EUVIMEDEUVIMED entdecken
Crossref

Medical Electronic Nose System for Rapid Diagnosis of Bloodstream Infection

Jie Min, Jiali Li, Yimin Zhou, Pengxin Li, Wei Liang, Li Gao, Wang Li

IEEE Sensors Journal · 01.06.2026 · 26 (11) · 17568-17577

DOI: 10.1109/jsen.2026.3683390 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Sensors Journal, erschienen am 01.06.2026. Verfasst von Jie Min und 6 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Gefördert von: Chongqing Municipal Key Project for Technology Innovation and Application Development.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Early identification of bloodstream infection (BSI) is critical for timely therapeutic intervention. However, conventional blood culture methods are limited by prolonged turnaround times and low sensitivity. This study proposes a rapid diagnostic approach for BSI based on the detection of alterations in the profile of volatile organic compounds (VOCs) in exhaled breath using an electronic nose (e-nose). Breath samples were collected from 132 blood culture-positive patients and 97 healthy volunteers (comprising hospital staff, as well as university students and teachers). A sensor array composed of 14 gas sensors was employed to capture the changes in VOCs. Three feature extraction models [including principal component analysis (PCA), linear discriminant analysis (LDA), and hybrid dictionary-rocket architecture (HYDRA)] and six classification models (including random forest (RF), LDA, multilayer perceptron (MLP), support vector machine (SVM), logistic regression (LR), and ridge regression) were systematically evaluated. Results demonstrated that HYDRA combined with ridge regression achieved optimal performance via a held-out test set, yielding an average accuracy of 0.87, precision of 0.90, sensitivity of 0.88, F1-score of 0.89, and AUC of 0.94 within the 95% CI. However, the healthy nonhospitalized controls may introduce confounding from differences in medication, metabolic status, and diet. Nevertheless, e-nose-based breath analysis shows great potential as a noninvasive tool for early BSI diagnosis and warrants further validation with more clinically comparable controls.

Förderung

  • Chongqing Municipal Key Project for Technology Innovation and Application Development

Zitieren

APA

Min, J., Li, J., Zhou, Y., Li, P., Liang, W., Gao, L., & Li, W. (2026). Medical Electronic Nose System for Rapid Diagnosis of Bloodstream Infection. IEEE Sensors Journal, 26(11), 17568-17577. https://doi.org/10.1109/jsen.2026.3683390

Vancouver

Min J, Li J, Zhou Y, Li P, Liang W, Gao L, et al. Medical Electronic Nose System for Rapid Diagnosis of Bloodstream Infection. IEEE Sensors Journal. 2026;26(11):17568-17577. doi:10.1109/jsen.2026.3683390

BibTeX
@article{Min2026,
  author = {Min, Jie and Li, Jiali and Zhou, Yimin and Li, Pengxin and Liang, Wei and Gao, Li and Li, Wang},
  title = {Medical Electronic Nose System for Rapid Diagnosis of Bloodstream Infection},
  journal = {IEEE Sensors Journal},
  year = {2026},
  volume = {26},
  number = {11},
  pages = {17568--17577},
  doi = {10.1109/jsen.2026.3683390}
}
Kostenlos starten →