Advances in imaging techniques for the study of individual bacteria and their pathophysiology

Abstract

Bacterial heterogeneity is pivotal for adaptation to diverse environments, posing significant challenges in microbial diagnostics and therapeutic interventions. Recent advancements in high-resolution optical microscopy have revolutionized our ability to observe and characterize individual bacteria, offering unprecedented insights into their metabolic states and behaviors at the single-cell level. This review discusses the transformative impact of various high-resolution imaging techniques, including fluorescence and label-free imaging, which have enhanced our understanding of bacterial pathophysiology. These methods provide detailed visualizations that are crucial for developing targeted treatments and improving clinical diagnostics. We highlight the integration of these imaging techniques with computational tools, which has facilitated rapid, accurate pathogen identification and real-time monitoring of bacterial responses to treatments. The ongoing development of these optical imaging technologies promises to significantly advance our understanding of microbiology and to catalyze the translation of these insights into practical healthcare solutions.

Full text

1
Advances in imaging techniques for the study of individual bacteria
and their pathophysiology
Dohyeon Lee1,2+, Hyun-Seung Lee4+, Moosung Lee1,2, Minhee Kang5,6, Geon Kim1,2, Tae Yeul Kim7*, Nam Yong
Lee7, and YongKeun Park1,2,3*
1Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon, 34141, Republic of
Korea
2KAIST Institute for Health Science and Technology, KAIST, Daejeon, 34141, Republic of Korea
3Tomocube Inc., Daejeon, 34109, Republic of Korea
4Department of Laboratory Medicine, Wonkwang University Hospital, Iksan, 54538, Republic of Korea
5Biomedical Engineering Research Center, Smart Healthcare Research Institute, Samsung Medical Center, Seoul,
06351, Republic of Korea
6Department of Medical Device Management and Research, Samsung Advanced Institute for Health Sciences &
Technology, Sungkyunkwan University, Seoul, 06351, Republic of Korea
7Department of Laboratory Medicine and Genetics, Samsung Medical Center, Sungkyunkwan University School
of Medicine, Seoul, 06351, Republic of Korea
+Equally contributed.
*Correspondence: yk.park@kaist.ac.kr and voltaire0925@gmail.com
Abstract
Bacterial heterogeneity is pivotal for adaptation to diverse environments, posing significant challenges in
microbial diagnostics and therapeutic interventions. Recent advancements in high-resolution optical microscopy
have revolutionized our ability to observe and characterize individual bacteria, offering unprecedented insights
into their metabolic states and behaviors at the single-cell level. This review discusses the transformative impact
of various high-resolution imaging techniques, including fluorescence and label-free imaging, which have
enhanced our understanding of bacterial pathophysiology. These methods provide detailed visualizations that are
crucial for developing targeted treatments and improving clinical diagnostics. We highlight the integration of
these imaging techniques with computational tools, which has facilitated rapid, accurate pathogen identification
and real-time monitoring of bacterial responses to treatments. The ongoing development of these optical imaging
technologies promises to significantly advance our understanding of microbiology and to catalyze the translation
of these insights into practical healthcare solutions.
Keywords
Bacteria imaging, High-resolution, Optical microscopy, Single-cell imaging, Fluorescence imaging, Label-free
imaging

2
1. Introduction
Bacteria are among the simplest unicellular organisms, yet they exhibit significant heterogeneity in their
physiology and behavior. This heterogeneity, arising from genetic, environmental, or stochastic factors, critically
impacts how bacteria adapt to and interact with their surroundings1,2. Understanding bacterial heterogeneity is
therefore a foundational challenge in microbiology, as it plays a crucial role in bacterial survival and adaptability
across diverse environments.
To investigate bacterial heterogeneity, optical microscopy techniques have been widely employed. Recently,
breakthroughs in high-resolution optical microscopy have made it possible to study individual bacteria with
unparalleled specificity and spatial resolution, enabling researchers to observe their metabolic states, structural
organization, and dynamic interactions. These imaging technologies are transforming clinical microbiology and
infectious disease research, offering rapid and precise tools for bacterial identification and real-time monitoring
of bacterial responses to treatments. For example, various high-resolution microscopic techniques have been used
to study the division of bacteria in biofilms, where bacteria divide different tasks to optimize their survival, such
as nutrient acquisition, defense, and attachment3-7. Such insights can lead to new strategies for managing biofilms,
a major cause of medical device-related infections.
Identifying bacterial species and determining antimicrobial susceptibility is essential for guiding treatment
decisions in bacterial infections, yet traditional identification methods, including biochemical assays8-11, serologic
tests12, staining13,14, and conventional antimicrobial susceptibility testing (AST) methods, including broth
microdilution15, often lack speed, sensitivity, or specificity. Chromogenic cultures16,17 and optical sensor-based
approaches8,10,18-21 offer simplicity and cost-effectiveness but require large sample volumes and may yield
inaccurate AST results in mixed infections. Mass spectrometry, while rapid and affordable for many
applications20,22, still provides limited identification data due to restricted libraries23. Similarly, genetic methods
offer highly sensitive bacterial detection with minimal sample requirements but at a high cost and with potential
for therapy-related false positives24-30. Imaging-based technology detecting small colonies has been a potential
solution for rapid bacterial identification31,32. These microbiological methods have been clinically tested and
commercialized (Table 1). Recently, high-resolution, single-cell-based imaging has become a promising
alternative because identification and AST are determined within a few cell cycles.
Moreover, high-resolution microscopy enables exploration of bacterial heterogeneity beyond simple
population-level measurements, such as growth rate and biochemical composition33-36. By allowing single-cell
visualization, these imaging methods reveal the diversity of bacterial behaviors within clonal populations and
interactions between individual bacteria and external stressors. For example, high-resolution imaging has shown
how bacteria can divide tasks within biofilms, enhancing their collective survival under challenging conditions.
This review discusses the transformative role of high-resolution optical microscopy in revealing bacterial
properties at the single-cell level and advancing our understanding of bacterial functions. We begin with an
overview of high-resolution microscopic techniques, from fluorescence to label-free imaging, and explore their
contributions to real-time bacterial identification, locomotion analysis, biofilm research, and monitoring cellular
responses to environmental changes. Additionally, we highlight the potential for integrating these imaging
techniques with microfluidic platforms and advanced image processing, which promises to further enhance their
impact on both preclinical research and clinical diagnostics.

3
Table 1 | Microbiological techniques for bacterial identification and AST for bloodstream infections in clinical microbiology laboratories
Mechanism Method or product Manufacturer Automation
Samples
Turnaround
time AST Advantage disadvantage
Blood
Culture-
positive
broth
Sub-
culture
Staining Manual staining13,14 - X . O O 0.2 h X
Relatively rapid and simple.
Cost-effective.
Labor intensive. Low sensitivity/specificity. Needs
confirmatory tests.
Biochemical
test
Manual biochemical
test
- X . . O X
Cost-effective. Ti me-consuming, labor-intensive. Complicated
procedures. Requires large numbers of bacteria.
API8-10
bioMerieux
X
.
.
O
4–24 h
X
BBL Crystal9
BD Diagnostics
X
.
.
O
4-20 h
X
RapID11
Thermo Fisher
X
.
.
O
4-6 h
X
Serologic
test
Manual serologic test12
- X . O O 1-3 day X
Alternative when pathogen
culture is difficult.
Time-consuming, Requires acute and convalescent
specimens.
Chromogeni
c culture
CHROMID16
bioMerieux
X
.
O
O
24 h
O
Simple, standardized method.
Time-consuming. Requires additional confirmatory
tests.
Spectra MRSA17
Thermo Fisher
X
.
O
O
24 h
O
Optic sensor-
based
MicroScan
Walk Awa y
8,10,18,19
Beckman Coulter O . . O 6–16 h O
Cost-effective. Detects
resistance mechanisms.
Requires large numbers of bacteria. Discrepancy in
AST results. Different test kits for each type of
bacteria.
VITEK10,18-20
bioMerieux
O
.
.
O
6–16 h
O
Sensititre ARIS21
Thermo Fisher
O
.
.
O
18–24 h
O
BD Phoenix10,19
BD Diagnostics
O
.
.
O
6–16 h
O
Mass
spectrometry
VITEK MS20,22
bioMerieux
X
.
O
O
<0.5 h
X
Accurate and rapid. Cost-
effective.
Requires large numbers of bacteria. Additional
procedure for bacteria concentration, when culture
positive broth was directly used. Inaccuracy in
combined infections. Requires an additional
procedure for AST.
Biotyper22 Bruker Daltonics X . O O <0.5 h O
Genetic test
T2MR24
T2 Biosystems
O
O
.
.
3–5 h
X
Requires small numbers of
bacteria. Accurate and rapid.
AST results with resistance
mechanism detection. Panel
flexibility with multiplex PCR
and primer design modification.
High cost.
Various sensitivities according to structure
of certain bacterial matrix. Therapy-related false
positive. Inaccuracy in combined infection.
Disc
repancy of AST results compared to the
reference
method.
Magicplex Sepsis real-
time test
25
Seegene X O . . 6 h O
iDTECT Dx Blood26
PathoQuest
X
O
.
.
2–3 days
O
LightCycler SeptiFast27
Roche
O
O
.
.
3–4 h
X
SepsiTest28
Molzym
O
O
.
.
8–12 h
X
FilmArray BCID229
bioMerieux
O
.
O
.
1 h
O
GeneXpert MRSA/SA
Blood Culture
30
Cepheid O . O . 1 h O
Imaging
Accelerated Pheno
system
31
Accelerate
Diagnostics
O . O . 7 h O
Requires small numbers of
bacteria. Rapid.
AST result with
resistance mechanism detection
Requires fresh culture-positive broth.
Discrepancy in
AST results.
dRAST32
QuantaMatrix
O
.
O
.
5–7 h
O
1

4
2. Recent advances in high-resolution microscopy for live bacterial imaging
The transparent nature of unlabeled bacterial cells presents a considerable challenge for researchers aiming to
achieve high-contrast imaging. Traditional approaches, such as staining and fluorescence labeling, have been
employed to visualize bacterial cells and their substructures with enhanced contrast or molecular specificity.
However, fixed-cell imaging methods are inherently limited, as they can only capture dead cells. Additionally,
genetic transfections for fluorescence imaging may alter the physiological characteristics of live bacteria and
introduce issues such as photobleaching and phototoxicity.
To address these limitations, various non-invasive optical microscopy techniques have been developed,
allowing researchers to capture the dynamic activities of live bacterial cells. Nonetheless, achieving high
spatiotemporal resolution and sufficient contrast to visualize substructures and molecular details—particularly in
motile bacteria—remains a significant challenge.
Recent advances in microscopy have successfully overcome many of these obstacles, enabling high-resolution
imaging of living bacterial cells. These techniques can be broadly categorized into two classes: high-resolution
fluorescence microscopy and label-free microscopy. High-resolution fluorescence microscopy requires labeling,
while label-free techniques do not rely on any exogenous markers. Pushing the technical boundaries of both spatial
and temporal resolution has opened new possibilities for in-depth studies of subcellular structures, bacteria-
bacteria interactions, and long-term investigations of bacterial pathophysiology.
In this section, we will explore recent advances in high-resolution microscopy for live bacterial imaging (Fig.
1). We will discuss how these innovations have enabled researchers to surpass the challenges of imaging unlabeled
bacterial cells with high spatiotemporal resolution. This includes developments in super-resolution microscopy,
structured illumination microscopy, and light-sheet microscopy. Furthermore, we will cover label-free techniques,
such as phase-contrast microscopy, differential interference contrast microscopy, and quantitative phase imaging,
which facilitate the imaging of live bacteria without the need for labeling.
2.1 High-resolution fluorescence microscopy
Fluorescence microscopy is a well-established tool in biological research, allowing for the visualization of specific
organelles and molecules by attaching different fluorescent markers to multiple objects of interest. This enables
the observation of separate structures through multi-channel imaging. Alongside fluorescence microscopy, non-
invasive optical microscopy techniques have been developed to visualize color-tagged bacterial cells while
keeping them alive. The primary goal of high-resolution fluorescence microscopy for live bacterial imaging has
been to enhance spatiotemporal resolution and extend the imaging window, effectively increasing the spacetime-
bandwidth product. Techniques such as optical sectioning microscopy, non-uniform excitation microscopy, and
single-molecule microscopy exemplify the advancements made in these areas.

5
Fig. 1 Schematic diagrams of high-resolution microscopy techniques are categorized based on their
dependence on bacterial sample labeling. (a-c) High-resolution fluorescence imaging techniques require
sample labeling before microscopic imaging. (a) Optical sectioning microscopy removes fluorescent
signals outside the focal plane for high-contrast imaging. (b) Nonuniform excitation microscopy
achieves high-resolution imaging utilizing an engineered excitation beam. (c) Single-molecule
microscopy tracks the exact positions of blinking fluorophores in a few nanometer scales. (d-f) Label-
free imaging techniques enable the imaging of raw bacterial samples. (d) Endogenous molecular
microscopy utilizes the excitation-relaxation of specific endogenous molecules to generate optical
fingerprints. (e) Conventional optical microscopy and (f) quantitative phase imaging (QPI) visualize
bacteria by contrasting or visualizing light-scattering patterns produced by the bacterial cells.
Optical sectioning microscopy
Optical sectioning microscopy enhances the contrast of images by suppressing fluorescence noise outside the
focal plane (Fig. 1a). Confocal microscopy and light sheet microscopy are two popular techniques in high-
resolution fluorescence microscopy for imaging live bacterial cells. Confocal microscopy uses a laser beam to
excite fluorescently labeled molecules in a sample, and a pinhole in front of the detector to selectively capture the
emitted light from the focal plane37. This technique provides high contrast and spatial resolution by removing out-
of-focus light, which results in a clearer image of the sample. However, confocal microscopy is relatively slow
and may cause phototoxicity and photobleaching due to its high-power laser excitation.
Light sheet microscopy, on the other hand, uses a thin sheet of light to excite the fluorescently labeled sample
in a plane perpendicular to the direction of imaging38. The emitted light is then detected by a camera positioned
perpendicular to the plane of illumination, which allows for fast and gentle imaging of living bacterial cells. This
technique has the advantage of high-speed imaging, high contrast, and minimal photodamage, making it an

6
excellent option for long-term imaging of live cells. Light sheet microscopy is particularly useful for studying
bacterial dynamics, such as in biofilms, where it can provide a 3D view of the sample.
Lattice light sheet microscopy is an advanced version of light sheet microscopy that employs a lattice pattern
of light sheets for even greater imaging speed and reduced phototoxicity39. Lattice light sheet microscopy allows
for ultrafast imaging at up to 1,000 frames per second, which is fast enough to capture rapid cellular processes
such as bacterial motility and cell division. Additionally, the use of multiple light sheets in lattice light sheet
microscopy results in less photodamage to the sample since each individual sheet is lower in intensity than a single
sheet in traditional light sheet microscopy. This makes it possible to image live cells for longer periods without
damaging them.
Nonuniform excitation microscopy
Nonuniform excitation microscopy yields super-resolution by giving gated illumination onto the sample through
spatially nonuniform light (Fig. 1b). Structured illumination microscopy (SIM) uses sinusoidal illuminations to
evoke the mixing of frequencies between harmonic patterns and the sample. Reconstruction of the digital image
reveals previously inaccessible high-frequency components encoded into the observed image, providing lateral
resolution beyond the diffraction limit. To further improve the resolution, saturated (nonlinear) SIM can be
employed by illuminating high-powered sinusoidal beams and saturating the fluorescent response40,41. However,
phototoxicity and photobleaching problems limit its applications. Alternatively, stimulated emission depletion
(STED) microscopy is one of the earliest super-resolution methods widely utilized in microbiology until now42.
In contrast with nonlinear SIM, STED microscopy sends the fluorophores into a dark state except for the focal
spot. This de-excitation or depletion of the surrounding emission can be achieved using a high-powered donut-
shaped depletion beam. Because it is a scanning-based method, STED microscopy takes advantage of rapid super-
resolution imaging of bacteria by restricting the field of view. Additionally, photostable43 or reversely photo-
switchable44 proteins can reduce photobleaching and phototoxicity problems for live-cell imaging.
Single-molecule microscopy
Single-molecule localization microscopy (SMLM) is another super-resolution method that exploits the stochastic
emission process of fluorophores (Fig. 1c). By separating blinking fluorophores from time-lapse fluorescence
images, SMLM enables super-resolution without specialized instrumentation45,46. Rapid-acquisition SMLM
enables visualization of spatially clustered protein distributions, chromosome dynamics tracking, and protein
molecule diffusion in living bacteria47-51.
2.2. Label-free microscopy
Photobleaching and phototoxicity are the significant hurdles of fluorescence microscopy that prevent its
widespread use in microbiology. Furthermore, conventional fluorescent proteins require sufficient oxygen to emit
light, thus making this technique incompatible with anaerobic bacteria52. Several label-free approaches have been
devised to capture the intrinsic properties of living bacteria to overcome such challenges.
Phase contrast microscopy

7
Throughout the history of microscopy, scientists have relied on scattered light to visualize unlabeled bacteria
(Fig. 1e). Bright-field microscopy, phase-contrast microscopy, and differential interference contrast (DIC)
microscopy are widely used label-free techniques for imaging bacteria. Bright-field microscopy is a simple and
widely used optical microscopy technique that provides contrast based on the absorption and scattering of light
by the sample. While it is useful for observing the overall morphology of bacterial cells and providing information
about bacterial growth and division, it lacks high image contrast due to weak scattering of bacteria, resulting in
poor contrast of the refractive index of bacteria and their surrounding media.
Phase-contrast microscopy enhances contrast between objects of similar refractive index, enabling visualization
of live bacteria without labeling53. This method is particularly useful for observing bacterial motility and flagellar
movement, providing insights into the mechanisms of bacterial motility. Differential interference contrast (DIC)
microscopy utilizes polarized light to generate contrast based on variations in refractive index and optical path
length of the sample54. This technique enhances the contrast of edges and boundaries, making it particularly useful
for observing bacterial structures and the interactions between bacteria and their environment. Label-free phase-
contrast techniques allow for imaging live bacteria without fluorescent labeling or genetic modifications,
providing a valuable tool for studying bacterial physiology and behavior. However, they only provide qualitative
imaging information and are limited to 2D.
Label-free molecular imaging
Autofluorescence and Raman scattering microscopy are two label-free imaging techniques that are becoming
increasingly popular in microbiology. Autofluorescence microscopy takes advantage of the intrinsic fluorescence
of certain endogenous biomolecules in bacterial cells, such as NADH, flavins, and porphyrins, to generate contrast
and image live bacteria without the need for exogenous labels55. The emitted fluorescence spectra can also provide
information on the metabolic state of the bacteria56, allowing for the monitoring of bacterial physiology and
behavior over time. Via fluorescence lifetime microscopy (FLIM), auto-fluorophores could be able to unravel the
internal condition of bacterial cells non-invasively57. Autofluorescence microscopy has been used to study a
variety of bacteria, including monitoring citrus greening diseases in plants caused by Candidatus Liberibacter
asiaticus58, laser-induced fluorescence spectroscopy imaging of murine gastrointestinal tract59.
Raman scattering microscopy is another label-free technique that provides information on the chemical
composition of bacterial cells60,61. Raman scattering occurs when incident light interacts with chemical bonds in
the sample, leading to a shift in the energy of the scattered photons. By analyzing the spectrum of the scattered
light, the chemical composition of the sample can be determined. Raman scattering microscopy can provide
molecular specificity and has been used to image bacterial cells and study their interactions with their environment.
A recent study took advantage of chemical specificity through deuterium-tagging and used this to image the
metabolic activity of live bacteria upon antibiotics susceptibility test62-66. However, the technique can be
challenging due to the low signal intensity of Raman scattering and the potential for sample damage from the
high-intensity laser used for excitation. Advances in Raman scattering microscopy, such as surface-enhanced
Raman scattering (SERS) and coherent Raman scattering (CRS), have improved sensitivity and reduced
photodamage, making it a promising tool for studying bacterial physiology and behavior60,67.
Quantitative phase imaging

8
Quantitative phase imaging (QPI) is a label-free technique that exploits refractive index as an intrinsic quantitative
imaging contrast68 (Fig. 1f). 2D QPI techniques, such as interferometric microscopy or digital holographic
microscopy, measure the optical phase delay that provides information about the optical thickness and refractive
index of biological samples. 3D QPI techniques, also known as holotomography69, reconstruct the 3D RI
distribution of a biological sample from multiple 2D optical measurements.
QPI is a powerful tool for observing living bacteria due to its high sensitivity and ability to capture fine details
of cells and their behavior without the need for labeling or staining. This technique has been extensively used in
various applications, including morphological and structural analysis70, and growth kinetics49. Quantitative
measurement of the refractive index of bacteria and their surroundings by QPI allows for the precise detection of
small changes in cell morphology and dynamics71, making it a valuable tool for studying bacterial physiology and
behavior. Holotomography, one of the 3D QPI techniques, has enabled the quantification of
polyhydroxyalkanoates (PHA) produced by individual bacteria, which includes the concentration, volume, and
dry mass of PHAs in live unlabeled bacteria72, and the investigation of antibacterial activities of engineered films73
or applied antibiotics74 by monitoring the dynamics of individual bacteria.
Moreover, QPI has potential applications in clinical diagnostics. For instance, QPI has been utilized to detect
anthrax spores by analyzing 2D optical field images of individual bacterial spores using a machine-learning
approach75. High-resolution label-free imaging of the growth of individual bacteria upon antibiotic treatments by
QPI also suggests potential for rapid label-free image-based antibiotic susceptibility testing (AST)74.
Holotomography, coupled with a deep-learning algorithm, was used to identify bacterial pathogens commonly
found in sepsis76. Hence, QPI offers a promising alternative to traditional staining and culturing methods for rapid
detection and identification of bacterial pathogens in clinical settings.
Electrical imaging
Label-free electrical techniques represent an exciting frontier in the study of cellular systems. These techniques
provide impedance and electrochemical information on cellular properties, including redox potential77, cell–cell
adhesion78 , and real-time kinetics79, which are often challenging to capture with optical imaging methods. Due to
their label-free nature, electrical imaging techniques offer the unique advantage of monitoring cellular behavior
over extended periods with minimal perturbation to cellular physiology. This makes them particularly well-suited
for investigating microbial populations' responses to varying environmental conditions or AST80 .
Recent advancements in complementary metal-oxide-semiconductor (CMOS) microelectrode array (MEA)
technology have enabled high-resolution, high-throughput functional imaging of unlabeled live-cell cultures
through in situ impedance and electrochemical measurements81-83. CMOS MEAs provide an unparalleled
opportunity to capture the real-time responses of adherent cells, such as those found in biofilms, with remarkable
precision. As the feature sizes of CMOS chips continue to shrink, the spatial resolution of electrical imaging is
expected to further improve, allowing for subcellular characterization of bacteria. This opens new avenues for
exploring microbial physiology at an unprecedented level of detail.
3. High-resolution imaging techniques for the study of bacterial physiology

9
High-resolution imaging techniques can be used to identify bacteria and study their physiological properties, such
as locomotion activity, biofilm formation, and cellular response. Fluorescence-tagged molecules can provide
information about subcellular structures or represent the metabolic state of the cell. Autofluorescence imaging of
endogenous markers, such as NADH/NAD+, can reveal cellular metabolism through redox reactions. Quantitative
phase imaging (QPI) can provide dry mass information through phase shift measurements, which have a linear
relationship with protein concentration. These techniques have been actively studied in the microbiological field
to gain insights into bacterial physiology and behavior. In the following subsections, we describe details of such
applications that are actively studied in the microbiological field.
3.1 Bacterial species identification
Fig. 2 (a) FISH analysis detects bacteria in a species-specific manner through the usage of specific
fluorescent probes. (b) Bacterial phasor fingerprint map generated by FLIM, where phasor shifts are
considered as metabolic changes within the cells. (c) Identification of anthrax spores using custom
neural network. (d) A neural activation profile resulted from artificial neural network processing of a
3D refractive index tomogram. (a) is adapted from Ref.84. (b) is adapted from Ref.85. (c) is adapted from
Ref. 75. (d) is adapted from Ref.76.
The examination of bacteria using conventional optical microscopy cannot distinguish between different bacterial
species, as they exhibit similar shapes and sizes. However, bacterial species identification is crucial for the
diagnosis and treatment of infectious diseases, and various identification methods are used in clinical
microbiology laboratories. Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass
spectrometry has been widely used for bacterial species identification, but it requires high initial cost for MALDI-
TOF equipment86.
Fluorescence in situ hybridization (FISH) employed imaging has been proven to be a reliable method for
bacterial detection under complex environments since fluorescence tagging enables chemical imaging with high

10
specificity. FISH targets a specific sequence of DNA or RNA using complementary probes, providing chemical
imaging with high specificity87. FISH has been used to detect bacteria in various applications, ranging from
quantifying microbial populations in beverages84 to laboratory diagnostics88 (Fig. 2a). Smartphone-based FISH
microscopy has also been introduced, further extending the application scope89.
Despite the high specificity of FISH microscopy, it is limited to identifying only the bacterial species that have
available probes. As an alternative, autofluorescence microscopy can be used for label-free bacterial identification.
Autofluorescence is an intrinsic modality of fluorescence emitted by molecules naturally residing in cells. The
potential of autofluorescence microscopy for bacterial detection was demonstrated by selectively imaging
pathogenic bacteria, including Mycobacterium tuberculosis90. In a recent study, biomarkers specific to different
bacterial species were found in the autofluorescence images obtained under a two-photon setting of fluorescence
lifetime imaging microscopy (FLIM)85 (Fig. 2b). The fingerprint patterns of five different species were located in
the phasor of FLIM measurement, which resulted from the autofluorescence of metabolic molecules.
QPI imaging provides label-free images that offer direct evidence for certain bacterial species. The rich
morphological profiling obtained from QPI can be statistically leveraged using machine learning-based
classification. Jo et al. suggested the potential of QPI for bacterial identification using linear machine learning to
classify bacterial species from angular spectrum features that were made accessible through QPI measurements91.
Another study successfully distinguished spores of multiple Bacillus species, including the hazardous B. anthracis,
by introducing deep learning to classify images acquired with a portable QPI unit75 (Fig. 2c). Recently, bacterial
identification using QPI has been advanced by combining 3D QPI and an advanced 3D deep learning
architecture76 (Fig. 2d). The high label-free single-cell identification performance was exploited to distinguish
among 19 bacterial species of bloodstream infection pathogens.
3.2 Bacterial locomotion
Fig. 3 (a) Dark-field microscopy capture an image sequence of a single bacteria at 420 frames per second.
Bacteria show three different motility phases. (b) Conventional phase contrast 3D tracking of a single
bacterium without mechanical refocusing. They measured the axial position of bacteria via cross-
correlation between diffracted patterns and a reference library. (a) is adapted from Ref.92,93. (b) is
adapted from Ref.94.

11
Measuring the locomotion of individual bacteria is important for understanding bacterial behavior, interactions,
and responses to stimuli95. High-resolution optical microscopy has been used to study bacterial motility at the
individual cell level, allowing for the measurement of the speed, direction, and trajectory of bacterial movement.
These techniques have provided insights into the mechanisms of bacterial motility and the behavior of bacterial
populations in response to environmental factors.
Representative studies using high-resolution optical microscopy for measuring the locomotion of individual
bacteria include the use of phase contrast microscopy to study the behavior of flagellar motors92,96,97, the use of
dark-field microscopy to observe the movements of bacteria in response to chemical stimuli, and the use of high-
speed fluorescence microscopy visualize flagellar motion98,99. To identify the bacterial motion in a real
environment, several studies analyze super-diffusive properties in colloidal media100,101 or colonies102. Phase
contrast microscopy and QPI methods have also been utilized to extract the longitudinal position of bacteria94,97,103-
105 (Fig. 3b). These techniques have provided valuable information on the behavior of individual bacteria and their
responses to various stimuli, providing insights into the complex interactions and dynamics of bacterial
populations.
3.3 Bacterial morphology and biofilm formation
Fig. 4 (a) Light-sheet microscopy reveals single-cell morphology under cytoplasmic punctum staining
enabling tracking of each cell in colony formation. (b) Deep learning was applied for single bacterial
cell morphology tracking in light-sheet microscopy. (c) Metabolic status from autofluorescence signals
display the heterogenous feature in a monoclonal group. (a) is adapted from Ref.106. (b) is adapted from
Ref.107. (c) is adapted from Ref.108.
Three-dimensional (3D) single-cell imaging has revolutionized the study of bacterial morphology and behaviors.
For instance, 3D structured illumination microscopy (SIM) has revealed a torus-like topological structure of the
circular chromosome inside living Escherichia coli, known as heterogeneous109.
Biofilms, the macroscopic multicellular communities formed by bacteria, represent another critical area of
microbiological research. The aggregate of bacteria provides shelter from harsh environments or agents, and the
matrix is optimal for capturing nutrients for their growth6. The safe microbial shelter challenge to solve biofouling

12
and biocorrosion issues4 or bacterial disease110. Deciphering the successful cooperative behavior of bacteria may
also provide effective countermeasures against bacteria. Biofilms have been studied using volumetric fluorescence
microscopy techniques that have recently circumvented the limitations of conventional 2D fluorescence
microscopy.
Fluorescence microscopy is a prevalent tool for biofilm imaging, allowing optical sectioning microscopy with
3D resolution to reduce noise from neighboring bacteria outside the focal plane. The technological advances
enabled quantitative analysis of 3D biofilm dynamics111. Spinning disk confocal microscopy has been used to
track thousands of cells112 and the disrupted collectivity of biofilms upon antibiotic treatment113. Similarly, light-
sheet fluorescence microscopy has revealed the fountain-like flow of bacterial communities during biofilm
formation106 (Fig. 4a). Cell counting and segmentation of light-sheet fluorescence images of living biofilms can
be effectively carried out using a deep learning-based image analysis workflow107 (Fig. 4b). Autofluorescence
studies have identified the metabolic activities of individual bacteria while forming biofilms108, providing insight
into the mechanism of biofilm development (Fig. 4c). Bimodal pigmentation capacity in Bacillus pumilus SF214
has been found to exhibit two types of cells, each group contributing differently to biofilm formation. Studying
cooperative dynamics at a single-cell resolution will provide better understanding of the mechanisms of biofilm
development114,115.
3.4 Monitoring cellular dynamics
Modifying the genetic composition and expression profiles of bacterial species to monitor their metabolic
pathways is of great interest in microbiology. Bacteria can convert carbon and nitrogen sources into various
intracellular and extracellular biopolymers, so regulating or designing this cell factory can have benefits such as
pathogenesis modulation and material production116. PHA biopolymer accumulation processes were compared in
two different recombinant species117.
3.5 Image-based AST
AST is routinely performed in clinical microbiology laboratories to guide selection of appropriate antibiotics.
Conventional AST methods including the disk diffusion test, Epsilometer test, and broth dilution test are simple,
but they require considerable time for growth feature inspection under the presence of antibiotics. High-resolution
imaging tools, on the other hand, can enable the observation of phenotypic changes of bacterial cells at a single-
cell level within a few doubling times. By monitoring bacterial growth from the very beginning, the culturing time
needed to obtain AST results can be reduced.
Various microscopic imaging approaches have been used to provide rapid AST results. Fluorescence
microscopy is an imaging technique widely used for rapid AST. Lu et al. demonstrated AST at a single-cell level
by analyzing bacterial growth using time-lapse fluorescence images of individual bacteria loaded into
microchannels under various antibiotic environments84,118 (Fig. 5a). Mohan et al. proposed a similar approach to
develop a microfluidic biosensor platform119. They fabricated a microfluidic chip in the form of an array in which
both the antibiotics and the concentration were varied, and fluorescence images were used to monitor cell growth
and death.

13
Fig. 5 High-resolution imaging techniques used for antimicrobial susceptibility testing (AST). (a)
Single-cell AST performed on E. coli in microchannels provides rapid determination of antimicrobial-
resistant profiles. (b) Growth rate comparison between single-cell traps with and without antibiotics
enables fast AST in less than 30 minutes. (c) The disc agarose channel system is validated for drug
susceptibility testing of M. tuberculosis strains. (d) Time-lapse 3D imaging through optical diffraction
tomography reveals the response of B. subtilis to ampicillin treatment. (a) is adapted from Ref.84,118. (b)
is adapted from Ref.120. (c) is adapted from Ref.121,122. (d) is adapted from Ref.74.
Recently, Label-free imaging methods, including bright-field microscopy, have also been used to assess
antibiotic susceptibility in conjunction with similar micro-loading platforms. Matsumoto et al. assessed drug
susceptibility using bright-field images from a microfluidic agarose channel chip, with bacterial number and size
used to carry out AST in agreement with the broth dilution test123. Another AST method based on bright-field
microscopy utilized arrays of microscopic wells to efficiently and multiplexedly monitor bacteria in various
antibiotic environments. Veses-Garcia et al. seeded and monitored bacteria in over 600 conditions using a slide
of nanoliter-sized wells124. Baltekin et al. devised a microfluidic trap to monitor bacterial growth rate120 (Fig. 5b),
while Jung et al. proposed micro-culture channels to determine critical drug concentrations for Mycobacterium

14
tuberculosis121,125 (Fig. 5c). Alternatively, QPI can also characterize bacteria under antibiotic influence, with time-
lapse analysis of 3D QPI measurements indicating the effectiveness and dose of antibiotics36,74 (Fig. 4d). More
recently, a blood culture-free ultra-rapid AST platform was reported126, which bypasses traditional blood culturing
and performs susceptibility profiling directly from whole blood. This approach reduced the turnaround time by
more than 40–60 hours compared to conventional AST workflows, demonstrating significant potential for faster
clinical decision-making in critical cases such as sepsis.
4. Discussion
We explored the latest innovations in microscopy techniques and invited scientists from diverse research areas to
bring novel bioengineering tools to solve new questions in microbiology. Specialized fluorescence tagging and
label-free approaches provide endogenous optical fingerprints that allow us to observe biomolecular processes
and metabolic states of individual bacteria, respectively. High-resolution imaging methods enable the real-time
visualization of bacteria trajectories in 2D and 3D space, such as monitoring the construction process of biofilms,
where the same types of bacteria participate in constructing their habitat by taking on specialized tasks.
Understanding the mechanism and importance of functional specialization will unveil the strategy of bacteria to
survive in a harsh environment. In the clinical approach, the advance and commercialization of culture-free
identification and monitoring systems will significantly reduce the death rates due to pathogenic diseases.
Compared to other methods, such as molecular or mass spectrometric approaches, the novel image-based
optical techniques have shown competitive performance for rapidity and accuracy of bacterial identification and
AST126,127. Significant efforts have been made to improve the performance of these techniques to the level
comparable to that of the reference methods of bacterial identification and AST; however, several clinical studies
reported limitations of these techniques for identification in multi-bacterial infections and AST for specific
bacteria, such as Pseudomonas aeruginosa128. Further improvements and research might be required to meet
clinical needs.
A microfluidic approach will ultimately be a key technique for assessing and sorting a multitude of bacteria for
automated monitoring. While high-resolution microscopy provides detailed morphological traits, the small field
of view limits the throughput of the setups. A microfluidic chip platform scales down the monitoring system into
the microscope window, which relieves the throughput issues. The miniature processing unit tests the small size
of samples efficiently. The potential progress of antibiotic susceptibility tests by using a micro-loading chip has
been described118,119,123, which enables the automatic analysis of morphological changes in single bacteria under
various antimicrobial conditions129. Recently, microfluidic chips have been used to generate a gradient of
antibiotic concentration to determine the minimal inhibitory concentration130,131. The combination of high-
resolution imaging modalities with the fine fluidic controller will facilitate the development of an automated high-
throughput monitoring system.
While deep learning-based microbiology has proven useful alongside various imaging modalities, label-free
imaging particularly attracts the utilization of deep learning. The label-free image provides consistent data suitable
for deep learning by profiling bacteria without perturbations irrelevant to physiology. This is in contrast to the
label-based image contrast, which suffers from signal degradation and stochastic errors during labeling or imaging.

15
Consistency in data acquisition prevents overfitting issues, as the discrepancy between the training data and new
data can be minimized without the above-mentioned errors. This advantage in data quality highlights label-free
imaging as a favorable option in deep image-based microbiology.
One of the challenges in adopting deep learning for image analysis is the limited interpretability of its
underlying working principle. It will be essential to allow human experts to understand, review, and adjust the
neural networks to scale up deep image-based microbiology and satisfy the reliability required for a biomedical
assay. However, accounting for the network operation remains perplexing, particularly since the designs for neural
networks have become more complex for high performance over time122,132,133. Early heuristic methods to locate
the network's region of interest134,135 have been followed by a pioneering implementation of Bayesian deep
learning that demonstrated the quantification of uncertainties lying in the data and network136. Additionally,
researchers have started to suggest the need and candidates for a public platform to share and evaluate the data
and network models137. We believe that the technical advances that unveil the operation of neural networks and
the establishment of a universal platform will accelerate the extensive biomedical application of deep image-based
microbiology.
Overall, the innovations in high-resolution microscopy techniques discussed in this review have provided new
insights and tools for studying bacterial morphology, physiology, and antibiotic susceptibility. These advances
have the potential to significantly impact both basic research and clinical applications. Looking ahead, there is
great potential for further improvements and integration of these imaging modalities with microfluidic
technologies, deep learning, and other novel approaches. As the field of microbiology continues to evolve, we
believe that the continued development and application of high-resolution imaging techniques will play an
increasingly important role in advancing our understanding of bacterial biology and improving human health.
List of abbreviations
AST: antimicrobial susceptibility testing, STED: stimulated emission depletion, SIM: structured illumination
microscopy, SMLM: Single-molecule localization microscopy, FISH: Fluorescence in situ hybridization, FLIM:
fluorescence lifetime microscopy, QPI: quantitative phase imaging, 3D: Three-dimensional
Declarations
Availability of data and materials
Not applicable.
Competing interests
The authors declare no competing interest.
Funding
This work was supported by National Research Foundation of Korea (RS-2024-00442348, 2015R1A3A2066550,
2022M3H4A1A02074314), and Korea Institute for Advancement of Technology (KIAT) through the
International Cooperative R&D program (P0028463).

16
Authors’ contributions
All authors contributed to writing subsections and revised the manuscript. NYL and YKP supervised the projects.
Acknowledgments
Not applicable.
References
1 Davis, K. M. & Isberg, R. R. Defining heterogeneity within bacterial populations via single cell
approaches. Bioessays 38, 782-790 (2016).
2 Dewachter, L., Fauvart, M. & Michiels, J. Bacterial heterogeneity and antibiotic survival:
understanding and combatting persistence and heteroresistance. Molecular cell 76, 255-267
(2019).
3 Hall-Stoodley, L., Costerton, J. W. & Stoodley, P. Bacterial biofilms: from the natural
environment to infectious diseases. Nat Rev Microbiol 2, 95-108 (2004).
https://doi.org/10.1038/nrmicro821
4 De Carvalho, C. C. Marine biofilms: a successful microbial strategy with economic
implications. Frontiers in marine science 5, 126 (2018).
5 Koo, H., Allan, R. N., Howlin, R. P., Stoodley, P. & Hall-Stoodley, L. Targeting microbial
biofilms: current and prospective therapeutic strategies. Nat Rev Microbiol 15, 740-755 (2017).
https://doi.org/10.1038/nrmicro.2017.99
6 Flemming, H. C. et al. Biofilms: an emergent form of bacterial life. Nat Rev Microbiol 14, 563-
575 (2016). https://doi.org/10.1038/nrmicro.2016.94
7 Van Acker, H., Van Dijck, P. & Coenye, T. Molecular mechanisms of antimicrobial tolerance
and resistance in bacterial and fungal biofilms. Trends Microbiol 22, 326-333 (2014).
https://doi.org/10.1016/j.tim.2014.02.001
8 O'hara, C., Tenover, F. & Miller, J. Parallel comparison of accuracy of API 20E, Vitek GNI,
MicroScan Walk/Away Rapid ID, and Becton Dickinson Cobas Micro ID-E/NF for
identification of members of the family Enterobacteriaceae and common gram-negative, non-
glucose-fermenting bacilli. Journal of clinical microbiology 31, 3165-3169 (1993).
9 Moll, W. M., Ungerechts, J., Marklein, G. & Schaal, K. P. Comparison of BBL crystal® ANR
ID kit and API rapid ID 32 A for identification of anaerobic bacteria. Zentralblatt für
Bakteriologie 284, 329-347 (1996).
10 Lamy, B. et al. Accuracy of 6 commercial systems for identifying clinical Aeromonas isolates.
Diagnostic microbiology and infectious disease 67, 9-14 (2010).
11 Kitch, T. T., Jacobs, M. R. & Appelbaum, P. C. Evaluation of RapID onE system for
identification of 379 strains in the family Enterobacteriaceae and oxidase-negative, gram-
negative nonfermenters. Journal of clinical microbiology 32, 931-934 (1994).
12 Claxton, n. P. & Masterton, R. Rapid organism identification from Bactec NR blood culture
media in a diagnostic microbiology laboratory. Journal of clinical pathology 47, 796-798
(1994).
13 Guarner, J., Street, C., Matlock, M., Cole, L. & Brierre, F. Improving Gram stain proficiency
in hospital and satellite laboratories that do not have microbiology. Clin Chem Lab Med 55,
458-461 (2017). https://doi.org/10.1515/cclm-2016-0556
14 Madison, B. M. Application of stains in clinical microbiology. Biotech Histochem 76, 119-125
(2001).
15 Barghouthi, S. A. A universal method for the identification of bacteria based on general PCR
primers. Indian journal of microbiology 51, 430-444 (2011).
16 Kuch, A., Stefaniuk, E., Ozorowski, T. & Hryniewicz, W. New selective and differential
chromogenic agar medium, chromID VRE, for screening vancomycin-resistant Enterococcus
species. Journal of microbiological methods 77, 124-126 (2009).

17
17 Peterson, J. F. et al. Spectra MRSA, a new chromogenic agar medium to screen for methicillin-
resistant Staphylococcus aureus. Journal of clinical microbiology 48, 215-219 (2010).
18 Juan, C. et al. Challenges for accurate susceptibility testing, detection and interpretation of β-
lactam resistance phenotypes in Pseudomonas aeruginosa: results from a Spanish multicentre
study. Journal of Antimicrobial Chemotherapy 68, 619-630 (2013).
19 Khan, A. et al. Evaluation of the Vitek 2, Phoenix, and MicroScan for antimicrobial
susceptibility testing of Stenotrophomonas maltophilia. Journal of Clinical Microbiology 59,
e00654-00621 (2021).
20 Choi, Y. et al. Scanner-free and wide-field endoscopic imaging by using a single multimode
optical fiber. Phys Rev Lett 109, 203901 (2012).
https://doi.org/10.1103/PhysRevLett.109.203901
21 Chapin, K. C. & Musgnug, M. C. Validation of the automated reading and incubation system
with Sensititre plates for antimicrobial susceptibility testing. Journal of clinical microbiology
41, 1951-1956 (2003).
22 Ferrand, J. et al. Comparison of Vitek MS and MALDI Biotyper for identification of
Actinomycetaceae of clinical importance. Journal of Clinical Microbiology 54, 782-784 (2016).
23 Carbonnelle, E. et al. MALDI-TOF mass spectrometry tools for bacterial identification in
clinical microbiology laboratory. Clinical biochemistry 44, 104-109 (2011).
24 Nguyen, M. H. et al. Performance of the T2Bacteria panel for diagnosing bloodstream
infections: a diagnostic accuracy study. Annals of internal medicine 170, 845-852 (2019).
25 Zboromyrska, Y. et al. Evaluation of the Magicplex™ Sepsis real-time test for the rapid
diagnosis of bloodstream infections in adults. Frontiers in cellular and infection microbiology
9, 56 (2019).
26 Parize, P. et al. Untargeted next-generation sequencing-based first-line diagnosis of infection
in immunocompromised adults: a multicentre, blinded, prospective study. Clinical
Microbiology and Infection 23, 574. e571-574. e576 (2017).
27 Dark, P. et al. Accuracy of LightCycler® Septi F ast for the detection and identification of
pathogens in the blood of patients with suspected sepsis: a systematic review and meta-analysis.
Intensive care medicine 41, 21-33 (2015).
28 Stevenson, M. et al. Sepsis: the LightCycler SeptiFast Test MGRADE®, SepsiTest™ and
IRIDICA BAC BSI assay for rapidly identifying bloodstream bacteria and fungi-a systematic
review and economic evaluation. Health Technology Assessment (Winchester, England) 20, 1-
246 (2016).
29 Blaschke, A. J. et al. Rapid identification of pathogens from positive blood cultures by
multiplex polymerase chain reaction using the FilmArray system. Diagnostic microbiology and
infectious disease 74, 349-355 (2012).
30 Spencer, D. H., Sellenriek, P. & Burnham, C.-A. D. Validation and implementation of the
GeneXpert MRSA/SA blood culture assay in a pediatric setting. American journal of clinical
pathology 136, 690-694 (2011).
31 Cenci, E. et al. Accelerate Pheno™ blood culture detection system: a literature review. Future
Microbiology 15, 1595-1605 (2020).
32 Kim, J. H. et al. Direct rapid antibiotic susceptibility test (dRAST) for blood culture and its
potential usefulness in clinical practice. J Med Microbiol 67, 325-331 (2018).
https://doi.org/10.1099/jmm.0.000678
33 Alvarez-Ordonez, A., Mouwen, D., Lopez, M. & Prieto, M. Fourier transform infrared
spectroscopy as a tool to characterize molecular composition and stress response in foodborne
pathogenic bacteria. Journal of microbiological methods 84, 369-378 (2011).
34 Buzalewicz, I. et al. Integrated multi-channel optical system for bacteria characterization and
its potential use for monitoring of environmental bacteria. Biomedical Optics Express 10, 1165-
1183 (2019).
35 Tasaki, S., Nakayama, M. & Shoji, W. Self-organization of bacterial communities against
environmental pH variation: Controlled chemotactic motility arranges cell population
structures in biofilms. PloS one 12, e0173195 (2017).

18
36 Beal, J. et al. Robust estimation of bacterial cell count from optical density. Communications
biology 3, 1-29 (2020).
37 Pawley, J. Handbook of biological confocal microscopy. Vol. 236 (Springer Science &
Business Media, 2006).
38 Huisken, J., Swoger, J., Del Bene, F., Wittbrodt, J. & Stelzer, E. H. Optical sectioning deep
inside live embryos by selective plane illumination microscopy. Science 305, 1007-1009 (2004).
39 Chen, B.-C. et al. Lattice light-sheet microscopy: imaging molecules to embryos at high
spatiotemporal resolution. Science 346, 1257998 (2014).
40 Gustafsson, M. G. Nonlinear structured-illumination microscopy: wide-field fluorescence
imaging with theoretically unlimited resolution. Proceedings of the National Academy of
Sciences 102, 13081-13086 (2005).
41 Li, D. et al. Extended-resolution structured illumination imaging of endocytic and cytoskeletal
dynamics. Science 349, aab3500 (2015).
42 Hell, S. W. & Wichmann, J. Breaking the diffraction resolution limit by stimulated emission:
stimulated-emission-depletion fluorescence microscopy. Opt. Lett. 19, 780-782 (1994).
43 Morozova, K. S. et al. Far-red fluorescent protein excitable with red lasers for flow cytometry
and superresolution STED nanoscopy. Biophysical journal 99, L13-L15 (2010).
44 Hofmann, M., Eggeling, C., Jakobs, S. & Hell, S. W. Breaking the diffraction barrier in
fluorescence microscopy at low light intensities by using reversibly photoswitchable proteins.
Proceedings of the National Academy of Sciences 102, 17565-17569 (2005).
45 Betzig, E. et al. Imaging intracellular fluorescent proteins at nanometer resolution. science 313,
1642-1645 (2006).
46 Rust, M. J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical
reconstruction microscopy (STORM). Nature methods 3, 793-796 (2006).
47 Stracy, M. et al. Live-cell superresolution microscopy reveals the organization of RNA
polymerase in the bacterial nucleoid. Proceedings of the National Academy of Sciences 112,
E4390-E4399 (2015).
48 Biteen, J. S. et al. Super-resolution imaging in live Caulobacter crescentus cells using
photoswitchable EYFP. Nature methods 5, 947-949 (2008).
49 Mir, M. et al. Optical measurement of cycle-dependent cell growth. Proceedings of the
National Academy of Sciences 108, 13124-13129 (2011).
50 Badrinarayanan, A., Reyes-Lamothe, R., Uphoff, S., Leake, M. C. & Sherratt, D. J. In vivo
architecture and action of bacterial structural maintenance of chromosome proteins. Science
338, 528-531 (2012).
51 Vink, J. N. et al. Direct visualization of native CRISPR target search in live bacteria reveals
cascade DNA surveillance mechanism. Molecular Cell 77, 39-50. e10 (2020).
52 García-Bayona, L. et al. Nanaerobic growth enables direct visualization of dynamic cellular
processes in human gut symbionts. Proceedings of the National Academy of Sciences 117,
24484-24493 (2020).
53 Zernike, F. How I Discovered Phase Contrast. Science 121, 345-349 (1955).
https://doi.org/10.1126/science.121.3141.345
54 Allen, R. & David, G. The Zeiss-Nomarski differential interference equipment for transmitted-
light microscopy. Zeitschrift fur wissenschaftliche Mikroskopie und mikroskopische Technik 69,
193-221 (1969).
55 Koenig, K., Hibst, R., Meyer, H., Flemming, G. & Schneckenburger, H. in Dental applications
of lasers. 170-180 (Spie).
56 Bhartia, R. et al. Label-free bacterial imaging with deep-UV-laser-induced native fluorescence.
Applied and environmental microbiology 76, 7231-7237 (2010).
57 Datta, R., Heaster, T. M., Sharick, J. T., Gillette, A. A. & Skala, M. C. Fluorescence lifetime
imaging microscopy: fundamentals and advances in instrumentation, analysis, and applications.
Journal of biomedical optics 25, 071203 (2020).
58 Pereira, F. M. V. et al. Laser-induced fluorescence imaging method to monitor citrus greening
disease. Computers and electronics in agriculture 79, 90-93 (2011).

19
59 Hariri, L. P. et al. Ex vivo optical coherence tomography and laser-induced fluorescence
spectroscopy imaging of murine gastrointestinal tract. Comparative Medicine 57, 175-185
(2007).
60 Zumbusch, A., Holtom, G. R. & Xie, X. S. Three-Dimensional Vibrational Imaging by
Coherent Anti-Stokes Raman Scattering. Physical Review Letters 82, 4142-4145 (1999).
https://doi.org/10.1103/PhysRevLett.82.4142
61 Freudiger, C. W. et al. Label-Free Biomedical Imaging with High Sensitivity by Stimulated
Raman Scattering Microscopy. Science 322, 1857 (2008).
https://doi.org/10.1126/science.1165758
62 Hong, W. et al. Antibiotic susceptibility determination within one cell cycle at single-bacterium
level by stimulated Raman metabolic imaging. Analytical chemistry 90, 3737-3743 (2018).
63 Zhang, M. et al. Rapid determination of antimicrobial susceptibility by stimulated Raman
scattering imaging of D2O metabolic incorporation in a single bacterium. Advanced Science 7,
2001452 (2020).
64 Zhang, W. et al. Rapid antimicrobial susceptibility testing by stimulated Raman scattering
metabolic imaging and morphological deformation of bacteria. Analytica Chimica Acta 1168,
338622 (2021).
65 Li, J. et al. Three-dimensional tomographic microscopy technique with multi-frequency
combination with partially coherent illuminations. Biomedical Optics Express 9, 2526-2542
(2018). https://doi.org/10.1364/BOE.9.002526
66 Kang, I., Pang, S., Zhang, Q., Fang, N. & Barbastathis, G. Recurrent neural network reveals
transparent objects through scattering media. Optics Express 29, 5316-5326 (2021).
https://doi.org/10.1364/OE.412890
67 Jarvis, R. M., Brooker, A. & Goodacre, R. Surface-enhanced Raman spectroscopy for bacterial
discrimination utilizing a scanning electron microscope with a Raman spectroscopy interface.
Analytical Chemistry 76, 5198-5202 (2004).
68 Park, Y., Depeursinge, C. & Popescu, G. Quantitative phase imaging in biomedicine. Nature
photonics 12, 578-589 (2018).
69 Kim, G. et al. Holotomography. Nature Reviews Methods Primers 4, 51 (2024).
70 Jo, Y. et al. Angle-resolved light scattering of individual rod-shaped bacteria based on Fourier
transform light scattering. Scientific reports 4, 1-6 (2014).
71 Shin, J., Kim, G., Park, J., Lee, M. & Park, Y. Long-term label-free assessments of individual
bacteria using three-dimensional quantitative phase imaging and hydrogel-based
immobilization. Scientific Reports 13, 46 (2023).
72 Choi, S. Y., Oh, J., Jung, J., Park, Y. & Lee, S. Y. Three-dimensional label-free visualization
and quantification of polyhydroxyalkanoates in individual bacterial cell in its native state.
Proceedings of the National Academy of Sciences 118, e2103956118 (2021).
73 Kim, T. I. et al. Antibacterial Activities of Graphene Oxide–Molybdenum Disulfide
Nanocomposite Films. ACS Applied Materials & Interfaces 9, 7908-7917 (2017).
https://doi.org/10.1021/acsami.6b12464
74 Oh, J. et al. Three-dimensional label-free observation of individual bacteria upon antibiotic
treatment using optical diffraction tomography. Biomedical optics express 11, 1257-1267
(2020).
75 Jo, Y. et al. Holographic deep learning for rapid optical screening of anthrax spores. Science
advances 3, e1700606 (2017).
76 Kim, G. et al. Rapid species identification of pathogenic bacteria from a minute quantity
exploiting three-dimensional quantitative phase imaging and artificial neural network. Light:
Science & Applications 11, 190 (2022).
77 Moriarty-Craige, S. E. & Jones, D. P. Extracellular thiols and thiol/disulfide redox in
metabolism. Annu. Rev. Nutr. 24, 481-509 (2004).
78 Srinivasan, B. et al. TEER measurement techniques for in vitro barrier model systems. Journal
of laboratory automation 20, 107-126 (2015).
79 Lee, R. et al. Dielectric imaging for differentiation between cancer and inflammation in vivo.
Scientific reports 7, 13137 (2017).

20
80 Spencer, D. C. et al. A fast impedance-based antimicrobial susceptibility test. Nature
communications 11, 5328 (2020).
81 Abbott, J. et al. Multi-parametric functional imaging of cell cultures and tissues with a CMOS
microelectrode array. Lab on a Chip 22, 1286-1296 (2022).
82 Chitale, S. et al. A semiconductor 96-microplate platform for electrical-imaging based high-
throughput phenotypic screening. Nature Communications 14, 7576 (2023).
83 Incandela, J. T., Hu, K., Joshi, P., Rosenstein, J. K. & Larkin, J. W. Non-Optical, Label-free
Electrical Capacitance Imaging of Microorganisms. bioRxiv, 2024.2009. 2025.615041 (2024).
84 Gey, A., Werckenthin, C., Poppert, S. & Straubinger, R. K. Identification of pathogens in
mastitis milk samples with fluorescent in situ hybridization. J Vet Diagn Invest 25, 386-394
(2013). https://doi.org/10.1177/1040638713486113
85 Bhattacharjee, A., Datta, R., Gratton, E. & Hochbaum, A. I. Metabolic fingerprinting of bacteria
by fluorescence lifetime imaging microscopy. Scientific reports 7, 1-10 (2017).
86 Singhal, N., Kumar, M., Kanaujia, P. K. & Virdi, J. S. MALDI-TOF mass spectrometry: an
emerging technology for microbial identification and diagnosis. Frontiers in microbiology 6,
791 (2015).
87 Amann, R. & Fuchs, B. M. Single-cell identification in microbial communities by improved
fluorescence in situ hybridization techniques. Nature Reviews Microbiology 6, 339-348 (2008).
88 Frickmann, H. et al. Fluorescence in situ hybridization (FISH) in the microbiological diagnostic
routine laboratory: a review. Critical reviews in microbiology 43, 263-293 (2017).
89 Müller, V. et al. Identification of pathogenic bacteria in complex samples using a smartphone
based fluorescence microscope. RSC advances 8, 36493-36502 (2018).
90 Patino, S. et al. Autofluorescence of mycobacteria as a tool for detection of Mycobacterium
tuberculosis. Journal of Clinical Microbiology 46, 3296-3302 (2008).
91 Jo, Y. et al. Label-free identification of individual bacteria using Fourier transform light
scattering. Optics express 23, 15792-15805 (2015).
92 Son, K., Guasto, J. S. & Stocker, R. Bacteria can exploit a flagellar buckling instability to
change direction. Nature physics 9, 494-498 (2013).
93 Hillman, T. R. et al. Digital optical phase conjugation for delivering two-dimensional images
through turbid media. Scientific Reports 3, 1-5 (2013). https://doi.org/10.1038/srep01909
94 Taute, K., Gude, S., Tans, S. & Shimizu, T. High-throughput 3D tracking of bacteria on a
standard phase contrast microscope. Nature communications 6, 1-9 (2015).
95 Son, K., Brumley, D. R. & Stocker, R. Live from under the lens: exploring microbial motility
with dynamic imaging and microfluidics. Nature Reviews Microbiology 13, 761-775 (2015).
96 Vissers, T. et al. Dynamical analysis of bacteria in microscopy movies. Plos one 14, e0217823
(2019).
97 Acres, J. & Nadeau, J. 2D vs 3D tracking in bacterial motility analysis. Biophysics (2021).
98 Hyon, Y., Powers, T. R., Stocker, R. & Fu, H. C. The wiggling trajectories of bacteria. Journal
of Fluid Mechanics 705, 58-76 (2012).
99 Lee, K. C. et al. Quantitative phase imaging flow cytometry for ultra‐large‐scale single‐cell
biophysical phenotyping. Cytometry Part A 95, 510-520 (2019).
100 Kamdar, S. et al. The colloidal nature of complex fluids enhances bacterial motility. Nature
603, 819-823 (2022).
101 Makarchuk, S., Braz, V. C., Araújo, N. A., Ciric, L. & Volpe, G. Enhanced propagation of
motile bacteria on surfaces due to forward scattering. Nature communications 10, 1-12 (2019).
102 Ariel, G. et al. Swarming bacteria migrate by Lévy Walk. Nature communications 6, 1-6 (2015).
103 Cheong, F. C. et al. Rapid, high-throughput tracking of bacterial motility in 3D via phase-
contrast holographic video microscopy. Biophysical journal 108, 1248-1256 (2015).
104 Molaei, M. & Sheng, J. Imaging bacterial 3D motion using digital in-line holographic
microscopy and correlation-based de-noising algorithm. Optics express 22, 32119-32137
(2014).
105 Jericho, S. et al. In-line digital holographic microscopy for terrestrial and exobiological
research. Planetary and Space Science 58, 701-705 (2010).

21
106 Qin, B. et al. Cell position fates and collective fountain flow in bacterial biofilms revealed by
light-sheet microscopy. Science 369, 71-77 (2020).
107 Zhang, M. et al. Non-invasive single-cell morphometry in living bacterial biofilms. Nature
communications 11, 1-13 (2020).
108 Manzo, N. et al. Pigmentation and sporulation are alternative cell fates in Bacillus pumilus
SF214. PLoS One 8, e62093 (2013).
109 Wu, F. et al. Direct imaging of the circular chromosome in a live bacterium. Nature
communications 10, 1-9 (2019).
110 Dewasthale, S., Mani, I. & Vasdev, K. Microbial biofilm: current challenges in health care
industry. J Appl Biotechnol Bioeng 5, 160-164 (2018).
111 Hartmann, R. et al. Quantitative image analysis of microbial communities with BiofilmQ.
Nature microbiology 6, 151-156 (2021).
112 Basu, R. et al. Cytotoxic T Cells Use Mechanical Force to Potentiate Target Cell Killing. Cell
165, 100-110 (2016). https://doi.org/10.1016/j.cell.2016.01.021
113 Díaz-Pascual, F. et al. Breakdown of Vibrio cholerae biofilm architecture induced by
antibiotics disrupts community barrier function. Nature microbiology 4, 2136-2145 (2019).
114 Giri, S., Waschina, S., Kaleta, C. & Kost, C. Defining division of labor in microbial
communities. Journal of molecular biology 431, 4712-4731 (2019).
115 Momeni, B. Division of labor: how microbes split their responsibility. Current Biology 28,
R697-R699 (2018).
116 Moradali, M. F. & Rehm, B. H. A. Bacterial biopolymers: from pathogenesis to advanced
materials. Nature Reviews Microbiology 18, 195-210 (2020). https://doi.org/10.1038/s41579-
019-0313-3
117 Choi, S. Y., Oh, J., Jung, J., Park, Y. & Lee, S. Y. Three-dimensional label-free visualization
and quantification of polyhydroxyalkanoates in individual bacterial cell in its native state.
Proceedings of the National Academy of Sciences 118 (2021).
118 Lu, Y. et al. Single cell antimicrobial susceptibility testing by confined microchannels and
electrokinetic loading. Analytical chemistry 85, 3971-3976 (2013).
119 Mohan, R. et al. A multiplexed microfluidic platform for rapid antibiotic susceptibility testing.
Biosensors and Bioelectronics 49, 118-125 (2013).
120 Baltekin, Ö., Boucharin, A., Tano, E., Andersson, D. I. & Elf, J. Antibiotic susceptibility testing
in less than 30 min using direct single-cell imaging. Proceedings of the National Academy of
Sciences 114, 9170-9175 (2017).
121 Jung, Y.-G. et al. A rapid culture system uninfluenced by an inoculum effect increases
reliability and convenience for drug susceptibility testing of Mycobacterium tuberculosis.
Scientific reports 8, 1-11 (2018).
122 Ban, S. et al. Optical properties of acute kidney injury measured by quantitative phase imaging.
Biomedical optics express 9, 921-932 (2018).
123 Matsumoto, Y. et al. A microfluidic channel method for rapid drug-susceptibility testing of
Pseudomonas aeruginosa. PloS one 11, e0148797 (2016).
124 Veses-Garcia, M. et al. Rapid phenotypic antibiotic susceptibility testing of uropathogens using
optical signal analysis on the nanowell slide. Frontiers in microbiology 9, 1530 (2018).
125 Jung, Y. G. et al. A rapid culture system uninfluenced by an inoculum effect increases reliability
and convenience for drug susceptibility testing of Mycobacterium tuberculosis. Sci Rep 8, 8651
(2018). https://doi.org/10.1038/s41598-018-26419-z
126 Kim, T. H. et al. Blood culture-free ultra-rapid antimicrobial susceptibility testing. Nature 632,
893-902 (2024).
127 Dare, R. K. et al. Clinical impact of Accelerate Pheno rapid blood culture detection system in
bacteremic patients. Clinical Infectious Diseases 73, e4616-e4626 (2021).
128 Descours, G. et al. Evaluation of the Accelerate Pheno™ system for rapid identification and
antimicrobial susceptibility testing of Gram-negative bacteria in bloodstream infections.
European Journal of Clinical Microbiology & Infectious Diseases 37, 1573-1583 (2018).
129 Zhang, K., Qin, S., Wu, S., Liang, Y. & Li, J. Microfluidic systems for rapid antibiotic
susceptibility tests (ASTs) at the single-cell level. Chemical science 11, 6352-6361 (2020).

22
130 Kim, S., Lee, S., Kim, J.-K., Chung, H. J. & Jeon, J. S. Microfluidic-based observation of local
bacterial density under antimicrobial concentration gradient for rapid antibiotic susceptibility
testing. Biomicrofluidics 13 (2019).
131 Kim, K., Kim, S. & Jeon, J. S. Visual estimation of bacterial growth level in microfluidic culture
systems. Sensors 18, 447 (2018).
132 He, K., Zhang, X., Ren, S. & Sun, J. in Proceedings of the IEEE conference on computer vision
and pattern recognition. 770-778.
133 Huang, G., Liu, Z., Van Der Maaten, L. & Weinberger, K. Q. in Proceedings of the IEEE
conference on computer vision and pattern recognition. 4700-4708.
134 Zhou, B., Khosla, A., Lapedriza, A., Oliva, A. & Torralba, A. in Proceedings of the IEEE
conference on computer vision and pattern recognition. 2921-2929.
135 Selvaraju, R. R. et al. in Proceedings of the IEEE international conference on computer vision.
618-626.
136 Kendall, A. & Gal, Y. What uncertainties do we need in bayesian deep learning for computer
vision? Advances in neural information processing systems 30 (2017).
137 Spahn, C. et al. DeepBacs for multi-task bacterial image analysis using open-source deep
learning approaches. Communications biology 5, 1-18 (2022).