Multi-target fluorescence staining of bacteria smears enables rapid machine learning-assisted species classification

Maxence Galvan; Michael Fujarski; Can Beslendi; Frieder Schaumburg; Julian Varghese; Johannes Liesche

doi:10.1002/mlf2.70076

mLife ›› 2026, Vol. 5 ›› Issue (2) :229 -238. DOI: 10.1002/mlf2.70076

METHOD

Multi-target fluorescence staining of bacteria smears enables rapid machine learning-assisted species classification

Author information +

History +

PDF (4385KB)

Abstract

Rapid identification of bacterial species from patient samples is crucial for clinical decision-making. In severe infections, such as bloodstream infections, the early start of an effective treatment is directly associated with reduced mortality rates. Current rapid species identification methods, such as matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) or multiplex PCR, require specialized hardware and extensive technical support that prevents application in resource-limited settings. Here, we present a staining and imaging procedure for bacterial smears using fluorescent dyes directed against intracellular structures and cell wall components. Data on relevant features were extracted from segmented images and used to train a machine learning (ML) model for species classification. The method was tested on clinical isolates from 126 patients. For the seven most common bacteria, the classification performance, indicated by area under the receiver operating characteristic (ROC) curve, ranged from 0.8 (Klebsiella pneumoniae) to 1 (Pseudomonas aeruginosa). Species that were not part of the training dataset, were reliably classified as unknown species. These results hold promise for the identification of further species, particularly Enterobacterales, and clinical application.

Keywords

artificial intelligence (AI) / bloodstream infection / fluorescence microscopy / rapid diagnostics / resource-limited settings

Cite this article

Download citation ▾

Maxence Galvan, Michael Fujarski, Can Beslendi, Frieder Schaumburg, Julian Varghese, Johannes Liesche. Multi-target fluorescence staining of bacteria smears enables rapid machine learning-assisted species classification. mLife, 2026, 5 (2) : 229-238 DOI:10.1002/mlf2.70076

登录浏览全文

4963

注册一个新账户忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Van Heuverswyn J, Valik JK, Desirée van der Werff S, Hedberg P, Giske C, Nauclér P. Association between time to appropriate antimicrobial treatment and 30-day mortality in patients with bloodstream infections: a retrospective cohort study. Clin Infect Dis. 2023; 76: 469–478.

[2]	Froböse NJ, Idelevich EA, Schaumburg F. Short incubation of positive blood cultures on solid media for species identification by MALDI-TOF MS: which agar is the fastest? Microbiol Spectr. 2021; 9:e0003821.

[3]	Idelevich EA, Seifert H, Sundqvist M, Scudeller L, Amit S, Balode A, et al. Microbiological diagnostics of bloodstream infections in Europe—an ESGBIES survey. Clin Microbiol Infect. 2019; 25: 1399–1407.

[4]	Fall B, Lo CI, Samb-Ba B, Perrot N, Diawara S, Gueye MW, et al. The ongoing revolution of MALDI-TOF mass spectrometry for microbiology reaches tropical Africa. Am Soc Trop Med Hyg. 2015; 92: 641–647.

[5]	Croxatto A, Prod'hom G, Greub G. Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev. 2012; 36: 380–407.

[6]	Rudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR, et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the global burden of disease study. Lancet. 2020; 395: 200–211.

[7]	Fuchs A, Tufa TB, Hörner J, Hurissa Z, Nordmann T, Bosselmann M, et al. Clinical and microbiological characterization of sepsis and evaluation of sepsis scores. PLoS One. 2021; 16:e0247646.

[8]	Lewis JM, Feasey NA, Rylance J. Aetiology and outcomes of sepsis in adults in sub-Saharan Africa: a systematic review and meta-analysis. Crit Care. 2019; 23: 212.

[9]	Beveridge T. Use of the Gram stain in microbiology. Biotech Histochem. 2001; 76: 111–118.

[10]	Fazii P, Ciancaglini E, Riario Sforza G. Differential fluorescent staining method for detection of bacteria in blood cultures, cerebrospinal fluid and other clinical specimens. Eur J Clin Microbiol Infect Dis. 2002; 21: 373–378.

[11]	Procop GW. Molecular diagnostics for the detection and characterization of microbial pathogens. Clin Infect Dis. 2007; 45: S99–S111.

[12]	Taniguchi M, Lindsay J. Absorption and fluorescence spectra of organic compounds from 40 sources: archives, repositories, databases, and literature search engines. Proc SPIE 11256, Reporters, Markers, Dyes, Nanoparticles, and Molecular Probes for Biomedical Applications. 2020; XII:112560J.

[13]	Yoon SA, Park SY, Cha Y, Gopala L, Lee MH. Strategies of detecting bacteria using fluorescence-based dyes. Front Chem. 2021; 9:743923.

[14]	Schleifer KH, Stackebrandt E. Molecular systematics of prokaryotes. Annu Rev Microbiol. 1983; 37: 143–187.

[15]	Schleifer KH, Kandler O. Peptidoglycan types of bacterial cell walls and their taxonomic implications. Bacteriol Rev. 1972; 36: 407–477.

[16]	Henry J, Endres JL, Sadykov MR, Bayles KW, Svechkarev D. Fast and accurate identification of pathogenic bacteria using excitation–emission spectroscopy and machine learning. Sensors Diagn. 2024; 3: 1253–1262.

[17]	Burns BL, Rhoads DD, Misra A. The use of machine learning for image analysis artificial intelligence in clinical microbiology. J Clin Microbiol. 2023; 61:e02336-21.

[18]	Sandmann S, Schaumburg F, Varghese J. GEFAAR: a generic framework for the analysis of antimicrobial resistance providing statistics and cluster analyses. Sci Rep. 2023; 13: 16922.

[19]	Cutler KJ, Stringer C, Lo TW, Rappez L, Stroustrup N, Brook Peterson S, et al. Omnipose: a high-precision morphology-independent solution for bacterial cell segmentation. Nat Methods. 2022; 19: 1438–1448.

[20]	van der Walt S, Schönberger JL, Nunez-Iglesias J, Boulogne F, Warner JD, Yager N, et al. scikit-image: image processing in Python. PeerJ. 2014; 2: e453.

[21]	Chen W, Qiu M, Paizs P, Sadowski M, Ramonaite T, Zborovsky L, et al. Universal, untargeted detection of bacteria in tissues using metabolomics workflows. Nat Commun. 2025; 16: 165.

[22]	Collins JT, Knapper J, Stirling J, Mduda J, Mkindi C, Mayagaya V, et al. Robotic microscopy for everyone: the OpenFlexure microscope. Biomed Opt Express. 2020; 11: 2447–2460.

[23]	Kong DS, Thorsen TA, Babb J, Wick ST, Gam JJ, Weiss R, et al. Open-source, community-driven microfluidics with metafluidics. Nat Biotechnol. 2017; 35: 523–529.

[24]	Liu W, Miao L, Li X, Xu Z. Development of fluorescent probes targeting the cell wall of pathogenic bacteria. Coord Chem Rev. 2021; 429: 213646.

[25]	Katz S. Coagulase test protocol. Washington DC, USA: American Society for Microbiology; 2010. https://asm.org/asm/media/protocol-images/coagulase-test-protocol.pdf?ext=.pdf

[26]	Shields P, Cathcart L. Oxidase test protocol. Washington DC, USA: American Society for Microbiology; 2010. https://asm.org/getattachment/00ce8639-8e76-4acb-8591-0f7b22a347c6/oxidase-test-protocol-3229.pdf

[27]	Holmes CL, Albin OR, Mobley HLT, Bachman MA. Bloodstream infections: mechanisms of pathogenesis and opportunities for intervention. Nat Rev Microbiol. 2024; 23: 210–224.