Scholarlink

Purpose <p>To conduct a scoping review to assess the extent and type of evidence on the use of IRT as a diagnostic tool for thyroid nodule detection and classification.</p> Methods <p>This review followed the JBI Scoping Review methodology and included studies involving IRT assessments of the thyroid gland in patients from clinical, academic, or experimental settings. Electronic searches were conducted in PubMed, Embase, Web of Science, Scopus, and Google Scholar.</p> Results <p>A total of 28 studies published between 2009 and 2025 were included after screening 2,687 records. Most studies used static IRT (SIRT) or dynamic IRT (DIRT), many of them integrating artificial intelligence (AI) tools. The assessed studies demonstrated temperature differences between malignant nodules, benign nodules, and healthy tissues, particularly under thermal stress. AI classifiers, especially convolutional neural networks (CNN), have demonstrated enhanced diagnostic performance, with reported accuracies reaching up to 98.4%. However, results varied due to differences in imaging protocols, camera resolution, and patient-specific factors such as subcutaneous fat thickness.</p> Conclusion <p>IRT shows promising potential as an adjunctive diagnostic tool for assessing thyroid nodules, particularly when combined with AI and dynamic protocols. However, standardization and further validation in diverse populations are needed.</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

The role of infrared thermography in thyroid nodule detection and classification: a scoping review

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月1日

查看全文
  • Diego Filipe Bezerra Silva,
  • Diana Lance,
  • Elis Ângela Batistella,
  • Scott Adams,
  • Cassiano Francisco Weege Nonaka,
  • Daniela Pita de Melo

摘要

Purpose

To conduct a scoping review to assess the extent and type of evidence on the use of IRT as a diagnostic tool for thyroid nodule detection and classification.

Methods

This review followed the JBI Scoping Review methodology and included studies involving IRT assessments of the thyroid gland in patients from clinical, academic, or experimental settings. Electronic searches were conducted in PubMed, Embase, Web of Science, Scopus, and Google Scholar.

Results

A total of 28 studies published between 2009 and 2025 were included after screening 2,687 records. Most studies used static IRT (SIRT) or dynamic IRT (DIRT), many of them integrating artificial intelligence (AI) tools. The assessed studies demonstrated temperature differences between malignant nodules, benign nodules, and healthy tissues, particularly under thermal stress. AI classifiers, especially convolutional neural networks (CNN), have demonstrated enhanced diagnostic performance, with reported accuracies reaching up to 98.4%. However, results varied due to differences in imaging protocols, camera resolution, and patient-specific factors such as subcutaneous fat thickness.

Conclusion

IRT shows promising potential as an adjunctive diagnostic tool for assessing thyroid nodules, particularly when combined with AI and dynamic protocols. However, standardization and further validation in diverse populations are needed.