with data sizes of 100, 200, 300, 400, and 500 datasets. The resampling procedures were repeated four times. Each subset was used to train and create a total of 60 AI models. The prediction accuracy of these models was evaluated using growth prediction errors at the lower lip landmark, labrale inferius, as a benchmark indicator. The prediction errors of the 60 AI models were analyzed according to the number of subjects and data sizes. Results: Prediction error decreased as the data size increased. However, increasing the number of subjects within the growth data led to higher prediction errors. Notably, the increase in prediction error caused by adding more subjects was more substantial than the improvement achieved by increasing the data size. Conclusions: The findings suggest that developing highly accurate AI-based craniofacial growth prediction models remains a significant challenge, even with extensive datasets.},

keywords = {artificial intelligence, craniofacial growth, data quantity, individual variability, prediction error},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Leveraging Generative Artificial Intelligence in Teaching, Scholarship and Dental Education: Use Cases and Reflections},

author = {Jonas Bianchi and Meixun Zheng},

url = {https://onlinelibrary.wiley.com/doi/pdf/10.1111/ocr.12949?getft_integrator=scopus&utm_source=scopus},

doi = {10.1111/ocr.12949},

year  = {2025},

date = {2025-05-30},

urldate = {2025-05-30},

journal = {Orthodontics and Craniofacial Research},

pages = {1-8},

abstract = {The objective of this narrative review is to explore the role of generative artificial intelligence (genAI) in dental education, highlighting its emerging applications, potential benefits and implementation challenges. Since the launch of ChatGPT in 2022, genAI tools have gained traction in academic and clinical settings, enabling content generation, translation, summarisation, 

exam preparation and basic clinical planning. This review presents a series of illustrative use cases demonstrating how genAI has been integrated into teaching, research and clinical workflows in dental and orthodontic training. Each example underscores how AI can support faculty in course design, assist students with learning complex concepts and provide real time feedback for exam analysis and academic writing. However, the implementation of genAI is not without limitations. The review addresses common concerns, including misinformation, data privacy, fabricated references and ethical use in clinical contexts. Although the benefits of genAI are promising, this review emphasises the importance of human oversight and institutional policies to guide ethical and effective use. In conclusion, genAI offers valuable support in dental education when used responsibly. Continued dialogue among educators, students and policymakers is essential to ensure that AI tools are integrated thoughtfully 

and equitably into academic practice.},

keywords = {artificial intelligence, dental education, machine learning},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Explainable artificial intelligence to quantify adenoid hypertrophy-related upper airway obstruction using 3D Shape Analysis},

author = {Claudia Trindade Mattos and Lucie Dole and Sergio Luiz Mota-Júnior and Adriana de Alcantara Cury-Saramago and Jonas Bianchi and Heesoo Oh and Karine Evangelista and José Valladares-Neto and Antonio Carlos de Oliveira Ruellas and Juan Carlos Prieto and Lucia Helena Soares Cevidanes },

url = {https://www.sciencedirect.com/science/article/pii/S0300571225001344},

doi = {https://doi.org/10.1016/j.jdent.2025.105689},

issn = {0300-5712},

year  = {2025},

date = {2025-05-01},

journal = {Journal of Dentistry},

volume = {156},

abstract = {Objectives: To develop and validate an explainable Artificial Intelligence (AI) model for classifying and quantifying upper airway obstruction related to adenoid hypertrophy using three-dimensional (3D) shape analysis of cone-beam computed tomography (CBCT) scans. Methods: 400 CBCT scans of patients aged 5–18 years were analyzed. Nasopharyngeal airway obstruction (NAO) ratio was calculated to label scans into four grades of obstruction severity, used as the ground truth. Upper airway surface meshes were used to train a deep learning model combining multiview and point-cloud approaches for 3D shape analysis and obstruction severity classification and quantification. Surface Gradient weighted Class Activation Mapping (SurfGradCAM) generated explainability heatmaps. Performance was evaluated using area under the curve (AUC), precision, recall, F1-score, mean absolute error, root mean squared error, and correlation coefficients. Results: The explainable AI model demonstrated strong performance in both classification and quantification tasks. The AUC values for the classification task ranged from 0.77 to 0.94, with the highest values of 0.88 and 0.94 for Grades 3 and 4, respectively, indicating excellent discriminative ability for identifying more severe cases of obstruction. The SurfGradCAM-generated heatmaps consistently highlighted the most relevant regions of the upper airway influencing the AI’s decision-making process. In the quantification task, the regression model successfully predicted the NAO ratio, with a strong correlation coefficient of 0.854 (p < 0.001) and R2 = 0.728, explaining a substantial proportion of the variance in NAO ratios. Conclusions: The proposed explainable AI model, using 3D shape analysis, demonstrated strong performance in classifying and quantifying adenoid hypertrophy-related upper airway obstruction in CBCT scans. Clinical significance: This AI model provides clinicians with a reliable, automated tool for standardized adenoid hypertrophy assessment. The model’s explainable nature enhances clinical confidence and patient communication, potentially improving diagnostic workflow and treatment planning.},

keywords = {3D shape analysis, Adenoid hypertrophy, artificial intelligence, CBCT, Upper Airway obstruction},

pubstate = {published},

tppubtype = {article}

}

@article{Roseth2025,

title = {Comparison of individualized facial growth prediction models using artificial intelligence and partial least squares based on the Mathews growth collection},

author = {Jeffrey Roseth and Jong-Hak Kim and Jun-Ho Moon and Dong-Yub Ko and Heesoo Oh and Shin-Jae Lee and Heeyeon Suh},

url = {https://angle-orthodontist.kglmeridian.com/view/journals/angl/95/3/article-p249.xml},

doi = {10.2319/082124-687.1},

year  = {2025},

date = {2025-01-31},

journal = {The Angle Orthodontist},

volume = {95},

issue = {3},

pages = {249-258},

abstract = {Objectives: To develop facial growth prediction models using artificial intelligence (AI) under various conditions, and to compare performance of these models with each other as well as with the partial least squares (PLS) growth prediction model. Materials and Methods: Longitudinal lateral cephalograms from 33 subjects in the Mathews growth collection were utilized. A total of 1257 pairs of before and after growth lateral cephalograms were included. In each image, 46 hard and 32 soft tissue landmarks were manually identified. Growth prediction models were constructed using a deep learning method based on TabNet deep neural network and partial least squares (PLS) method. Prediction accuracies of the two methods were compared. Results: On average, artificial intelligence (AI) showed 0.61 mm less prediction error than PLS. Among the 77 predicted landmarks, AI was more accurate than PLS in 60 landmarks. When comparing AI models with varying numbers of training epochs, those with higher epochs yielded more accurate predictions. Overall, PLS and AI exhibited greater prediction errors for soft tissue and mandibular landmarks compared to hard tissue and maxillary landmarks. However, AI showed a smaller increase in prediction error in areas with greater variability. Conclusions: AI proved to be a valuable growth prediction method, with clinically acceptable prediction errors averaging 1.49 mm for 45 hard tissue landmarks and 1.71 mm for 32 soft tissue landmarks. PLS accurately predicted landmarks with low variability. However, AI generally outperformed PLS, particularly for landmarks in the lower part of the craniofacial structure and soft tissue, where uncertainty is considerable. (Angle Orthod. 2025;95:249–258.)},

keywords = {artificial intelligence, Growth prediction, longitudinal craniofacial growth records},

pubstate = {published},

tppubtype = {article}

}

@article{Kim2025,

title = {Craniofacial growth prediction models based on cephalometric landmarks in Korean and American children},

author = {Jong-Hak Kim and Jun-Ho Moon and Jeffrey Roseth and Heeyeon Suh and Heesoo Oh and Shin-Jae Lee},

url = {https://angle-orthodontist.kglmeridian.com/view/journals/angl/95/2/article-p219.xml},

doi = {10.2319/052324-399.1},

year  = {2025},

date = {2025-01-13},

journal = {The Angle Orthodontist},

volume = {95},

issue = {2},

pages = {219-226},

abstract = {Objectives: To compare differences in craniofacial growth prediction results for Korean and American children according to growth prediction models developed using Korean and American longitudinal growth data. Materials and Methods: Growth prediction models based on cephalometric landmarks were built for each population using longitudinally taken lateral cephalograms of Korean children and American children of northern European origin. The sample sizes of the serial datasets were 679 and 1257 for Korean and American children, respectively. On each cephalogram, 78 cephalometric landmarks were identified. The prediction models were based on the partial least squares method with 160 input and 154 output variables. For each group, growth was predicted by applying the prediction models developed using data from the same and different populations. The growth prediction results were compared and analyzed. Results: The growth prediction results obtained with the prediction model developed using data from the same population were more accurate (P<.0001). The results distinctively visualized the discrepancies in the growth prediction results if different population types were not considered. Conclusions: Applying a growth prediction model generated using data from the same population may be desirable. (Angle Orthod. 2025;95:219–226.)},

keywords = {artificial intelligence, Growth prediction, Longitudinal studies, machine learning},

pubstate = {published},

tppubtype = {article}

}

@article{deOliveira2024,

title = {Artificial intelligence as a prediction tool for orthognathic surgery assessment},

author = {Pedro Henrique José de Oliveira and Tengfei Li and Haoyue Li and João Roberto Gonçalves and Ary Santos-Pinto and Luiz Gonzaga Gandini Junior and Lucia Soares Cevidanes and Claudia Toyama and Guilherme Paladini Feltrin and Antonio Augusto Campanha and Melchiades Alves de Oliveira Junior and Jonas Bianchi},

url = {https://doi.org/10.1111/ocr.12805},

doi = {10.1111/ocr.12805},

issn = {1601-6335},

year  = {2024},

date = {2024-04-21},

journal = {Orthodontics & Craniofacial Research},

volume = {27},

issue = {5},

pages = {785-794},

abstract = {Introduction: An ideal orthodontic treatment involves qualitative and quantitative measurements of dental and skeletal components to evaluate patients' discrepancies, such as facial, occlusal, and functional characteristics. Deciding between orthodontics and orthognathic surgery remains challenging, especially in borderline patients. Advances in technology are aiding clinical decisions in orthodontics. The increasing availability of data and the era of big data enable the use of artificial intelligence to guide clinicians' diagnoses. This study aims to test the capacity of different machine learning (ML) models to predict whether orthognathic surgery or orthodontics treatment is required, using soft and hard tissue cephalometric values. Methods: A total of 920 lateral radiographs from patients previously treated with either conventional orthodontics or in combination with orthognathic surgery were used, comprising n = 558 Class II and n = 362 Class III patients, respectively. Thirty-two measures were obtained from each cephalogram at the initial appointment. The subjects were randomly divided into training (n = 552), validation (n = 183), and test (n = 185) datasets, both as an entire sample and divided into Class II and Class III sub-groups. The extracted data were evaluated using 10 machine learning models and by a four-expert panel consisting of orthodontists (n = 2) and surgeons (n = 2). Results: The combined prediction of 10 models showed top-ranked performance in the testing dataset for accuracy, F1-score, and AUC (entire sample: 0.707, 0.706, 0.791; Class II: 0.759, 0.758, 0.824; Class III: 0.822, 0.807, 0.89). Conclusions: The proposed combined 10 ML approach model accurately predicted the need for orthognathic surgery, showing better performance in Class III patients.},

keywords = {artificial intelligence, Class II, Class III, orthodontics, Orthognathic Surgery},

pubstate = {published},

tppubtype = {article}

}

@article{Bianchi2023j,

title = {Interpretable artificial intelligence for classification of alveolar bone defect in patients with cleft lip and palate. },

author = {Miranda F and Choudhari V and Barone S and Anchling L and Hutin N and Gurgel M and et al},

url = {https://doi.org/10.1038/s41598-023-43125-7},

doi = {10.1038/s41598-023-43125-7},

year  = {2023},

date = {2023-09-22},

journal = {Scientific Reports},

volume = {15861},

abstract = {Cleft lip and/or palate (CLP) is the most common congenital craniofacial anomaly and requires bone grafting of the alveolar cleft. This study aimed to develop a novel classification algorithm to assess the severity of alveolar bone defects in patients with CLP using three-dimensional (3D) surface models and to demonstrate through an interpretable artificial intelligence (AI)-based algorithm the decisions provided by the classifier. Cone-beam computed tomography scans of 194 patients with CLP were used to train and test the performance of an automatic classification of the severity of alveolar bone defect. The shape, height, and width of the alveolar bone defect were assessed in automatically segmented maxillary 3D surface models to determine the ground truth classification index of its severity. The novel classifier algorithm renders the 3D surface models from different viewpoints and captures 2D image snapshots fed into a 2D Convolutional Neural Network. An interpretable AI algorithm was developed that uses features from each view and aggregated via Attention Layers to explain the classification. The precision, recall and F-1 score were 0.823, 0.816, and 0.817, respectively, with agreement ranging from 97.4 to 100% on the severity index within 1 group difference. The new classifier and interpretable AI algorithm presented satisfactory accuracy to classify the severity of alveolar bone defect morphology using 3D surface models of patients with CLP and graphically displaying the features that were considered during the deep learning model's classification decision.},

keywords = {3D landmark identification, alveolar bone defect, artificial intelligence, cleft lip, cleft lip and palate},

pubstate = {published},

tppubtype = {article}

}

@article{Bianchi2023g,

title = {Artificial Intelligence Applications in Dentistry},

author = {Bianchi J},

url = {https://doi.org/10.1080/19424396.2023.2204566},

year  = {2023},

date = {2023-05-31},

journal = {Journal of the California Dental Association },

volume = {51},

issue = {1},

keywords = {artificial intelligence, CHAT-GPT, DALL-E AI system, dentistry, OpenAI},

pubstate = {published},

tppubtype = {article}

}

@article{Bianchi2023f,

title = {Artificial intelligence applications in orthodontics. },

author = {Miranda F and Barone S and Gillot M and Baquero B and Anchling L and Hutlin B and et al},

url = {https://doi.org/10.1080/19424396.2023.2195585},

year  = {2023},

date = {2023-04-13},

urldate = {2023-04-13},

journal = {Journal of the California Dental Association },

volume = {51},

issue = {1},

abstract = {Objective

This manuscript describes strategies for assessment of precision of several diagnostic artificial intelligence (AI) tools in orthodontics, available open-source image analysis platforms, as well as the use of three-dimensional (3D) surface models and superimpositions.



Results

The advances described in this manuscript present perspectives on the controversies of whether AI is smarter than clinicians and may replace human clinical decisions. A thorough orthodontic diagnosis requires comprehensive 3D analysis of the interrelationships among the dentition, craniofacial skeleton and soft tissues. Forecasts have indicated that 3D printing technology will provide more than 60% of all dental treatment needs by 2025, and orthodontic companies as well as remote monitoring companies are already using AI technology, being it essential that the clinicians are prepared and knowledgeable with the technology advances now available.



Conclusions

The AI applications in orthodontics rely on the implementation into diagnostic image records, data analysis for clinical practice and research applications. Continuous training and validation of the AI orthodontic image tools are essential for improving the performance and generalizability of these methods.},

keywords = {artificial intelligence, imaging, orthodontics, three-dimensional},

pubstate = {published},

tppubtype = {article}

}

@article{Bianchi2022d,

title = {Quantitative bone imaging biomarkers and joint space analysis of the articular Fossa in temporomandibular joint osteoarthritis using artificial intelligence models},

author = {Mackie T and Al Turkestani N and Bianchi J and Li T and Ruellas A and Gurgel M and Benavides E and Soki F and Cevidanes L},

url = {https://www.frontiersin.org/articles/10.3389/fdmed.2022.1007011/full},

doi = {https://doi.org/10.3389/fdmed.2022.1007011},

year  = {2022},

date = {2022-09-19},

journal = {Frontiers in Dental Medicine},

abstract = {Temporomandibular joint osteoarthritis (TMJ OA) is a disease with a multifactorial etiology, involving many pathophysiological processes, and requiring comprehensive assessments to characterize progressive cartilage degradation, subchondral bone remodeling, and chronic pain. This study aimed to integrate quantitative biomarkers of bone texture and morphometry of the articular fossa and joint space to advance the role of imaging phenotypes for diagnosis of Temporomandibular Joint Osteoarthritis (TMJ OA) in early to moderate stages by improving the performance of machine-learning algorithms to detect TMJ OA status. Ninety-two patients were prospectively enrolled (184 h-CBCT scans of the right and left mandibular condyles), divided into two groups: 46 control and 46 TMJ OA subjects. No significant difference in the articular fossa radiomic biomarkers was found between TMJ OA and control patients. The superior condyle-to-fossa distance (p < 0.05) was significantly smaller in diseased patients. The interaction effects of the articular fossa radiomic biomarkers enhanced the performance of machine-learning algorithms to detect TMJ OA status. The LightGBM model achieved an AUC 0.842 to diagnose the TMJ OA status with Headaches and Range of Mouth Opening Without Pain ranked as top features, and top interactions of VE-cadherin in Serum and Angiogenin in Saliva, TGF-β1 in Saliva and Headaches, Gender and Muscle Soreness, PA1 in Saliva and Range of Mouth Opening Without Pain, Lateral Condyle Grey Level Non-Uniformity and Lateral Fossa Short Run Emphasis, TGF-β1 in Serum and Lateral Fossa Trabeculae number, MMP3 in Serum and VEGF in Serum, Headaches and Lateral Fossa Trabecular spacing, Headaches and PA1 in Saliva, and Headaches and BDNF in Saliva. Our preliminary results indicate that condyle imaging features may be more important in regards to main effects, but the fossa imaging features may have a larger contribution in terms of interaction effects. More studies are needed to optimize and further enhance machine-learning algorithms to detect early markers of disease, improve prediction of disease progression and severity to ultimately better serve clinical decision support systems in the treatment of patients with TMJ OA.},

keywords = {articular fossa, artificial intelligence, hr-CBCT, imaging biomarkers, joint space, temporomandibular osteoarthritis},

pubstate = {published},

tppubtype = {article}

}