可避免的活检?验证基于人工智能的决策支持软件在不确定甲状腺结节中的应用。

IF 3.2 2区 医学 Q1 SURGERY
Surgery Pub Date : 2024-10-12 DOI:10.1016/j.surg.2024.07.074
Christopher J Carnabatu, David T Fetzer, Alexander Tessnow, Shelby Holt, Vivek R Sant
{"title":"可避免的活检?验证基于人工智能的决策支持软件在不确定甲状腺结节中的应用。","authors":"Christopher J Carnabatu, David T Fetzer, Alexander Tessnow, Shelby Holt, Vivek R Sant","doi":"10.1016/j.surg.2024.07.074","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Multiple artificial intelligence (AI) systems have been approved to risk-stratify thyroid nodules through sonographic characterization. We sought to validate the ability of one such AI system, Koios DS (Koios Medical, Chicago, IL), to aid in improving risk stratification of indeterminate thyroid nodules.</p><p><strong>Methods: </strong>A retrospective single-institution dataset was compiled of 28 cytologically indeterminate thyroid nodules having undergone molecular testing and surgical resection, with surgical pathology categorized as malignant or benign. Nodules were retrospectively evaluated with Koios DS. After nodule selection, automated and AI-adapter-derived Thyroid Imaging Reporting and Data System (TI-RADS) levels were recorded, and agreement with radiologist-derived levels was assessed using Cohen's κ statistic. The performance of malignancy classification was compared between the radiologist and AI-adapter. Biopsy thresholds were re-evaluated using the AI-adapter.</p><p><strong>Results: </strong>In this cohort, 7 (25%) nodules were malignant on surgical pathology. The median nodule size was 2.4 cm (interquartile range: 1.8-2.9 cm). Median radiologist and automated TI-RADS levels were both 4, with κ 0.25 (\"fair agreement\"). Malignancy classification by the radiologist provided sensitivity 100%, specificity 33.3%, positive predictive value (PPV) 33.3%, and negative predictive value (NPV) 100%, compared with the AI-adapter's performance with sensitivity 85.7%, specificity 76.2%, PPV 54.5%, and NPV 94.1%. Using the AI-adapter, 14 of 28 biopsies would have been deferred, 13 of which were surgically benign.</p><p><strong>Conclusion: </strong>Koios automated and radiologist-derived TI-RADS levels were in consistent agreement for indeterminate thyroid nodules. Malignancy reclassification with the AI-adapter improved PPV at minimal cost to NPV. Risk stratification with the addition of the AI-adapter may allow for more accurate patient counseling and the avoidance of biopsies in select cases that would otherwise be cytologically indeterminate.</p>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":" ","pages":""},"PeriodicalIF":3.2000,"publicationDate":"2024-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Avoidable biopsies? Validating artificial intelligence-based decision support software in indeterminate thyroid nodules.\",\"authors\":\"Christopher J Carnabatu, David T Fetzer, Alexander Tessnow, Shelby Holt, Vivek R Sant\",\"doi\":\"10.1016/j.surg.2024.07.074\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Multiple artificial intelligence (AI) systems have been approved to risk-stratify thyroid nodules through sonographic characterization. We sought to validate the ability of one such AI system, Koios DS (Koios Medical, Chicago, IL), to aid in improving risk stratification of indeterminate thyroid nodules.</p><p><strong>Methods: </strong>A retrospective single-institution dataset was compiled of 28 cytologically indeterminate thyroid nodules having undergone molecular testing and surgical resection, with surgical pathology categorized as malignant or benign. Nodules were retrospectively evaluated with Koios DS. After nodule selection, automated and AI-adapter-derived Thyroid Imaging Reporting and Data System (TI-RADS) levels were recorded, and agreement with radiologist-derived levels was assessed using Cohen's κ statistic. The performance of malignancy classification was compared between the radiologist and AI-adapter. Biopsy thresholds were re-evaluated using the AI-adapter.</p><p><strong>Results: </strong>In this cohort, 7 (25%) nodules were malignant on surgical pathology. The median nodule size was 2.4 cm (interquartile range: 1.8-2.9 cm). Median radiologist and automated TI-RADS levels were both 4, with κ 0.25 (\\\"fair agreement\\\"). Malignancy classification by the radiologist provided sensitivity 100%, specificity 33.3%, positive predictive value (PPV) 33.3%, and negative predictive value (NPV) 100%, compared with the AI-adapter's performance with sensitivity 85.7%, specificity 76.2%, PPV 54.5%, and NPV 94.1%. Using the AI-adapter, 14 of 28 biopsies would have been deferred, 13 of which were surgically benign.</p><p><strong>Conclusion: </strong>Koios automated and radiologist-derived TI-RADS levels were in consistent agreement for indeterminate thyroid nodules. Malignancy reclassification with the AI-adapter improved PPV at minimal cost to NPV. Risk stratification with the addition of the AI-adapter may allow for more accurate patient counseling and the avoidance of biopsies in select cases that would otherwise be cytologically indeterminate.</p>\",\"PeriodicalId\":22152,\"journal\":{\"name\":\"Surgery\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":3.2000,\"publicationDate\":\"2024-10-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Surgery\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1016/j.surg.2024.07.074\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"SURGERY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Surgery","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1016/j.surg.2024.07.074","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"SURGERY","Score":null,"Total":0}
引用次数: 0

摘要

背景:多种人工智能(AI)系统已被批准用于通过声学特征描述对甲状腺结节进行风险分层。我们试图验证 Koios DS(Koios Medical,芝加哥,伊利诺伊州)这一人工智能系统在帮助改善不确定甲状腺结节风险分层方面的能力:方法:对28个细胞学不确定的甲状腺结节进行了分子检测和手术切除,手术病理分为恶性和良性。使用 Koios DS 对结节进行了回顾性评估。选择结节后,记录自动和人工智能适配器得出的甲状腺成像报告和数据系统(TI-RADS)级别,并使用 Cohen's κ 统计量评估与放射科医生得出的级别的一致性。比较了放射科医生和 AI-adapter 的恶性肿瘤分类效果。使用人工智能适配器重新评估了活检阈值:结果:在这批患者中,7 个(25%)结节经手术病理检查为恶性。结节大小的中位数为 2.4 厘米(四分位间范围:1.8-2.9 厘米)。放射科医生和自动 TI-RADS 分级的中位数均为 4,κ为 0.25("相当一致")。放射科医生的恶性肿瘤分类灵敏度为 100%,特异性为 33.3%,阳性预测值 (PPV) 为 33.3%,阴性预测值 (NPV) 为 100%,而人工智能适配器的灵敏度为 85.7%,特异性为 76.2%,阳性预测值 (PPV) 为 54.5%,阴性预测值 (NPV) 为 94.1%。如果使用人工智能适配器,28 例活检中有 14 例会被推迟,其中 13 例是良性的:结论:对于不确定的甲状腺结节,Koios自动和放射科医生得出的TI-RADS水平一致。使用 AI 适配器进行恶性肿瘤再分类提高了 PPV,但对 NPV 的影响最小。使用人工智能适配器进行风险分层可为患者提供更准确的咨询,并避免对细胞学未确定的特定病例进行活检。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Avoidable biopsies? Validating artificial intelligence-based decision support software in indeterminate thyroid nodules.

Background: Multiple artificial intelligence (AI) systems have been approved to risk-stratify thyroid nodules through sonographic characterization. We sought to validate the ability of one such AI system, Koios DS (Koios Medical, Chicago, IL), to aid in improving risk stratification of indeterminate thyroid nodules.

Methods: A retrospective single-institution dataset was compiled of 28 cytologically indeterminate thyroid nodules having undergone molecular testing and surgical resection, with surgical pathology categorized as malignant or benign. Nodules were retrospectively evaluated with Koios DS. After nodule selection, automated and AI-adapter-derived Thyroid Imaging Reporting and Data System (TI-RADS) levels were recorded, and agreement with radiologist-derived levels was assessed using Cohen's κ statistic. The performance of malignancy classification was compared between the radiologist and AI-adapter. Biopsy thresholds were re-evaluated using the AI-adapter.

Results: In this cohort, 7 (25%) nodules were malignant on surgical pathology. The median nodule size was 2.4 cm (interquartile range: 1.8-2.9 cm). Median radiologist and automated TI-RADS levels were both 4, with κ 0.25 ("fair agreement"). Malignancy classification by the radiologist provided sensitivity 100%, specificity 33.3%, positive predictive value (PPV) 33.3%, and negative predictive value (NPV) 100%, compared with the AI-adapter's performance with sensitivity 85.7%, specificity 76.2%, PPV 54.5%, and NPV 94.1%. Using the AI-adapter, 14 of 28 biopsies would have been deferred, 13 of which were surgically benign.

Conclusion: Koios automated and radiologist-derived TI-RADS levels were in consistent agreement for indeterminate thyroid nodules. Malignancy reclassification with the AI-adapter improved PPV at minimal cost to NPV. Risk stratification with the addition of the AI-adapter may allow for more accurate patient counseling and the avoidance of biopsies in select cases that would otherwise be cytologically indeterminate.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Surgery
Surgery 医学-外科
CiteScore
5.40
自引率
5.30%
发文量
687
审稿时长
64 days
期刊介绍: For 66 years, Surgery has published practical, authoritative information about procedures, clinical advances, and major trends shaping general surgery. Each issue features original scientific contributions and clinical reports. Peer-reviewed articles cover topics in oncology, trauma, gastrointestinal, vascular, and transplantation surgery. The journal also publishes papers from the meetings of its sponsoring societies, the Society of University Surgeons, the Central Surgical Association, and the American Association of Endocrine Surgeons.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信