Rapid, non-invasive breath analysis for enhancing detection of silicosis using mass spectrometry and interpretable machine learning.

IF 3.7 4区 医学 Q1 BIOCHEMICAL RESEARCH METHODS
Merryn J Baker, Jeff Gordon, Aruvi Thiruvarudchelvan, Deborah Yates, William A Donald
{"title":"Rapid, non-invasive breath analysis for enhancing detection of silicosis using mass spectrometry and interpretable machine learning.","authors":"Merryn J Baker, Jeff Gordon, Aruvi Thiruvarudchelvan, Deborah Yates, William A Donald","doi":"10.1088/1752-7163/adbc11","DOIUrl":null,"url":null,"abstract":"<p><p>Occupational lung diseases, such as silicosis, are a significant global health concern, especially with increasing exposure to engineered stone dust. Early detection of silicosis is helpful for preventing disease progression, but existing diagnostic methods, including x-rays, computed tomography scans, and spirometry, often detect the disease only at late stages. This study investigates a rapid, non-invasive diagnostic approach using atmospheric pressure chemical ionization-mass spectrometry (APCI-MS) to analyze volatile organic compounds (VOCs) in exhaled breath from 31 silicosis patients and 60 healthy controls. Six different interpretable machine learning (ML) models with Shapley additive explanations (SHAP) were applied to classify these samples and determine VOC features that contribute the most significantly to model accuracy. The extreme gradient boosting classifier demonstrated the highest performance, achieving an area under the receiver-operator characteristic curve of 0.933 with the top ten SHAP features. The<i>m</i>/<i>z</i>442 feature, potentially corresponding to leukotriene-E3, emerged as a significant predictor for silicosis. The VOC sampling and measurement process takes less than five minutes per sample, highlighting its potential suitability for large-scale population screening. Moreover, the ML models are interpretable through SHAP, providing insights into the features contributing to the model's predictions. This study suggests that APCI-MS breath analysis could enable early and non-invasive diagnosis of silicosis, helping to improve disease outcomes.</p>","PeriodicalId":15306,"journal":{"name":"Journal of breath research","volume":" ","pages":""},"PeriodicalIF":3.7000,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of breath research","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1088/1752-7163/adbc11","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"BIOCHEMICAL RESEARCH METHODS","Score":null,"Total":0}
引用次数: 0

Abstract

Occupational lung diseases, such as silicosis, are a significant global health concern, especially with increasing exposure to engineered stone dust. Early detection of silicosis is helpful for preventing disease progression, but existing diagnostic methods, including x-rays, computed tomography scans, and spirometry, often detect the disease only at late stages. This study investigates a rapid, non-invasive diagnostic approach using atmospheric pressure chemical ionization-mass spectrometry (APCI-MS) to analyze volatile organic compounds (VOCs) in exhaled breath from 31 silicosis patients and 60 healthy controls. Six different interpretable machine learning (ML) models with Shapley additive explanations (SHAP) were applied to classify these samples and determine VOC features that contribute the most significantly to model accuracy. The extreme gradient boosting classifier demonstrated the highest performance, achieving an area under the receiver-operator characteristic curve of 0.933 with the top ten SHAP features. Them/z442 feature, potentially corresponding to leukotriene-E3, emerged as a significant predictor for silicosis. The VOC sampling and measurement process takes less than five minutes per sample, highlighting its potential suitability for large-scale population screening. Moreover, the ML models are interpretable through SHAP, providing insights into the features contributing to the model's predictions. This study suggests that APCI-MS breath analysis could enable early and non-invasive diagnosis of silicosis, helping to improve disease outcomes.

利用质谱法和可解释的机器学习进行快速、无创的呼吸分析,以加强对矽肺病的检测。
职业性肺病,如矽肺病,是一个重大的全球健康问题,特别是随着越来越多地接触工程石尘。矽肺的早期发现有助于预防疾病进展,但现有的诊断方法,包括x射线,计算机断层扫描和肺活量测定法,通常只能在晚期发现疾病。本研究探讨了一种快速、无创的诊断方法,使用大气压化学电离-质谱(APCI-MS)分析31名矽肺患者和60名健康对照者呼出气体中的挥发性有机化合物(VOCs)。六种不同的可解释机器学习(ML)模型与Shapley加性解释(SHAP)被应用于分类这些样本,并确定对模型准确性贡献最大的VOC特征。极端梯度增强分类器表现出最好的性能,在接收算子特征曲线下的面积为0.933,具有前10个SHAP特征。m/z 442特征,可能对应于白三烯- e3,成为矽肺病的重要预测因子。挥发性有机化合物的采样和测量过程每个样本需要不到五分钟,突出了其潜在的适合大规模人群筛选。此外,机器学习模型可以通过SHAP进行解释,从而深入了解有助于模型预测的特征。本研究表明,APCI-MS呼吸分析可以实现矽肺的早期和非侵入性诊断,有助于改善疾病预后。
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
Journal of breath research
Journal of breath research BIOCHEMICAL RESEARCH METHODS-RESPIRATORY SYSTEM
CiteScore
7.60
自引率
21.10%
发文量
49
审稿时长
>12 weeks
期刊介绍: Journal of Breath Research is dedicated to all aspects of scientific breath research. The traditional focus is on analysis of volatile compounds and aerosols in exhaled breath for the investigation of exogenous exposures, metabolism, toxicology, health status and the diagnosis of disease and breath odours. The journal also welcomes other breath-related topics. Typical areas of interest include: Big laboratory instrumentation: describing new state-of-the-art analytical instrumentation capable of performing high-resolution discovery and targeted breath research; exploiting complex technologies drawn from other areas of biochemistry and genetics for breath research. Engineering solutions: developing new breath sampling technologies for condensate and aerosols, for chemical and optical sensors, for extraction and sample preparation methods, for automation and standardization, and for multiplex analyses to preserve the breath matrix and facilitating analytical throughput. Measure exhaled constituents (e.g. CO2, acetone, isoprene) as markers of human presence or mitigate such contaminants in enclosed environments. Human and animal in vivo studies: decoding the ''breath exposome'', implementing exposure and intervention studies, performing cross-sectional and case-control research, assaying immune and inflammatory response, and testing mammalian host response to infections and exogenous exposures to develop information directly applicable to systems biology. Studying inhalation toxicology; inhaled breath as a source of internal dose; resultant blood, breath and urinary biomarkers linked to inhalation pathway. Cellular and molecular level in vitro studies. Clinical, pharmacological and forensic applications. Mathematical, statistical and graphical data interpretation.
×
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