Scholarlink

Introduction <p>Artificial Intelligence (AI) is a rapidly evolving technology, with various applications in dentistry including diagnosis, treatment planning, and prognosis. There are various AI-based applications for dental practitioners, however, their real-world evaluation through deployement studies is scarce, as most of the studies are validation studies. This review explores the potential pitfalls of focusing solely on technical performance metrics when evaluating AI-based applications in dentistry while overlooking the importance of clinical applicability.</p> Methods <p>An electronic search was performed on PubMed and Scopus while a manual search was conducted on Google Scholar&#xa0;“Dentistry”, “Dental”, “Artificial Intelligence”, “Deep Learning, “Machine Learning”, “Applications”, “Diagnocat”, “CephX”, “Denti.AI”, “VideaAI”, “Smile Designer”, “Overjet”, “DentalXR.AI”, “Smilo.AI”, “Smile.AI”, “Pearl”, “AI deployment challenges in dental practice”, “AI for treatment planning in dentistry”, “AI in dental imaging”, and “AI-based diagnosis in dentistry”.</p> Results <p>The electronic search yielded a total of 34 studies, while 10 additional studies were obtained through a manual search, resulting in a total of 44 studies included in this review.&#xa0;&#xa0;Among the 44 studies analyzed, 26 studies were retrospective, while 7 studies utilized a comparative design. The remaining studies comprised of 3 observational, 5 validation, 2 cross-sectional, and 1 prospective study.&#xa0;Further to evaluate the identified applications, relevant companies were contacted via email. Only one company’s representative responded, offering a limited trial version which was insufficient for evaluating the application’s effectiveness. AI technologies may offer lots of benefits for dental practice by enhancing patient-health-based outcomes, however, real-world applications are necessary to ensure its safety.</p> Conclusion <p>This work highlights the need for conducting deployment studies for such AI-based dental applications to translate and implement them into dental practice. Collaboration with stakeholders and dental practitioners to assess the use of such applications is of paramount importance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Concerns regarding deployment of AI-based applications in dentistry – a review

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月25日

查看全文
  • Abhishek Lal,
  • Ayesha Nooruddin,
  • Fahad Umer

摘要

Introduction

Artificial Intelligence (AI) is a rapidly evolving technology, with various applications in dentistry including diagnosis, treatment planning, and prognosis. There are various AI-based applications for dental practitioners, however, their real-world evaluation through deployement studies is scarce, as most of the studies are validation studies. This review explores the potential pitfalls of focusing solely on technical performance metrics when evaluating AI-based applications in dentistry while overlooking the importance of clinical applicability.

Methods

An electronic search was performed on PubMed and Scopus while a manual search was conducted on Google Scholar “Dentistry”, “Dental”, “Artificial Intelligence”, “Deep Learning, “Machine Learning”, “Applications”, “Diagnocat”, “CephX”, “Denti.AI”, “VideaAI”, “Smile Designer”, “Overjet”, “DentalXR.AI”, “Smilo.AI”, “Smile.AI”, “Pearl”, “AI deployment challenges in dental practice”, “AI for treatment planning in dentistry”, “AI in dental imaging”, and “AI-based diagnosis in dentistry”.

Results

The electronic search yielded a total of 34 studies, while 10 additional studies were obtained through a manual search, resulting in a total of 44 studies included in this review.  Among the 44 studies analyzed, 26 studies were retrospective, while 7 studies utilized a comparative design. The remaining studies comprised of 3 observational, 5 validation, 2 cross-sectional, and 1 prospective study. Further to evaluate the identified applications, relevant companies were contacted via email. Only one company’s representative responded, offering a limited trial version which was insufficient for evaluating the application’s effectiveness. AI technologies may offer lots of benefits for dental practice by enhancing patient-health-based outcomes, however, real-world applications are necessary to ensure its safety.

Conclusion

This work highlights the need for conducting deployment studies for such AI-based dental applications to translate and implement them into dental practice. Collaboration with stakeholders and dental practitioners to assess the use of such applications is of paramount importance.