To compare the skeletal, dentoalveolar, soft-tissue, and treatment-burden effects of skeletal anterior traction (SAT) versus fixed anterior traction (FAT) in growing patients with skeletal Class III malocclusion. PubMed/MEDLINE, Embase, Scopus, Web of Science, and CENTRAL were searched from inception through 15 May 2021, with citation tracking through 30 June 2021. Randomised and non-randomised comparative studies directly evaluating SAT versus FAT in growing Class III patients were eligible. Two reviewers independently screened records, extracted data, and assessed risk of bias with RoB 2 or ROBINS-I as appropriate; certainty was interpreted with GRADE. Random-effects meta-analysis was used for compatible continuous outcomes. The quantitative dataset comprised six comparative studies. In that dataset, SAT was associated with greater maxillary advancement (MD +1.23, 95% CI 0.88 to 1.57), greater intermaxillary sagittal correction (MD +1.38, 95% CI 1.04 to 1.73), less forward mandibular positional change (MD -0.94, 95% CI -1.38 to -0.49), less clockwise mandibular rotation (MD -1.12, 95% CI -1.79 to -0.45), less maxillary-incisor proclination (MD -3.12, 95% CI -4.05 to -2.18), and improved facial convexity (MD +1.42, 95% CI 0.96 to 1.89). Publication-level reconciliation identified two additional pre-2021 direct comparisons already cited in the manuscript and six direct comparative reports published after the original database-search boundary; these were not retrospectively mixed into the retained pooled estimates without a reproducible re-extraction of the original dataset. The retained pooled dataset suggests that SAT shifts early Class III correction toward a more skeletal and less dentoalveolar pattern, but the exact magnitude of the skeletal advantage is uncertain. RCT-only syntheses published subsequently report only small incremental maxillary effects of skeletal anchorage, while the reduction in maxillary-incisor proclination is more consistent. The pooled estimates in this manuscript should therefore be treated as provisional until a complete multi-database update and re-extraction incorporate all eligible direct comparisons.

Artificial intelligence (AI) is increasingly being integrated into orthodontic research and clinical practice. This review paper consists of three parts and aims to establish a foundational understanding of AI concepts for readers from non-technical backgrounds while familiarising them with its applications and challenges within the field of orthodontics. The first part provides a comprehensive introduction to the basic concepts of AI, including definitions of machine learning and deep learning models, alongside various learning types: Supervised, unsupervised, and reinforcement learning. To explain how learning occurs and how to optimise it during model training, key features of the "learning" process are introduced. A detailed overview of supervised learning is presented, including data preparation, model selection, and performance evaluation. Following this introduction, different architectures and algorithms of machine learning and neural networks are examined, with examples of their applications in orthodontics, including benefits for clinical decision-making, image processing, and patient monitoring. Finally, this review will consider innovations in AI that hold promise for enhancing orthodontic practice, such as the use of synthetic data, transformers, generative models, and large language models.

Artificial intelligence (AI) is increasingly affecting orthodontic practice and research, yet its adoption requires a careful understanding of both its capabilities and its limitations. As the final part of a three-section review, this paper focuses on the practical cautions and constraints that accompany current AI developments in orthodontics. Topics such as data quality, educational preparedness and the ethical and legal responsibilities surrounding AI use are discussed as essential considerations for preventing errors, misuse and misinterpretation. Technical challenges, including model generalizability, transparency and interpretability, are elucidated in relation to their impact on clinical applicability. Finally, new trends and future directions in the evolution of AI within orthodontics and other fields are described. Together, these perspectives aim to provide a realistic and balanced perspective of how AI can be integrated responsibly into orthodontic care and research.