Journal of Microscopy,
Journal Year:
2023,
Volume and Issue:
295(1), P. 61 - 82
Published: June 3, 2023
Abstract
Images
are
at
the
core
of
most
modern
biological
experiments
and
used
as
a
major
source
quantitative
information.
Numerous
algorithms
available
to
process
images
make
them
more
amenable
be
measured.
Yet
nature
output
that
is
useful
for
given
experiment
uniquely
dependent
upon
question
being
investigated.
Here,
we
discuss
3
main
types
information
can
extracted
from
microscopy
data:
intensity,
morphology,
object
counts
or
categorical
labels.
For
each,
describe
where
they
come
from,
how
measured,
what
may
affect
relevance
these
measurements
in
downstream
data
analysis.
Acknowledging
makes
measurement
‘good’
ultimately
down
investigated,
this
review
aims
providing
readers
with
toolkit
challenge
quantify
their
own
critical
conclusions
drawn
bioimage
analysis
experiments.
Medical Image Analysis,
Journal Year:
2022,
Volume and Issue:
79, P. 102470 - 102470
Published: May 4, 2022
With
an
increase
in
deep
learning-based
methods,
the
call
for
explainability
of
such
methods
grows,
especially
high-stakes
decision
making
areas
as
medical
image
analysis.
This
survey
presents
overview
eXplainable
Artificial
Intelligence
(XAI)
used
A
framework
XAI
criteria
is
introduced
to
classify
analysis
methods.
Papers
on
techniques
are
then
surveyed
and
categorized
according
anatomical
location.
The
paper
concludes
with
outlook
future
opportunities
Nature Methods,
Journal Year:
2023,
Volume and Issue:
20(7), P. 1010 - 1020
Published: May 18, 2023
Abstract
The
Cell
Tracking
Challenge
is
an
ongoing
benchmarking
initiative
that
has
become
a
reference
in
cell
segmentation
and
tracking
algorithm
development.
Here,
we
present
significant
number
of
improvements
introduced
the
challenge
since
our
2017
report.
These
include
creation
new
segmentation-only
benchmark,
enrichment
dataset
repository
with
datasets
increase
its
diversity
complexity,
silver
standard
corpus
based
on
most
competitive
results,
which
will
be
particular
interest
for
data-hungry
deep
learning-based
strategies.
Furthermore,
up-to-date
leaderboards,
in-depth
analysis
relationship
between
performance
state-of-the-art
methods
properties
annotations,
two
novel,
insightful
studies
about
generalizability
reusability
top-performing
methods.
provide
critical
practical
conclusions
both
developers
users
traditional
machine
algorithms.
Abstract
In
this
review,
we
explore
the
growing
role
of
artificial
intelligence
(AI)
in
advancing
biomedical
applications
human
pluripotent
stem
cell
(hPSC)‐derived
organoids.
Stem
cell‐derived
organoids,
these
miniature
organ
replicas,
have
become
essential
tools
for
disease
modeling,
drug
discovery,
and
regenerative
medicine.
However,
analyzing
vast
intricate
datasets
generated
from
organoids
can
be
inefficient
error‐prone.
AI
techniques
offer
a
promising
solution
to
efficiently
extract
insights
make
predictions
diverse
data
types
microscopy
images,
transcriptomics,
metabolomics,
proteomics.
This
review
offers
brief
overview
organoid
characterization
fundamental
concepts
while
focusing
on
comprehensive
exploration
organoid‐based
modeling
evaluation.
It
provides
into
future
possibilities
enhancing
quality
control
fabrication,
label‐free
recognition,
three‐dimensional
image
reconstruction
complex
structures.
presents
challenges
potential
solutions
AI‐organoid
integration,
establishment
reliable
model
decision‐making
processes
standardization
research.
PLoS Computational Biology,
Journal Year:
2024,
Volume and Issue:
20(2), P. e1011890 - e1011890
Published: Feb. 20, 2024
Recent
advances
in
computer
vision
have
led
to
significant
progress
the
generation
of
realistic
image
data,
with
denoising
diffusion
probabilistic
models
proving
be
a
particularly
effective
method.
In
this
study,
we
demonstrate
that
can
effectively
generate
fully-annotated
microscopy
data
sets
through
an
unsupervised
and
intuitive
approach,
using
rough
sketches
desired
structures
as
starting
point.
The
proposed
pipeline
helps
reduce
reliance
on
manual
annotations
when
training
deep
learning-based
segmentation
approaches
enables
diverse
datasets
without
need
for
human
annotations.
We
trained
small
set
synthetic
reach
accuracy
levels
comparable
those
generalist
large
collection
manually
annotated
thereby
offering
streamlined
specialized
application
models.
Molecular Biology of the Cell,
Journal Year:
2021,
Volume and Issue:
32(9), P. 823 - 829
Published: April 19, 2021
Microscopy
images
are
rich
in
information
about
the
dynamic
relationships
among
biological
structures.
However,
extracting
this
complex
can
be
challenging,
especially
when
structures
closely
packed,
distinguished
by
texture
rather
than
intensity,
and/or
low
intensity
relative
to
background.
By
learning
from
large
amounts
of
annotated
data,
deep
accomplish
several
previously
intractable
bioimage
analysis
tasks.
Until
past
few
years,
however,
most
deep-learning
workflows
required
significant
computational
expertise
applied.
Here,
we
survey
new
open-source
software
tools
that
aim
make
deep-learning–based
image
segmentation
accessible
biologists
with
limited
experience.
These
take
many
different
forms,
such
as
web
apps,
plug-ins
for
existing
imaging
software,
and
preconfigured
interactive
notebooks
pipelines.
In
addition
surveying
these
tools,
overview
challenges
remain
field.
We
hope
expand
awareness
powerful
available
analysis.
