- Startseite
- Forschungsabteilungen
- Klinisch-Spektroskopische Diagnostik
- Publikationen
- Towards a blood culture-independent diagnostic tool for blood stream infections using Raman spectroscopy
Towards a blood culture-independent diagnostic tool for blood stream infections using Raman spectroscopy
in: Infection (2019)
Introduction: Clinical routine identification of blood stream infections is still problematic due to low colony forming unit numbers per ml of blood or difficult to meet culture conditions of certain pathogens [1]. Thus, progress in this traditionally microbiological area is necessary. Advances in immunology research have shown that Raman spectroscopy is a powerful tool to detect immune cell activation [2]. Objectives: In this study we used a blood culture- and pathogen isolation-independent diagnostic approach to identify bacterial and fungal infection from non-infected neutrophils. Methods: Freshly isolated neutrophils were co-incubated with Staphylococcus aureus as representative for Gram-positive bacteria, Escherichia coli for Gram-negative bacteria or Candida albicans for fungi. By using a High-Throughput-Screening-Raman Spectroscopy (HTS-RS) system nearly 20,000 cells from three different donors were measured and the data was fed into a random forest classification model. Furthermore, 2D Raman scans were recorded using a commercial Raman imaging system. Results: Infected neutrophils were correctly distinguished from noninfected cells with 92% accuracy, while neutrophils challenged with bacteria or fungi were successfully predicted with 90% accuracy. Even bacteria species prediction, in S. aureus and E. coli, was 84% accurate. When the Raman scans were analyzed, phagocytized yeasts could be visualized in Raman false color images without the need for labeling or staining. Conclusions: This proof-of-principle study shows that the Raman spectroscopic fingerprint of neutrophils carries the information about the activation state of the cell and furthermore, which pathogen species activated the cells.
DOI: Array