Self-Attention Mechanisms-Based Laryngoscopy Image Classification Technique for Laryngeal Cancer Detection.

IF 2.3 3区医学 Q1 OTORHINOLARYNGOLOGY Head and Neck-Journal for the Sciences and Specialties of the Head and Neck Pub Date : 2024-11-11 DOI:10.1002/hed.27999

Yi-Fan Kang, Lie Yang, Yi-Fan Hu, Kai Xu, Lan-Jun Cai, Bin-Bin Hu, Xiang Lu

{"title":"Self-Attention Mechanisms-Based Laryngoscopy Image Classification Technique for Laryngeal Cancer Detection.","authors":"Yi-Fan Kang, Lie Yang, Yi-Fan Hu, Kai Xu, Lan-Jun Cai, Bin-Bin Hu, Xiang Lu","doi":"10.1002/hed.27999","DOIUrl":null,"url":null,"abstract":"Background: The early diagnosis of laryngeal cancer (LCA) is crucial for prognosis, driving our search for an accurate, precise, and sensitive deep learning model to assist in LCA detection.Methods: We collected 5768 laryngoscopic images from 1462 patients and created the intelligent laryngeal cancer detection system (ILCDS) based on Swin-Transformer. Following training and validation, we assessed the ILCDS performance on the internal and external test sets and compared it with previous convolutional neural network (CNN) models and three professional laryngologists.Results: The ILCDS outperformed the six CNNs, with the highest accuracy of 92.78% and an area under the curve (AUC) of 0.9732. Despite a slight drop in performance on external sets, the ILCDS maintained the best superiority, with 85.79% accuracy and an AUC of 0.9550. Surpassing professional laryngologists, the ILCDS achieved 92.00% accuracy.Conclusions: The ILCDS offers high accuracy and stability for LCA detection, reducing the burden on laryngologists.","PeriodicalId":55072,"journal":{"name":"Head and Neck-Journal for the Sciences and Specialties of the Head and Neck","volume":" ","pages":""},"PeriodicalIF":2.3000,"publicationDate":"2024-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Head and Neck-Journal for the Sciences and Specialties of the Head and Neck","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1002/hed.27999","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"OTORHINOLARYNGOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: The early diagnosis of laryngeal cancer (LCA) is crucial for prognosis, driving our search for an accurate, precise, and sensitive deep learning model to assist in LCA detection.

Methods: We collected 5768 laryngoscopic images from 1462 patients and created the intelligent laryngeal cancer detection system (ILCDS) based on Swin-Transformer. Following training and validation, we assessed the ILCDS performance on the internal and external test sets and compared it with previous convolutional neural network (CNN) models and three professional laryngologists.

Results: The ILCDS outperformed the six CNNs, with the highest accuracy of 92.78% and an area under the curve (AUC) of 0.9732. Despite a slight drop in performance on external sets, the ILCDS maintained the best superiority, with 85.79% accuracy and an AUC of 0.9550. Surpassing professional laryngologists, the ILCDS achieved 92.00% accuracy.

Conclusions: The ILCDS offers high accuracy and stability for LCA detection, reducing the burden on laryngologists.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

基于自留机制的喉镜图像分类技术用于喉癌检测

背景：喉癌（LCA）的早期诊断对预后至关重要：喉癌（LCA）的早期诊断对预后至关重要，这促使我们寻找一种准确、精确、灵敏的深度学习模型来辅助喉癌检测：我们收集了来自 1462 名患者的 5768 张喉镜图像，并基于 Swin-Transformer 创建了智能喉癌检测系统（ILCDS）。经过训练和验证后，我们评估了智能喉癌检测系统在内部和外部测试集上的表现，并将其与之前的卷积神经网络（CNN）模型和三位专业喉科专家进行了比较：ILCDS 的表现优于六种 CNN，最高准确率为 92.78%，曲线下面积 (AUC) 为 0.9732。尽管在外部集上的表现略有下降，但 ILCDS 仍保持了最佳优势，准确率为 85.79%，AUC 为 0.9550。ILCDS 的准确率达到 92.00%，超过了专业喉科医师：ILCDS为LCA检测提供了高准确性和稳定性，减轻了喉科医师的负担。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Head and Neck-Journal for the Sciences and Specialties of the Head and Neck 医学-耳鼻喉科学

CiteScore

7.00

自引率

6.90%

发文量

278

审稿时长

1.6 months

期刊介绍： Head & Neck is an international multidisciplinary publication of original contributions concerning the diagnosis and management of diseases of the head and neck. This area involves the overlapping interests and expertise of several surgical and medical specialties, including general surgery, neurosurgery, otolaryngology, plastic surgery, oral surgery, dermatology, ophthalmology, pathology, radiotherapy, medical oncology, and the corresponding basic sciences.