P. Dyroey
J. Heitz
H. Studier
S. Johannsmeier
T. Ripken

Multimodal imaging of bacterial species for label-free classification

Proc. SPIE 12854, Label-free Biomedical Imaging and Sensing (LBIS)
Type: Zeitschriftenaufsatz (non-reviewed)
Bacterial infections often pose a time-sensitive concern; however, the current diagnostic methods are comprised of lengthy processes. This study explores label-free imaging techniques based on two-photon excitation of intrinsic molecules found in various bacterial species. Specifically, two-photon excitation microscopy (TPEF) and fluorescence lifetime imaging microscopy (FLIM) are employed. These methods have been extensively utilized in the area of animal cells, yielding favorable outcomes, yet their application to bacteria remains largely unexplored. Analogous to the work on animal cells, initial attention is directed towards the metabolic coenzymes, namely nicotinamide adenine dinucleotide (phosphate) (NADPH) and flavin adenine dinucleotide (FAD). The extra time component of the FLIM setup was used to further investigate differences between the decay curves of the emitted autofluorescence in their different growth phases, corresponding to different internal microenvironments. Bacteria were excited at 740nm and 900nm and TPEF and FLIM signals were recorded with fitting bandpass filters, respectively. Different species (Escherichia coli K12, Pseudomonas fluorescens and Staphylococcus aureus) and different growth phases were examined to identify characteristic signal combinations. The two image channels and both wavelengths are compared and plotted against each other. The results show that the bacteria can already be classified by their autofluorescence signals. The growth phase seems to have less influence than the species. Based on this data, a classification is made to uniquely identify them. For this purpose, the database will next be expanded to include additional species and the classification will be automated.