Automatic
characterization
of
malignant
disease
is
an
important
clinical
need
to
facilitate
early
detection
and
treatment
cancer.
A
deep
semi-supervised
transfer
learning
approach
was
developed
for
automated
whole-body
tumor
segmentation
prognosis
on
positron
emission
tomography
(PET)/computed
(CT)
scans
using
limited
annotations.
This
study
analyzed
five
datasets
consisting
408
prostate-specific
membrane
antigen
(PSMA)
PET/CT
prostate
cancer
patients
611
18F-fluorodeoxyglucose
(18F-FDG)
lung,
melanoma,
lymphoma,
head
neck,
breast
patients.
Transfer
generalized
the
task
across
PSMA
18F-FDG
PET/CT.
Imaging
measures
quantifying
molecular
burden
were
extracted
from
predicted
segmentations.
Prognostic
risk
models
evaluated
follow-up
measures,
Kaplan-Meier
survival
analysis,
response
assessment
with
prostate,
cancers,
respectively.
The
proposed
demonstrated
accurate
six
types.
Journal of Personalized Medicine,
Journal Year:
2024,
Volume and Issue:
14(3), P. 287 - 287
Published: March 7, 2024
Molecular
imaging
is
a
key
tool
in
the
diagnosis
and
treatment
of
prostate
cancer
(PCa).
Magnetic
Resonance
(MR)
plays
major
role
this
respect
with
nuclear
medicine
imaging,
particularly,
Prostate-Specific
Membrane
Antigen-based,
(PSMA-based)
positron
emission
tomography
computed
(PET/CT)
also
playing
rapidly
increasing
importance.
Another
technology
finding
growing
application
across
specifically
molecular
use
machine
learning
(ML)
artificial
intelligence
(AI).
Several
authoritative
reviews
are
available
MR-based
sparsity
PET/CT.
This
review
will
focus
on
AI
for
PCa.
It
aim
to
achieve
two
goals:
firstly,
give
reader
an
introduction
technologies
available,
secondly,
provide
overview
applied
PET/CT
The
clinical
applications
include
diagnosis,
staging,
target
volume
definition
planning,
outcome
prediction
monitoring.
ML
AL
techniques
discussed
radiomics,
convolutional
neural
networks
(CNN),
generative
adversarial
(GAN)
training
methods:
supervised,
unsupervised
semi-supervised
learning.
European Urology,
Journal Year:
2023,
Volume and Issue:
84(5), P. 491 - 502
Published: July 4, 2023
Prostate-specific
Membrane
Antigen
Reporting
and
Data
System
(PSMA-RADS)
was
introduced
for
standardized
reporting,
PSMA-RADS
version
1.0
allows
classification
of
lesions
based
on
their
likelihood
representing
a
site
prostate
cancer
PSMA-targeted
positron
emission
tomography
(PET).
In
recent
years,
this
system
has
extensively
been
investigated.
Increasing
evidence
accumulated
that
the
different
categories
reflect
actual
meanings,
such
as
true
positivity
in
4
5
lesions.
Interobserver
agreement
studies
demonstrated
high
concordance
among
broad
spectrum
68Ga-
or
18F-labeled,
PSMA-directed
radiotracers,
even
less
experienced
readers.
Moreover,
also
applied
to
challenging
clinical
scenarios
assist
decision-making,
example,
avoid
overtreatment
oligometastatic
disease.
Nonetheless,
with
an
increasing
use
1.0,
framework
shown
not
only
benefits,
but
limitations,
follow-up
assessment
locally
treated
Thus,
we
aimed
update
include
refined
set
order
optimize
lesion-level
characterization
best
decision-making
(PSMA-RADS
2.0).
Journal of Nuclear Medicine,
Journal Year:
2024,
Volume and Issue:
65(4), P. 643 - 650
Published: Feb. 29, 2024
Automatic
detection
and
characterization
of
cancer
are
important
clinical
needs
to
optimize
early
treatment.
We
developed
a
deep,
semisupervised
transfer
learning
approach
for
fully
automated,
whole-body
tumor
segmentation
prognosis
on
PET/CT.
Methods:
This
retrospective
study
consisted
611
18F-FDG
PET/CT
scans
patients
with
lung
cancer,
melanoma,
lymphoma,
head
neck
breast
408
prostate-specific
membrane
antigen
(PSMA)
prostate
cancer.
The
had
nnU-net
backbone
learned
the
task
PSMA
images
using
limited
annotations
radiomics
analysis.
True-positive
rate
Dice
similarity
coefficient
were
assessed
evaluate
performance.
Prognostic
models
imaging
measures
extracted
from
predicted
segmentations
perform
risk
stratification
based
follow-up
levels,
survival
estimation
by
Kaplan–Meier
method
Cox
regression
analysis,
pathologic
complete
response
prediction
after
neoadjuvant
chemotherapy.
Overall
accuracy
area
under
receiver-operating-characteristic
(AUC)
curve
assessed.
Results:
Our
yielded
median
true-positive
rates
0.75,
0.85,
0.87,
0.75
coefficients
0.81,
0.76,
0.83,
0.73
respectively,
task.
model
an
overall
0.83
AUC
0.86.
Patients
classified
as
low-
intermediate-
high-risk
mean
levels
18.61
727.46
ng/mL,
respectively
(P
<
0.05).
score
was
significantly
associated
univariable
multivariable
analyses
Predictive
only
pretherapy
both
pre-
posttherapy
accuracies
0.72
0.84
AUCs
respectively.
Conclusion:
proposed
demonstrated
accurate
in
across
6
types
scans.
Cancers,
Journal Year:
2024,
Volume and Issue:
16(3), P. 486 - 486
Published: Jan. 23, 2024
Early
detection
of
metastatic
prostate
cancer
(mPCa)
is
crucial.
Whilst
the
prostate-specific
membrane
antigen
(PSMA)
PET
scan
has
high
diagnostic
accuracy,
it
suffers
from
inter-reader
variability,
and
time-consuming
reporting
process.
This
systematic
review
was
registered
on
PROSPERO
(ID
CRD42023456044)
aims
to
evaluate
AI’s
ability
enhance
reporting,
diagnostics,
predictive
capabilities
for
mPCa
PSMA
scans.
Inclusion
criteria
covered
studies
using
AI
PET,
excluding
non-PSMA
tracers.
A
search
conducted
Medline,
Embase,
Scopus
inception
July
2023.
After
screening
249
studies,
11
remained
eligible
inclusion.
Due
heterogeneity
meta-analysis
precluded.
The
prediction
model
risk
bias
assessment
tool
(PROBAST)
indicated
a
low
overall
in
ten
though
only
one
incorporated
clinical
parameters
(such
as
age,
Gleason
score).
demonstrated
accuracy
(98%)
identifying
lymph
node
involvement
disease,
albeit
with
sensitivity
variation
(62–97%).
Advantages
included
distinguishing
bone
lesions,
estimating
tumour
burden,
predicting
treatment
response,
automating
tasks
accurately.
In
conclusion,
showcases
promising
enhancing
potential
scans
mPCa,
addressing
current
limitations
efficiency
variability.
Seminars in Nuclear Medicine,
Journal Year:
2023,
Volume and Issue:
54(1), P. 141 - 149
Published: June 24, 2023
Prostate-specific
membrane
antigen
(PSMA)
positron
emission
tomography/computed
tomography
(PET/CT)
has
emerged
as
an
important
imaging
technique
for
prostate
cancer.
The
use
of
PSMA
PET/CT
is
rapidly
increasing,
while
the
number
nuclear
medicine
physicians
and
radiologists
to
interpret
these
scans
limited.
Additionally,
there
variability
in
interpretation
among
readers.
Artificial
intelligence
techniques,
including
traditional
machine
learning
deep
algorithms,
are
being
used
address
challenges
provide
additional
insights
from
images.
aim
this
scoping
review
was
summarize
available
research
on
development
applications
AI
cancer
imaging.
A
systematic
literature
search
performed
PubMed,
Embase
Cinahl
according
Preferred
Reporting
Items
Systematic
Reviews
Meta-Analyses
(PRISMA)
guidelines.
total
26
publications
were
included
synthesis.
studies
focus
different
aspects
artificial
PET/CT,
detection
primary
tumor,
local
recurrence
metastatic
lesions,
lesion
classification,
tumor
quantification
prediction/prognostication.
Several
show
similar
performances
algorithms
compared
human
interpretation.
Few
tools
approved
clinical
practice.
Major
limitations
include
lack
external
validation
prospective
design.
Demonstrating
impact
utility
crucial
their
adoption
healthcare
settings.
To
take
next
step
towards
a
clinically
valuable
tool
that
provides
quantitative
data,
independent
needed
across
institutions
equipment
ensure
robustness.
Cancers,
Journal Year:
2023,
Volume and Issue:
15(13), P. 3416 - 3416
Published: June 29, 2023
Radical
prostatectomy
(RP)
is
the
main
treatment
of
prostate
cancer
(PCa).
Biochemical
recurrence
(BCR)
following
RP
remains
first
sign
aggressive
disease;
hence,
better
assessment
potential
long-term
post-RP
BCR-free
survival
crucial.
Our
study
aimed
to
evaluate
a
combined
clinical-deep
learning
(DL)
model
using
multiparametric
magnetic
resonance
imaging
(mpMRI)
for
predicting
in
PCa.
A
total
437
patients
with
PCa
who
underwent
mpMRI
followed
by
between
2008
and
2009
were
enrolled;
radiomics
features
extracted
from
T2-weighted
imaging,
apparent
diffusion
coefficient
maps,
contrast-enhanced
sequences
manually
delineating
index
tumors.
Deep
same
set
deep
neural
network
based
on
pretrained
EfficentNet-B0.
Here,
we
present
clinical
(six
variables),
model,
DL
(DLM-Deep
feature),
clinical–radiomics
(CRM-Multi),
clinical–DL
(CDLM-Deep
feature)
that
built
Cox
models
regularized
least
absolute
shrinkage
selection
operator.
We
compared
their
prognostic
performances
stratified
fivefold
cross-validation.
In
median
follow-up
61
months,
110/437
experienced
BCR.
CDLM-Deep
feature
achieved
best
performance
(hazard
ratio
[HR]
=
7.72),
DLM-Deep
(HR
4.37)
or
RM-Multi
2.67).
CRM-Multi
performed
moderately.
results
confirm
superior
our
mpMRI-derived
algorithm
over
conventional
radiomics.
Cancers,
Journal Year:
2023,
Volume and Issue:
15(19), P. 4746 - 4746
Published: Sept. 27, 2023
Prostate
cancer
is
the
most
frequent
epithelial
neoplasia
after
skin
in
men
starting
from
50
years
and
prostate-specific
antigen
(PSA)
dosage
can
be
used
as
an
early
screening
tool.
imaging
includes
several
radiological
modalities,
ranging
ultrasonography,
computed
tomography
(CT),
magnetic
resonance
to
nuclear
medicine
hybrid
techniques
such
single-photon
emission
(SPECT)/CT
positron
(PET)/CT.
Innovation
radiopharmaceutical
compounds
has
introduced
specific
tracers
with
diagnostic
therapeutic
indications,
opening
horizons
targeted
very
effective
clinical
care
for
patients
prostate
cancer.
The
aim
of
present
review
illustrate
current
knowledge
future
perspectives
medicine,
including
stand-alone
theragnostic
approaches,
management
initial
staging
advanced
disease.
Cancers,
Journal Year:
2024,
Volume and Issue:
16(10), P. 1809 - 1809
Published: May 9, 2024
Background:
The
aim
was
to
analyze
the
current
state
of
deep
learning
(DL)-based
prostate
cancer
(PCa)
diagnosis
with
a
focus
on
magnetic
resonance
(MR)
reconstruction;
PCa
detection/stratification/reconstruction;
positron
emission
tomography/computed
tomography
(PET/CT);
androgen
deprivation
therapy
(ADT);
biopsy;
associated
challenges
and
their
clinical
implications.
Methods:
A
search
PubMed
database
conducted
based
inclusion
exclusion
criteria
for
use
DL
methods
within
abovementioned
areas.
Results:
total
784
articles
were
found,
which,
64
included.
Reconstruction
prostate,
detection
stratification
cancer,
reconstruction
PET/CT,
ADT,
biopsy
analyzed
in
21,
22,
6,
7,
2,
6
studies,
respectively.
Among
studies
describing
MR-based
purposes,
datasets
field
power
3
T,
1.5
3/1.5
T
used
18/19/5,
0/1/0,
3/2/1
respectively,
6/7
analyzing
PET/CT
which
data
from
single
institution.
radiotracers,
[68Ga]Ga-PSMA-11,
[18F]DCFPyl,
[18F]PSMA-1007
5,
1,
1
study,
Only
two
that
context
DT
met
criteria.
Both
performed
single-institution
dataset
only
manual
labeling
training
data.
Three
each
biopsy,
single-
multi-institutional
datasets.
TeUS,
TRUS,
MRI
as
input
modalities
two,
three,
one
Conclusion:
models
show
promise
but
are
not
yet
ready
due
variability
methods,
labels,
evaluation
Conducting
additional
research
while
acknowledging
all
limitations
outlined
is
crucial
reinforcing
utility
effectiveness
DL-based
settings.
Frontiers in Oncology,
Journal Year:
2024,
Volume and Issue:
13
Published: Jan. 8, 2024
To
explore
the
feasibility
and
importance
of
deep
learning
(DL)
based
on
68Ga-prostate-specific
membrane
antigen
(PSMA)-11
PET/CT
in
predicting
pathological
upgrading
from
biopsy
to
radical
prostatectomy
(RP)
patients
with
prostate
cancer
(PCa).
Medical Review,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 23, 2025
Abstract
The
diagnosis
and
prognosis
of
Prostate
cancer
(PCa)
have
undergone
a
significant
transformation
with
the
advent
prostate-specific
membrane
antigen
(PSMA)-targeted
positron
emission
tomography
(PET)
imaging.
PSMA-PET
imaging
has
demonstrated
superior
performance
compared
to
conventional
methods
by
detecting
PCa,
its
biochemical
recurrence,
sites
metastasis
higher
sensitivity
specificity.
That
now
intersects
rapid
advances
in
artificial
intelligence
(AI)
–
including
emergence
generative
AI.
However,
there
are
unique
clinical
challenges
associated
that
still
need
be
addressed
ensure
continued
widespread
integration
into
care
research
trials.
Some
those
very
wide
dynamic
range
lesion
uptake,
benign
uptake
organs
may
adjacent
disease,
insufficient
large
datasets
for
training
AI
models,
as
well
artifacts
images.
Generative
e.g.,
adversarial
networks,
variational
autoencoders,
diffusion
language
models
played
crucial
roles
overcoming
many
such
across
various
modalities,
PET,
computed
tomography,
magnetic
resonance
imaging,
ultrasound,
etc.
In
this
review
article,
we
delve
potential
role
enhancing
robustness
utilization
image
analysis,
drawing
insights
from
existing
literature
while
also
exploring
current
limitations
future
directions
domain.