SSA-classifier based screening study for Alzheimer’s disease

IF 4.3 2区化学 Q1 SPECTROSCOPY

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy Pub Date : 2025-04-10 DOI:10.1016/j.saa.2025.126115

Zihao Qi , Zhigang Li , Peng Shan , Qiaoyun Wang , Weishang Sun

{"title":"SSA-classifier based screening study for Alzheimer’s disease","authors":"Zihao Qi , Zhigang Li , Peng Shan , Qiaoyun Wang , Weishang Sun","doi":"10.1016/j.saa.2025.126115","DOIUrl":null,"url":null,"abstract":"<div><div>Alzheimer’s is a disease (AD) that affects 10 % of individuals aged ≥ 65, is the most prevalent neurodegenerative disorder. We propose a diagnostic framework integrating plasma attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy with machine learning for AD screening. Four classifiers (SVM, Logistic Regression, XGBoost, LDA) were optimized using a modified Sparrow Search Algorithm (GSSA), benchmarked against its standard version (SSA) and Bayesian methods. GSSA-optimized classifiers demonstrated superior performance, with GSSA-XGBoost achieving peak metrics: 88.51 % accuracy (+2.30 % vs SSA-XGBoost), 95.35 % sensitivity, and 81.82 % specificity. Comparative test-set results revealed consistent improvements: SSA-optimized models attained 83.91 % (SVM), 77.01 % (Logistic), 86.21 % (XGBoost), and 79.31 % (LDA) accuracy, and Bayesian counterparts achieved 85.06 %, 80.46 %, 85.06 %, and 79.31 %,while GSSA-optimized models achieved 86.21 %,80.46 %,88.51 %,80.46 %,respectively. Moreover, GSSA further enhanced sensitivities to 97.67 % (SVM/LDA) and specificities to 81.82 % (XGBoost), outperforming both SSA and Bayesian approaches. This ATR-FTIR/GSSA-machine learning synergy shows significant potential as a minimally invasive AD screening tool, with XGBoost delivering optimal diagnostic balance. Our methodology advances spectroscopic biomarker discovery while demonstrating algorithmic optimization efficacy.</div></div>","PeriodicalId":433,"journal":{"name":"Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy","volume":"339 ","pages":"Article 126115"},"PeriodicalIF":4.3000,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy","FirstCategoryId":"92","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1386142525004214","RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"SPECTROSCOPY","Score":null,"Total":0}

引用次数: 0

Abstract

Alzheimer’s is a disease (AD) that affects 10 % of individuals aged ≥ 65, is the most prevalent neurodegenerative disorder. We propose a diagnostic framework integrating plasma attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy with machine learning for AD screening. Four classifiers (SVM, Logistic Regression, XGBoost, LDA) were optimized using a modified Sparrow Search Algorithm (GSSA), benchmarked against its standard version (SSA) and Bayesian methods. GSSA-optimized classifiers demonstrated superior performance, with GSSA-XGBoost achieving peak metrics: 88.51 % accuracy (+2.30 % vs SSA-XGBoost), 95.35 % sensitivity, and 81.82 % specificity. Comparative test-set results revealed consistent improvements: SSA-optimized models attained 83.91 % (SVM), 77.01 % (Logistic), 86.21 % (XGBoost), and 79.31 % (LDA) accuracy, and Bayesian counterparts achieved 85.06 %, 80.46 %, 85.06 %, and 79.31 %,while GSSA-optimized models achieved 86.21 %,80.46 %,88.51 %,80.46 %,respectively. Moreover, GSSA further enhanced sensitivities to 97.67 % (SVM/LDA) and specificities to 81.82 % (XGBoost), outperforming both SSA and Bayesian approaches. This ATR-FTIR/GSSA-machine learning synergy shows significant potential as a minimally invasive AD screening tool, with XGBoost delivering optimal diagnostic balance. Our methodology advances spectroscopic biomarker discovery while demonstrating algorithmic optimization efficacy.

Abstract Image

查看原文本刊更多论文

基于ssa分类器的阿尔茨海默病筛查研究

阿尔茨海默病是一种疾病（AD），影响10%的≥65岁的个体，是最普遍的神经退行性疾病。我们提出了一个将等离子体衰减全反射傅立叶变换红外（ATR-FTIR）光谱与机器学习相结合的诊断框架，用于AD筛查。使用改进的麻雀搜索算法（GSSA）对4个分类器（SVM、Logistic Regression、XGBoost、LDA）进行了优化，并对其标准版本（SSA）和贝叶斯方法进行了基准测试。gssa优化的分类器表现出优异的性能，GSSA-XGBoost达到峰值指标：准确率为88.51%（比SSA-XGBoost高2.30%），灵敏度为95.35%，特异性为81.82%。对比测试集结果显示了一致的改进：ssa优化模型的准确率分别为83.91% （SVM）、77.01% （Logistic）、86.21% （XGBoost）和79.31% (LDA)，贝叶斯模型的准确率分别为85.06%、80.46%、85.06%和79.31%，而gssa优化模型的准确率分别为86.21%、80.46%、88.51%、80.46%。此外，GSSA进一步将灵敏度提高到97.67% (SVM/LDA)，特异性提高到81.82% (XGBoost)，优于SSA和贝叶斯方法。这种ATR-FTIR/ gssa -机器学习协同作用显示出作为微创AD筛查工具的巨大潜力，XGBoost提供了最佳的诊断平衡。我们的方法推进了光谱生物标志物的发现，同时展示了算法优化的功效。

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

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 物理-光谱学

CiteScore

8.40

自引率

11.40%

发文量

1364

审稿时长

40 days

期刊介绍： Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy (SAA) is an interdisciplinary journal which spans from basic to applied aspects of optical spectroscopy in chemistry, medicine, biology, and materials science. The journal publishes original scientific papers that feature high-quality spectroscopic data and analysis. From the broad range of optical spectroscopies, the emphasis is on electronic, vibrational or rotational spectra of molecules, rather than on spectroscopy based on magnetic moments. Criteria for publication in SAA are novelty, uniqueness, and outstanding quality. Routine applications of spectroscopic techniques and computational methods are not appropriate. Topics of particular interest of Spectrochimica Acta Part A include, but are not limited to: Spectroscopy and dynamics of bioanalytical, biomedical, environmental, and atmospheric sciences, Novel experimental techniques or instrumentation for molecular spectroscopy, Novel theoretical and computational methods, Novel applications in photochemistry and photobiology, Novel interpretational approaches as well as advances in data analysis based on electronic or vibrational spectroscopy.