Nature Communications,
Journal Year:
2024,
Volume and Issue:
15(1)
Published: Jan. 2, 2024
Abstract
Droplet
microfluidics
enables
kHz
screening
of
picoliter
samples
at
a
fraction
the
cost
other
high-throughput
approaches.
However,
generating
stable
droplets
with
desired
characteristics
typically
requires
labor-intensive
empirical
optimization
device
designs
and
flow
conditions
that
limit
adoption
to
specialist
labs.
Here,
we
compile
comprehensive
droplet
dataset
use
it
train
machine
learning
models
capable
accurately
predicting
geometries
required
generate
aqueous-in-oil
oil-in-aqueous
single
double
emulsions
from
15
250
μm
rates
up
12000
Hz
for
different
fluids
commonly
used
in
life
sciences.
Blind
predictions
by
our
as-yet-unseen
fluids,
geometries,
materials
yield
accurate
results,
establishing
their
generalizability.
Finally,
an
easy-to-use
design
automation
tool
within
3
(<8%)
diameter,
facilitating
tailored
droplet-based
platforms
accelerating
utility
Advanced Science,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Feb. 8, 2025
Tailored
droplet
generation
is
crucial
for
microfluidics
that
involve
samples
of
varying
sizes.
However,
the
absence
precise
predictive
models
forces
platforms
to
rely
on
empiricism
derived
from
extensive
experiments,
underscoring
need
comprehensive
modeling
analysis.
To
address
this,
a
novel
customized
assembled
centrifugal
step
emulsifier
(CASE)
presented
by
incorporating
"jigsaw
puzzles"
design
efficiently
acquire
large-scale
experimental
data.
Numerical
simulations
are
utilized
analyze
fluid
configurations
during
emulsification,
identifying
key
connection
tube
determines
size.
By
training
and
verifying
with
simulation
datasets,
theoretical
model
established
allows
preliminary
size
frequency
an
average
error
rate
4.8%,
successfully
filling
critical
gap
in
existing
field.
This
empowers
CASE
achieve
all-in-one
functionality,
including
pre-design,
generation,
manipulation,
on-site
detection.
As
proof
concept,
multiscale
sample
analysis
ranging
nanoscale
nucleic
acids
microscale
bacteria
3D
cell
spheroids
realized
CASE.
In
summary,
this
platform
offers
valuable
guidance
emulsifiers
promotes
adoption
biochemical
assays.
Advanced Science,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 1, 2025
Abstract
Microfluidic
droplets,
with
their
unique
properties
and
broad
applications,
are
essential
in
chemical,
biological,
materials
synthesis
research.
Despite
the
flourishing
studies
on
artificial
intelligence‐accelerated
microfluidics,
most
research
efforts
have
focused
upstream
design
phase
of
microfluidic
systems.
Generating
user‐desired
droplets
still
remains
laborious,
inefficient,
time‐consuming.
To
address
long‐standing
challenges
associated
accurate
efficient
identification,
sorting,
analysis
morphology
generation
rate
single
double
emulsion
a
novel
machine
vision
approach
utilizing
deformable
detection
transformer
(DETR)
algorithm
is
proposed.
This
method
enables
rapid
precise
(detection
relative
error
<
4%
precision
>
94%)
across
various
scales
scenarios,
including
real‐world
simulated
environments.
identification
(MDIA),
web‐based
tool
powered
by
Deformable
DETR,
which
supports
transfer
learning
to
enhance
accuracy
specific
user
scenarios
developed.
MDIA
characterizes
diameter,
number,
frequency,
other
parameters.
As
more
training
data
added
users,
MDIA's
capability
universality
expand,
contributing
comprehensive
database
for
droplet
microfluidics.
The
work
highlights
potential
intelligence
advancing
regulation,
fabrication,
label‐free
analysis,
accelerating
biochemical
sciences
engineering.
Lab on a Chip,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 1, 2025
The
proposed
Bayesian
optimization-based
approach
enhances
micromixer
performance
by
optimizing
geometric
parameters,
significantly
reducing
required
number
of
simulations,
and
accelerating
the
design
process
compared
to
conventional
methods.
Biosensors,
Journal Year:
2024,
Volume and Issue:
14(12), P. 613 - 613
Published: Dec. 13, 2024
Microfluidic
devices
have
revolutionized
biosensing
by
enabling
precise
manipulation
of
minute
fluid
volumes
across
diverse
applications.
This
review
investigates
the
incorporation
machine
learning
(ML)
into
design,
fabrication,
and
application
microfluidic
biosensors,
emphasizing
how
ML
algorithms
enhance
performance
improving
design
accuracy,
operational
efficiency,
management
complex
diagnostic
datasets.
Integrating
microfluidics
with
has
fostered
intelligent
systems
capable
automating
experimental
workflows,
real-time
data
analysis,
supporting
informed
decision-making.
Recent
advances
in
health
diagnostics,
environmental
monitoring,
synthetic
biology
driven
are
critically
examined.
highlights
transformative
potential
ML-enhanced
systems,
offering
insights
future
trajectory
this
rapidly
evolving
field.
Lab on a Chip,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 1, 2025
The
human
microbiome
is
vital
for
health.
Droplet
microfluidics
offers
a
versatile
toolbox
research,
enabling
single-cell
sequencing,
cultivation,
and
functional
analyses
to
deepen
our
understanding
drive
innovations.
Microsystems & Nanoengineering,
Journal Year:
2025,
Volume and Issue:
11(1)
Published: Jan. 14, 2025
Abstract
Droplet
microfluidics
enable
high-throughput
screening,
sequencing,
and
formulation
of
biological
chemical
systems
at
the
microscale.
Such
devices
are
generally
fabricated
in
a
soft
polymer
such
as
polydimethylsiloxane
(PDMS).
However,
developing
design
masks
for
PDMS
can
be
slow
expensive
process,
requiring
an
internal
cleanroom
facility
or
using
external
vendor.
Here,
we
present
first
complete
droplet-based
component
library
low-cost
rapid
prototyping
electrode
integration.
This
fabrication
method
droplet
microfluidic
costs
less
than
$12
per
device
full
design-build-test
cycle
completed
within
day.
Discrete
components
generation,
re-injection,
picoinjection,
anchoring,
fluorescence
sensing,
sorting
were
built
characterized.
These
biocompatible,
low-cost,
high-throughput.
To
show
its
ability
to
perform
multistep
workflows,
these
used
assemble
“pixel"
arrays,
where
droplets
generated,
sensed,
sorted,
anchored
onto
grid
produce
images.
Analytical Chemistry,
Journal Year:
2025,
Volume and Issue:
unknown
Published: March 6, 2025
Double
emulsions
are
highly
structured
dispersion
systems
that
generate
double-layered
droplets.
offer
an
effective
platform
for
encapsulating
liquid
samples.
Multilayer
protection,
controlled
release
of
encapsulated
materials,
and
stability
make
double
superior
to
single
in
handling
sensitive
This
technology
is
widely
used
biology,
food
technology,
cosmetics,
environmental
sciences.
Microfluidic
emulsification
a
promising
method
producing
monodisperse
double-emulsion
droplets
with
high
encapsulation
efficiency.
Well-controlled
adjustment
the
core
size
shell
thickness
critical
applications
emulsions.
Changing
flow
rates
fluid
phases
most
straightforward
control
emulsion
sizes.
However,
double-emulsions
can
only
be
generated
within
small
range
rates.
Thus,
wide
without
changing
device
design
or
drastically
altering
properties
challenging.
Here,
we
demonstrate
facile
tunable
phases.
To
address
this
challenge,
developed
proof-of-concept
flexible
stretchable
microfluidic
capable
controlling
size,
thickness,
generation
frequency
by
adjusting
channel
dimensions
stretching
device.
We
incorporated
three
cases
assess
feasibility
process
emulsion.
demonstrated
increases
decreases
frequency.
Experimental
results
showed
∼84%
increase
volume
∼23%
applying
∼16%
strain.
innovative
approach
significantly
advances
field
droplet-based
microfluidics,
providing
on-site,
real-time
tunability
precision
reproducibility.
