Accuracy of artificial intelligence in the detection and segmentation of oral and maxillofacial structures using cone-beam computed tomography images: a systematic review and meta-analysis

Abstract

PurposeThe aim of the present systematic review and meta-analysis was to resolve the conflicts on the diagnostic accuracy of artificial intelligence systems in detecting and segmenting oral and maxillofacial structures using cone-beam computed tomography (CBCT) images.Material and methodsWe performed a literature search of the Embase, PubMed, and Scopus databases for reports published from their inception to 31 October 2022. We included studies that explored the accuracy of artificial intelligence in the automatic detection or segmentation of oral and maxillofacial anatomical landmarks or lesions using CBCT images. The extracted data were pooled, and the estimates were presented with 95% confidence intervals (CIs).ResultsIn total, 19 eligible studies were identified. As per the analysis, the overall pooled diagnostic accuracy of artificial intelligence was 0.93 (95% CI: 0.91-0.94). This rate was 0.93 (95% CI: 0.89-0.96) for anatomical landmarks based on 7 studies and 0.92 (95% CI: 0.90-0.94) for lesions according to 12 reports. Moreover, the pooled accuracy of detection and segmentation tasks for artificial intelligence was 0.93 (95% CI: 0.91-0.94) and 0.92 (95% CI: 0.85-0.95) based on 14 and 5 surveys, respectively.ConclusionsExcellent accuracy was observed for the detection and segmentation objectives of artificial intelligence using oral and maxillofacial CBCT images. These systems have the potential to streamline oral and dental healthcare services.