AIM: This systematic review aimed to investigate the performance of AI systems in identifying dental anomalies in paediatric dentistry and compare it with human performance.
DESIGN: A systematic search of Scopus, PubMed and Google Scholar was conducted from 2012 to 2022. Inclusion criteria were based on problem/patient/population, intervention/indicator, comparison and outcome scheme and specific keywords related to AI, DL, paediatric dentistry, dental anomalies, supernumerary and mesiodens. Six of 3918 initial pool articles were included, assessing nine DL sub-systems that used panoramic radiographs or cone-beam computed tomography. Article quality was assessed using QUADAS-2.
RESULTS: Artificial intelligence systems based on DL algorithms showed promising potential in enhancing the speed and accuracy of dental anomaly detection, with an average of 85.38% accuracy and 86.61% sensitivity. Human performance, however, outperformed AI systems, achieving 95% accuracy and 99% sensitivity. Limitations included a limited number of articles and data heterogeneity.
CONCLUSION: The potential of AI systems employing DL algorithms is highlighted in detecting dental anomalies in paediatric dentistry. Further research is needed to address limitations, explore additional anomalies and establish the broader applicability of AI in paediatric dentistry.
METHODS: Following a registered protocol, a modified e-Delphi study was applied over two rounds with a final consensus meeting. The threshold of consensus was set a priori at 75%. Agreed techniques were then categorized by four coders, according to behavioural learning theory, to sort techniques according to their mechanism of action.
RESULTS: The panel (n = 35) agreed on 42 DBS techniques from a total of 63 candidate labels and descriptions. Complete agreement was achieved regarding all labels and descriptions, while agreement was not achieved regarding distinctiveness for 17 techniques. In exploring underlying principles of learning, it became clear that multiple and differing principles may apply depending on the specific context and procedure in which the technique may be applied.
DISCUSSION: Experts agreed on what each DBS technique is, what label to use, and their description, but were less likely to agree on what distinguishes one technique from another. All techniques were describable but not comprehensively categorizable according to principles of learning. While objective consistency was not attained, greater clarity and consistency now exists. The resulting list of agreed terminology marks a significant foundation for future efforts towards understanding DBS techniques in research, education and clinical care.