GigaScience,
Journal Year:
2024,
Volume and Issue:
13
Published: Jan. 1, 2024
Abstract
Background
Deep
learning
has
revolutionized
medical
image
analysis
in
cancer
pathology,
where
it
had
a
substantial
clinical
impact
by
supporting
the
diagnosis
and
prognostic
rating
of
cancer.
Among
first
available
digital
resources
field
brain
is
glioblastoma,
most
common
fatal
At
histologic
level,
glioblastoma
characterized
abundant
phenotypic
variability
that
poorly
linked
with
patient
prognosis.
transcriptional
3
molecular
subtypes
are
distinguished
mesenchymal-subtype
tumors
being
associated
increased
immune
cell
infiltration
worse
outcome.
Results
We
address
genotype–phenotype
correlations
applying
an
Xception
convolutional
neural
network
to
discovery
set
276
hematozylin
eosin
(H&E)
slides
subtype
annotation
independent
The
Cancer
Genome
Atlas–based
validation
cohort
178
cases.
Using
this
approach,
we
achieve
high
accuracy
H&E-based
mapping
(area
under
curve
for
classical,
mesenchymal,
proneural
=
0.84,
0.81,
0.71,
respectively;
P
<
0.001)
regions
outcome
(univariable
survival
model
0.001,
multivariable
0.01).
latter
were
higher
tumor
density
(P
0.001),
cells
decreased
T-cell
0.017).
Conclusions
modify
well-known
architecture
accurately
map
spatial
distribution
predictive
outcome,
thereby
showcasing
relevance
artificial
intelligence–enabled
mining
Cancer Discovery,
Journal Year:
2024,
Volume and Issue:
14(5), P. 711 - 726
Published: March 21, 2024
Artificial
intelligence
(AI)
in
oncology
is
advancing
beyond
algorithm
development
to
integration
into
clinical
practice.
This
review
describes
the
current
state
of
field,
with
a
specific
focus
on
integration.
AI
applications
are
structured
according
cancer
type
and
domain,
focusing
four
most
common
cancers
tasks
detection,
diagnosis,
treatment.
These
encompass
various
data
modalities,
including
imaging,
genomics,
medical
records.
We
conclude
summary
existing
challenges,
evolving
solutions,
potential
future
directions
for
field.
Pharmacological Research,
Journal Year:
2024,
Volume and Issue:
unknown, P. 107534 - 107534
Published: Dec. 1, 2024
Breast
cancer
is
the
most
commonly
diagnosed
malignancy
and
fifth
leading
cause
of
deaths
worldwide.
Surgery
radiation
therapy
are
localized
therapies
for
early-stage
metastatic
breast
cancer.
The
management
determined
in
large
part
by
HER2
(human
epidermal
growth
factor
receptor
2),
HR
(hormone
receptor),
ER
(estrogen
PR
(progesterone
receptor)
status.
Our
views
evolving
as
its
molecular
hallmarks
examined,
which
now
include
immunohistochemical
markers
(ER,
PR,
HER2,
proliferation
marker
protein
Ki-67),
genomic
(BRCA1/2
PIK3CA),
immunomarkers
(tumor-infiltrating
lymphocytes
PDL1).
About
two-thirds
malignancies
HR-positive/HER2-negative;
accordingly,
endocrine-based
a
major
treatment
option
these
patients.
Hormonal
or
endocrine
includes
selective
estrogen
modulators
(SERMs)
such
raloxifene,
tamoxifen
toremifene,
estrogen-receptor
degraders
(SERDs)
including
elacestrant
fulvestrant,
aromatase
inhibitors
anastrozole,
letrozole,
exemestane.
A
variety
cytotoxic
chemotherapeutic
agents
used
to
treat
HR-negative
These
taxanes
(docetaxel,
nab-paclitaxel,
paclitaxel),
anthracyclines
(doxorubicin,
epirubicin),
anti-metabolites
(capecitabine,
gemcitabine,
fluorouracil,
methotrexate),
alkylating
(carboplatin,
cisplatin,
cyclophosphamide),
drugs
that
target
microtubules
(eribulin,
ixabepilone,
ado-trastuzumab
emtansine).
Patients
with
ER-positive
tumors
treated
5-10
years
chemotherapy.
For
patients
cancer,
standard
first-line
follow-up
options
targeted
approaches
CDK4/6
inhibitors,
PI3K
PARP
anti-PDL1
immunotherapy,
depending
on
tumor
type
profile.
Algorithms,
Journal Year:
2024,
Volume and Issue:
17(6), P. 254 - 254
Published: June 10, 2024
The
introduction
of
machine
learning
in
digital
pathology
has
deeply
impacted
the
field,
especially
with
advent
whole
slide
image
(WSI)
analysis.
In
this
review,
we
tried
to
elucidate
role
algorithms
diagnostic
precision,
efficiency,
and
reproducibility
results.
First,
discuss
some
most
used
tools,
including
QuPath,
HistoQC,
HistomicsTK,
provide
an
updated
overview
approaches
their
application
pathology.
Later,
report
how
these
tools
may
simplify
automation
WSI
analyses,
also
reducing
manual
workload
inter-observer
variability.
A
novel
aspect
review
is
its
focus
on
open-source
presented
a
way
that
help
adoption
process
for
pathologists.
Furthermore,
highlight
major
benefits
technologies,
aim
making
practical
guide
clinicians
seeking
implement
learning-based
solutions
specific
workflows.
Moreover,
emphasizes
crucial
limitations
related
data
quality
interpretability
models,
giving
insight
into
future
directions
research.
Overall,
work
tries
bridge
gap
between
more
recent
technological
progress
computer
science
traditional
clinical
practice,
supporting
broader,
yet
smooth,
Frontiers in Pharmacology,
Journal Year:
2025,
Volume and Issue:
15
Published: Jan. 10, 2025
Breast
cancer
is
the
most
commonly
diagnosed
worldwide.
Metal
metabolism
pivotal
for
regulating
cell
fate
and
drug
sensitivity
in
breast
cancer.
Iron
copper
are
essential
metal
ions
critical
maintaining
cellular
function.
The
accumulation
of
iron
triggers
distinct
death
pathways,
known
as
ferroptosis
cuproptosis,
respectively.
Ferroptosis
characterized
by
iron-dependent
lipid
peroxidation,
while
cuproptosis
involves
copper-induced
oxidative
stress.
They
increasingly
recognized
promising
targets
development
anticancer
drugs.
Recently,
compelling
evidence
demonstrated
that
interplay
between
plays
a
crucial
role
progression.
This
review
elucidates
converging
pathways
Moreover,
we
examined
value
genes
associated
with
clinical
diagnosis
treatment
cancer,
mainly
outlining
potential
co-targeting
approach.
Lastly,
delve
into
current
challenges
limitations
this
strategy.
In
general,
offers
an
overview
interaction
offering
valuable
perspectives
further
research
treatment.
