Scholarlink

Background <p>Artificial intelligence (AI) has increasingly permeated otorhinolaryngology (ENT), offering novel tools for diagnosis, surgical planning, and rehabilitation. Despite rapid technological advances, a comprehensive synthesis of AI applications across the subspecialties of otology, rhinology, laryngology, and head and neck oncology—with rigorous methodological appraisal—remains lacking. This scoping review aims to systematically map the extent, range, and nature of recent AI-driven innovations, evaluate their clinical utility, and identify prevailing barriers to implementation within otorhinolaryngology.</p> Methods <p>Following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines, a comprehensive literature search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore for articles published between January 2018 and March 2025. Search terms combined “artificial intelligence,” “machine learning,” and “deep learning” with ENT-specific subspecialty terms. Inclusion criteria encompassed original research and clinical trials addressing AI applications in ENT. Two independent reviewers screened titles, abstracts, and full texts. Studies were charted using a structured data extraction framework assessing study design, dataset characteristics, AI methodology, validation strategy, and performance metrics.</p> Results <p>From an initial yield of 342 records after duplicate removal, 51 studies met the inclusion criteria. In otologic and audiologic domains, deep learning algorithms demonstrated high accuracy (&gt; 90%) in automated otoscopic image classification and audiometric pattern recognition. Rhinologic and laryngologic applications included AI-enhanced imaging for chronic rhinosinusitis severity scoring and voice disorder classification. In head and neck oncology, AI contributed to early tumor detection through radiomics and optimized cervical nodal management. Critical appraisal revealed that the majority of studies (&gt; 75%) remain retrospective and single-center, with only approximately 15–25% reporting external validation. Prospective clinical utility assessment was rare (&lt; 10%).</p> Conclusion <p>AI holds substantial promise to augment diagnostic precision and surgical decision-making in otorhinolaryngology. However, the current evidence base is largely early-stage and proof-of-concept. Addressing the “AI chasm” between algorithmic performance and clinical deployment requires a paradigm shift toward multicenter prospective validation, standardized reporting, and rigorous evaluation of real-world clinical utility.</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

Artificial intelligence in otorhinolaryngology: a scoping review of diagnostic, surgical, and rehabilitative applications

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月3日

查看全文
  • Wan-Ling Lin,
  • Sheng-Han Chen,
  • Shih-Shuan Fang

摘要

Background

Artificial intelligence (AI) has increasingly permeated otorhinolaryngology (ENT), offering novel tools for diagnosis, surgical planning, and rehabilitation. Despite rapid technological advances, a comprehensive synthesis of AI applications across the subspecialties of otology, rhinology, laryngology, and head and neck oncology—with rigorous methodological appraisal—remains lacking. This scoping review aims to systematically map the extent, range, and nature of recent AI-driven innovations, evaluate their clinical utility, and identify prevailing barriers to implementation within otorhinolaryngology.

Methods

Following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines, a comprehensive literature search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore for articles published between January 2018 and March 2025. Search terms combined “artificial intelligence,” “machine learning,” and “deep learning” with ENT-specific subspecialty terms. Inclusion criteria encompassed original research and clinical trials addressing AI applications in ENT. Two independent reviewers screened titles, abstracts, and full texts. Studies were charted using a structured data extraction framework assessing study design, dataset characteristics, AI methodology, validation strategy, and performance metrics.

Results

From an initial yield of 342 records after duplicate removal, 51 studies met the inclusion criteria. In otologic and audiologic domains, deep learning algorithms demonstrated high accuracy (> 90%) in automated otoscopic image classification and audiometric pattern recognition. Rhinologic and laryngologic applications included AI-enhanced imaging for chronic rhinosinusitis severity scoring and voice disorder classification. In head and neck oncology, AI contributed to early tumor detection through radiomics and optimized cervical nodal management. Critical appraisal revealed that the majority of studies (> 75%) remain retrospective and single-center, with only approximately 15–25% reporting external validation. Prospective clinical utility assessment was rare (< 10%).

Conclusion

AI holds substantial promise to augment diagnostic precision and surgical decision-making in otorhinolaryngology. However, the current evidence base is largely early-stage and proof-of-concept. Addressing the “AI chasm” between algorithmic performance and clinical deployment requires a paradigm shift toward multicenter prospective validation, standardized reporting, and rigorous evaluation of real-world clinical utility.