Malaysian Journal of Fundamental and Applied Sciences,
Journal Year:
2024,
Volume and Issue:
20(2), P. 347 - 359
Published: April 24, 2024
Semen
analysis
is
an
important
for
male
infertility
primary
investigation
and
manual
semen
a
conventional
method
to
assess
it.
Manual
has
been
revealed
with
accuracy
precision
limitations
due
noncompliance
guidelines
procedures.
Sperm
motility
concentration
are
the
main
indicators
pregnancy
conception
rate
hence
they
were
selected
parameters
prediction.
Convolutional
neural
network
(CNN)
benefited
computer
vision
application
industry
in
recent
years
widely
applied
research
tasks.
In
this
paper,
three-dimensional
CNN
(3DCNN)
was
designed
extract
motion
temporal
features,
which
vital
sperm
For
concentration,
since
two-dimensional
(2DCNN)
efficient
recognizing
extracting
spatial
well-established
Residual
Network
(ResNet)
architecture
adopted
customized
Multimodal
learning
approach
technique
aggregate
learnt
features
from
different
deep
that
other
forms
of
modalities,
could
provide
model
better
insights
on
their
Hence,
multimodal
receive
both
image-based
(frames
extracted
video
samples)
video-based
(stacked
frames
pre-processed
input
well-extracted
The
results
obtained
using
proposed
methodology
have
surpassed
similar
works
who
used
approach.
motility,
its
best
achieved
average
mean
absolute
error
(MAE)
8.048,
competent
Pearson’s
correlation
coefficient
(RP)
value
0.853.
Materials Futures,
Journal Year:
2024,
Volume and Issue:
3(3), P. 035402 - 035402
Published: July 17, 2024
Abstract
Esophageal
cancer
(EC)
is
characterized
by
high
morbidity
and
mortality,
chemotherapy
has
become
an
indispensable
means
for
comprehensive
treatment.
However,
due
to
the
limitation
of
effective
in
vitro
disease
model,
development
chemotherapeutic
agents
still
faces
great
challenges.
In
this
paper,
we
present
a
novel
tumor
spheroid
on
chip
platform
based
inverse
opal
hydrogel
scaffolds
screen
EC
With
microfluidic
emulsion
approach,
were
generated
with
tunable
organized
pores,
which
could
provide
spatial
confinement
cell
growth.
Thus,
suspended
KYSE-70
cells
successfully
form
uniform
spheroids
scaffolds.
It
was
demonstrated
that
recapitulate
3D
growth
patterns
vivo
exhibited
higher
sensitivity
compared
monolayer
cells.
Besides,
employing
into
microfluidics
construct
esophageal
chip,
device
realize
high-throughput
generation
drug
screening,
indicating
its
promising
role
development.
Malaysian Journal of Fundamental and Applied Sciences,
Journal Year:
2024,
Volume and Issue:
20(2), P. 347 - 359
Published: April 24, 2024
Semen
analysis
is
an
important
for
male
infertility
primary
investigation
and
manual
semen
a
conventional
method
to
assess
it.
Manual
has
been
revealed
with
accuracy
precision
limitations
due
noncompliance
guidelines
procedures.
Sperm
motility
concentration
are
the
main
indicators
pregnancy
conception
rate
hence
they
were
selected
parameters
prediction.
Convolutional
neural
network
(CNN)
benefited
computer
vision
application
industry
in
recent
years
widely
applied
research
tasks.
In
this
paper,
three-dimensional
CNN
(3DCNN)
was
designed
extract
motion
temporal
features,
which
vital
sperm
For
concentration,
since
two-dimensional
(2DCNN)
efficient
recognizing
extracting
spatial
well-established
Residual
Network
(ResNet)
architecture
adopted
customized
Multimodal
learning
approach
technique
aggregate
learnt
features
from
different
deep
that
other
forms
of
modalities,
could
provide
model
better
insights
on
their
Hence,
multimodal
receive
both
image-based
(frames
extracted
video
samples)
video-based
(stacked
frames
pre-processed
input
well-extracted
The
results
obtained
using
proposed
methodology
have
surpassed
similar
works
who
used
approach.
motility,
its
best
achieved
average
mean
absolute
error
(MAE)
8.048,
competent
Pearson’s
correlation
coefficient
(RP)
value
0.853.
