Academic Clinical Fellow in Oral Surgery, Oral and Maxillofacial Department, Royal Devon and Exeter Hospital, Barrack Road, Exeter, EX2 5DW, UK.
Qatar University, College of Dental Medicine, QU Health, Doha 2713, Qatar.
Evid Based Dent. 2022 Mar;23(1):12-13. doi: 10.1038/s41432-022-0238-y. Epub 2022 Mar 25.
Data Sources Electronic search on PubMed, Cochrane, Scopus, Embase, Google Scholar, Saudi Digital Library and Web of Science, and hand searching carried out for studies published January 2000-March 2021. Language was restricted to English.Study selection Original research studies involving artificial intelligence technology for oral cancer diagnosis and prognosis prediction were considered. The studies had to provide quantitative data of their evaluation analysis. The exclusion criteria were reported. No limit was set on study design.Data extraction and synthesis The initial search yielded 628 articles. Following deduplication, 340 full-text articles were screened. QUADAS-2 tool was used to assess the quality of the included studies regarding diagnostic accuracy.Results A total of 16 studies were included with various study designs: 14 cross-sectional, one cohort and one retrospective study. Six studies reviewed the diagnosis aspect. All studies indicate an overall positive trend of artificial intelligence technology.Conclusions Artificial intelligence appears to have good accuracy in oral cancer diagnosis and its prediction.
在 PubMed、Cochrane、Scopus、Embase、Google Scholar、沙特数字图书馆和 Web of Science 上进行电子检索，并对 2000 年 1 月至 2021 年 3 月发表的研究进行手工检索。语言仅限于英语。
考虑了涉及人工智能技术的口腔癌诊断和预后预测的原始研究。这些研究必须提供其评估分析的定量数据。排除标准有报道。对研究设计没有限制。
最初的搜索产生了 628 篇文章。经过去重，筛选了 340 篇全文。使用 QUADAS-2 工具评估了纳入研究的诊断准确性的质量。
共纳入 16 项研究，研究设计各不相同：14 项横断面研究、1 项队列研究和 1 项回顾性研究。六项研究回顾了诊断方面。所有研究均表明人工智能技术具有总体积极趋势。
人工智能似乎在口腔癌诊断及其预测方面具有良好的准确性。
