Journal of Biophotonics,
Journal Year:
2022,
Volume and Issue:
15(7)
Published: Feb. 12, 2022
We
report
a
new
method
for
the
rapid
identification
of
pathogenic
bacterial
species
at
single-cell
level
that
combines
laser
tweezers
Raman
spectroscopy
(LTRS)
with
deep
learning
(DL).
LTRS
can
accurately
measure
spectra
(scRS)
without
destroying
and
labeling
cells.
Based
on
scRS
data,
DL
rapidly
identifies
bacteria.
measured
15
bacteria
using
homemade
LTRS.
For
each
species,
approximately,
160
cells
from
three
different
patients
were
measured,
one
patient's
data
used
as
test
set,
rest
after
being
augmented
was
training
set.
A
residual
network
(ResNet)
model,
trained
achieved
an
accuracy
94.53%
Moreover,
we
applied
gradient-weighted
class
activation
mapping
to
visualize
proposed
model.
Finally,
demonstrated
advantages
ResNet
over
traditional
machine-learning
algorithms.
Nature Communications,
Journal Year:
2023,
Volume and Issue:
14(1)
Published: Jan. 4, 2023
Abstract
Biopsy
is
the
recommended
standard
for
pathological
diagnosis
of
liver
carcinoma.
However,
this
method
usually
requires
sectioning
and
staining,
well-trained
pathologists
to
interpret
tissue
images.
Here,
we
utilize
Raman
spectroscopy
study
human
hepatic
samples,
developing
validating
a
workflow
in
vitro
intraoperative
cancer.
We
distinguish
carcinoma
tissues
from
adjacent
non-tumour
rapid,
non-disruptive,
label-free
manner
by
using
combined
with
deep
learning,
which
validated
metabolomics.
This
technique
allows
detailed
identification
cancer
tissues,
including
subtype,
differentiation
grade,
tumour
stage.
2D/3D
images
unprocessed
slices
submicrometric
resolution
are
also
acquired
based
on
visualization
molecular
composition,
could
assist
boundary
recognition
clinicopathologic
diagnosis.
Lastly,
potential
portable
handheld
system
illustrated
during
surgery
real-time
Analytical Chemistry,
Journal Year:
2023,
Volume and Issue:
95(19), P. 7552 - 7559
Published: May 4, 2023
Exosomes
are
a
class
of
extracellular
vesicles
secreted
by
cells,
which
can
be
used
as
promising
noninvasive
biomarkers
for
the
early
diagnosis
and
treatment
diseases,
especially
cancer.
However,
due
to
heterogeneity
exosomes,
it
remains
grand
challenge
distinguish
accurately
reliably
exosomes
from
clinical
samples.
Herein,
we
achieve
accurate
fuzzy
discrimination
human
serum
samples
breast
cancer
cervical
through
machine
learning-based
label-free
surface-enhanced
Raman
spectroscopy
(SERS),
using
"hot
spot"
rich
3D
plasmonic
AuNPs
nanomembranes
substrates.
Due
existence
some
weak
distinguishable
SERS
fingerprint
signals
high
sensitivity
method,
analysis
precisely
identify
three
(normal
cancerous)
cell
lines,
two
different
types
without
specific
labeling
biomarkers.
The
prediction
accuracy
based
on
learning
algorithm
was
up
91.1%
lines
(H8,
HeLa,
MCF-7
cell)-derived
exosomes.
Our
model
trained
with
spectra
cell-derived
could
reach
93.3%
Furthermore,
action
mechanism
chemotherapeutic
process
cells
revealed
dynamic
monitoring
profiling
secreted.
method
would
useful
postoperative
assessment
or
other
diseases
in
future.
PhotoniX,
Journal Year:
2023,
Volume and Issue:
4(1)
Published: July 7, 2023
Abstract
Raman
spectroscopy,
as
a
label-free
optical
technology,
has
widely
applied
in
tumor
diagnosis.
Relying
on
the
different
technologies,
conventional
diagnostic
methods
can
be
used
for
diagnosis
of
benign,
malignant
and
subtypes
tumors.
In
past
3
years,
addition
to
traditional
methods,
application
artificial
intelligence
(AI)
various
technologies
based
been
developing
at
an
incredible
speed.
Based
this,
three
technical
from
single
spot
acquisition
(conventional
surface-enhanced
spectroscopy)
imaging
are
respectively
introduced
analyzed
process
these
methods.
Meanwhile,
emerging
AI
applications
within
highlighted
presented.
Finally,
challenges
limitations
existing
prospects
AI-enabled
Materials,
Journal Year:
2024,
Volume and Issue:
17(7), P. 1621 - 1621
Published: April 2, 2024
Nanomanufacturing
and
digital
manufacturing
(DM)
are
defining
the
forefront
of
fourth
industrial
revolution—Industry
4.0—as
enabling
technologies
for
processing
materials
spanning
several
length
scales.
This
review
delineates
evolution
nanomaterials
nanomanufacturing
in
age
applications
medicine,
robotics,
sensory
technology,
semiconductors,
consumer
electronics.
The
incorporation
artificial
intelligence
(AI)
tools
to
explore
nanomaterial
synthesis,
optimize
processes,
aid
high-fidelity
nanoscale
characterization
is
discussed.
paper
elaborates
on
different
machine-learning
deep-learning
algorithms
analyzing
images,
designing
nanomaterials,
nano
quality
assurance.
challenges
associated
with
application
machine-
models
achieve
robust
accurate
predictions
outlined.
prospects
incorporating
sophisticated
AI
such
as
reinforced
learning,
explainable
(XAI),
big
data
analytics
material
process
innovation,
nanosystem
integration
ACS Nano,
Journal Year:
2021,
Volume and Issue:
15(11), P. 18023 - 18036
Published: Oct. 29, 2021
Cytokine
storm,
known
as
an
exaggerated
hyperactive
immune
response
characterized
by
elevated
release
of
cytokines,
has
been
described
a
feature
associated
with
life-threatening
complications
in
COVID-19
patients.
A
critical
evaluation
cytokine
storm
and
its
mechanistic
linkage
to
requires
innovative
immunoassay
technology
capable
rapid,
sensitive,
selective
detection
multiple
cytokines
across
wide
dynamic
range
at
high-throughput.
In
this
study,
we
report
machine-learning-assisted
microfluidic
nanoplasmonic
digital
meet
the
rising
demand
for
monitoring
Specifically,
assay
was
carried
out
using
facile
one-step
sandwich
format
three
notable
features:
(i)
microarray
patterning
technique
high-throughput,
multiantibody-arrayed
biosensing
chip
fabrication;
(ii)
ultrasensitive
imaging
utilizing
100
nm
silver
nanocubes
(AgNCs)
signal
transduction;
(iii)
rapid
accurate
machine-learning-based
image
processing
method
analysis.
The
developed
allows
simultaneous
six
single
run
working
ranges
1-10,000
pg
mL-1
ultralow
limits
down
0.46-1.36
minimum
3
μL
serum
samples.
whole
can
afford
6-plex
8
different
samples
6
repeats
each
sample
total
288
sensing
spots
less
than
min.
enhanced
convolutional
neural
network
(CNN)
dramatically
shortens
time
∼6,000
fold
much
simpler
procedure
while
maintaining
high
statistical
accuracy
compared
conventional
manual
counting
approach.
validated
gold-standard
enzyme-linked
immunosorbent
(ELISA)
utilized
profiling
positive
Our
results
demonstrate
promising
practical
tool
comprehensive
characterization
patients
that
holds
great
promise
intelligent
next
generation
monitoring.
Annual Review of Analytical Chemistry,
Journal Year:
2023,
Volume and Issue:
16(1), P. 205 - 230
Published: April 17, 2023
Infrared
(IR)
spectroscopic
imaging
records
spatially
resolved
molecular
vibrational
spectra,
enabling
a
comprehensive
measurement
of
the
chemical
makeup
and
heterogeneity
biological
tissues.
Combining
this
novel
contrast
mechanism
in
microscopy
with
use
artificial
intelligence
can
transform
practice
histopathology,
which
currently
relies
largely
on
human
examination
morphologic
patterns
within
stained
tissue.
First,
review
summarizes
IR
instrumentation
especially
suited
to
analyses
its
performance,
major
trends.
Second,
an
overview
data
processing
methods
application
machine
learning
is
given,
emphasis
emerging
deep
learning.
Third,
discussion
workflows
pathology
provided,
four
categories
proposed
based
complexity
analytical
performance
needed.
Last,
set
guidelines,
termed
experimental
specifications
for
are
help
standardize
diversity
approaches
area.
