Utilization of Explainable Artificial Intelligence (XAI)-Powered Computer-Aided Detection (CAD) System on Chest X-Ray Abnormalities in Health Check-Ups.

IF 0.6 4区医学 Q4 MEDICINE, RESEARCH & EXPERIMENTAL

Yonago acta medica Pub Date : 2025-07-23 eCollection Date: 2025-08-01 DOI:10.33160/yam.2025.08.002

Shizuka Nishii, Katsuyuki Tomita, Hirokazu Touge, Hiroyuki Yamamoto, Keiji Shigeshiro, Akira Yamasaki

{"title":"Utilization of Explainable Artificial Intelligence (XAI)-Powered Computer-Aided Detection (CAD) System on Chest X-Ray Abnormalities in Health Check-Ups.","authors":"Shizuka Nishii, Katsuyuki Tomita, Hirokazu Touge, Hiroyuki Yamamoto, Keiji Shigeshiro, Akira Yamasaki","doi":"10.33160/yam.2025.08.002","DOIUrl":null,"url":null,"abstract":"Background: We designed a single-center retrospective study comparing the performance of commercially explainable artificial intelligence (XAI)-powered computer-aided detection (CAD) system of abnormal findings on chest X-rays (CXR) with that of non-experts, and pulmonology experts.Methods: A total of 1,262 images of 1,262 subjects (mean age 49 years; 52% female) and 1,252 images of 1,252 subjects (mean age 51 years; 51% female) were obtained from DICOM formats in Hakuai Hospital Health Check-up Center, in the pre-and post-implementing XAI-powered CAD period, respectively. The ultimate decision of abnormality on CXR was made by two pulmonology experts. The diagnostic accuracy metrics were measured accuracy and negative predictive value (NPV) for detecting abnormality on CXR.Results: XAI-powered CAD systems achieved an accuracy of 0.84 (95% confidential interval [CI] 0.82-0.86) and NPV of 1.00 (95% CI 0.99-1.00) to detect the abnormality on CXR. For determining nodular shadows, it was found to be non-inferior to the pulmonology experts with an accuracy of 0.94 (95% CI 0.92-0.95), and NPV of 1.00 (95% CI 0.99-1.00). It tended to overestimate the abnormality of heart enlargement and pleural thickening with a tendency for lower sensitivity.Conclusion: It seems likely that in the future, the most accurate screening CXR will be a double check combining with the pulmonology experts with XAI-powered CAD systems.","PeriodicalId":23795,"journal":{"name":"Yonago acta medica","volume":"68 3","pages":"180-186"},"PeriodicalIF":0.6000,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12343184/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Yonago acta medica","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.33160/yam.2025.08.002","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/8/1 0:00:00","PubModel":"eCollection","JCR":"Q4","JCRName":"MEDICINE, RESEARCH & EXPERIMENTAL","Score":null,"Total":0}

引用次数: 0

Abstract

Background: We designed a single-center retrospective study comparing the performance of commercially explainable artificial intelligence (XAI)-powered computer-aided detection (CAD) system of abnormal findings on chest X-rays (CXR) with that of non-experts, and pulmonology experts.

Methods: A total of 1,262 images of 1,262 subjects (mean age 49 years; 52% female) and 1,252 images of 1,252 subjects (mean age 51 years; 51% female) were obtained from DICOM formats in Hakuai Hospital Health Check-up Center, in the pre-and post-implementing XAI-powered CAD period, respectively. The ultimate decision of abnormality on CXR was made by two pulmonology experts. The diagnostic accuracy metrics were measured accuracy and negative predictive value (NPV) for detecting abnormality on CXR.

Results: XAI-powered CAD systems achieved an accuracy of 0.84 (95% confidential interval [CI] 0.82-0.86) and NPV of 1.00 (95% CI 0.99-1.00) to detect the abnormality on CXR. For determining nodular shadows, it was found to be non-inferior to the pulmonology experts with an accuracy of 0.94 (95% CI 0.92-0.95), and NPV of 1.00 (95% CI 0.99-1.00). It tended to overestimate the abnormality of heart enlargement and pleural thickening with a tendency for lower sensitivity.

Conclusion: It seems likely that in the future, the most accurate screening CXR will be a double check combining with the pulmonology experts with XAI-powered CAD systems.

查看原文本刊更多论文

可解释人工智能（XAI）驱动的计算机辅助检测（CAD）系统在健康检查中胸部x线异常的应用。

背景：我们设计了一项单中心回顾性研究，比较商业上可解释的人工智能（XAI）驱动的计算机辅助检测（CAD）系统在胸部x光片（CXR）异常表现与非专家和肺科专家的表现。方法：1262例受试者(平均年龄49岁；女性占52%)，1252名受试者(平均年龄51岁；（51%为女性）分别在实施XAI-powered CAD之前和之后在白会医院健康检查中心的DICOM格式中获得。两名肺科专家最终决定是否在x光片上显示异常。诊断准确性指标为测量准确性和阴性预测值（NPV），用于检测CXR异常。结果：xai驱动的CAD系统检测CXR异常的准确率为0.84(95%置信区间[CI] 0.82-0.86)， NPV为1.00 （95% CI 0.99-1.00）。对于结节影的确定，其准确性为0.94 (95% CI 0.92-0.95)， NPV为1.00 (95% CI 0.99-1.00)，不逊于肺科专家。它倾向于高估心脏增大和胸膜增厚的异常，并倾向于降低敏感性。结论：似乎在未来，最准确的筛查CXR将是与肺科专家结合xai驱动的CAD系统进行双重检查。

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分钟内获得全文求助全文

来源期刊

Yonago acta medica MEDICINE, RESEARCH & EXPERIMENTAL-

CiteScore

1.60

自引率

0.00%

发文量

审稿时长

>12 weeks

期刊介绍： Yonago Acta Medica (YAM) is an electronic journal specializing in medical sciences, published by Tottori University Medical Press, 86 Nishi-cho, Yonago 683-8503, Japan. The subject areas cover the following: molecular/cell biology; biochemistry; basic medicine; clinical medicine; veterinary medicine; clinical nutrition and food sciences; medical engineering; nursing sciences; laboratory medicine; clinical psychology; medical education. Basically, contributors are limited to members of Tottori University and Tottori University Hospital. Researchers outside the above-mentioned university community may also submit papers on the recommendation of a professor, an associate professor, or a junior associate professor at this university community. Articles are classified into four categories: review articles, original articles, patient reports, and short communications.