NeuroImage,
Journal Year:
2021,
Volume and Issue:
232, P. 117893 - 117893
Published: Feb. 21, 2021
The
human
brain
exhibits
large-scale
spontaneous
fluctuations
that
account
for
most
of
its
total
energy
metabolism.
Independent
any
overt
function,
this
immense
ongoing
activity
likely
creates
or
maintains
a
potential
functional
reserve
to
facilitate
normal
function.
An
important
property
is
the
long-range
temporal
coherence,
which
can
be
characterized
by
resting
state
fMRI-based
entropy
mapping
(BEN),
relatively
new
method
has
gained
increasing
research
interest.
purpose
study
was
leverage
large
fMRI
and
behavioral
data
publicly
available
from
connectome
project
address
three
but
still
unknown
questions:
stability
rsfMRI-derived
BEN;
relationship
BEN
latent
reserve;
associations
neurocognition.
Our
results
showed
highly
stable
across
time;
in
default
mode
network
(DMN)
executive
control
(ECN)
related
negative
correlation
education
years;
lower
DMN/ECN
corresponds
higher
fluid
intelligence
better
task
performance.
These
suggest
temporally
trait;
may
provide
means
measure
bestows
functionality
enhanced
education.
Annual Review of Medicine,
Journal Year:
2020,
Volume and Issue:
72(1), P. 29 - 43
Published: Oct. 9, 2020
Impaired
cognition
is
common
in
many
neuropsychiatric
disorders
and
severely
compromises
quality
of
life.
Synchronous
electrophysiological
rhythms
represent
a
core
mechanism
for
sculpting
communication
dynamics
among
large-scale
brain
networks
that
underpin
its
breakdown
disorders.
Here,
we
review
an
emerging
neuromodulation
technology
called
transcranial
alternating
current
stimulation
has
shown
remarkable
early
results
rapidly
improving
various
domains
human
by
modulating
properties
rhythmic
network
synchronization.
Future
noninvasive
research
holds
promise
potentially
rescuing
activity
patterns
cognition,
setting
groundwork
the
development
drug-free,
circuit-based
therapeutics
people
with
cognitive
Frontiers in Neuroscience,
Journal Year:
2022,
Volume and Issue:
16
Published: April 18, 2022
Working
memory
is
a
fundamental
feature
of
biological
brains
for
perception,
cognition,
and
learning.
In
addition,
learning
with
working
memory,
which
has
been
show
in
conventional
artificial
intelligence
systems
through
recurrent
neural
networks,
instrumental
to
advanced
cognitive
intelligence.
However,
it
hard
endow
simple
neuron
model
understand
the
mechanisms
that
have
resulted
such
powerful
ability
at
neuronal
level.
This
article
presents
novel
self-adaptive
multicompartment
spiking
model,
referred
as
SAM,
spike-based
memory.
SAM
integrates
four
major
principles
including
sparse
coding,
dendritic
non-linearity,
intrinsic
dynamics,
spike-driven
We
first
describe
SAM's
design
explore
impacts
critical
parameters
on
its
dynamics.
then
use
build
networks
accomplish
several
different
tasks
supervised
MNIST
dataset
using
sequential
spatiotemporal
encoding,
noisy
spike
pattern
classification,
coding
during
detection,
meta-learning
applied
navigation
task
classification
task,
Our
experimental
results
highlight
energy
efficiency
robustness
these
wide
range
challenging
tasks.
The
effects
variations
are
also
explored,
hoping
offer
insight
into
underlying
brain.
attempt
integrate
capabilities
unified
single
multiple
timescale
competitive
performance
could
potentially
contribute
development
efficient
adaptive
neuromorphic
computing
various
applications
from
robotics
edge
computing.
NeuroImage,
Journal Year:
2021,
Volume and Issue:
232, P. 117893 - 117893
Published: Feb. 21, 2021
The
human
brain
exhibits
large-scale
spontaneous
fluctuations
that
account
for
most
of
its
total
energy
metabolism.
Independent
any
overt
function,
this
immense
ongoing
activity
likely
creates
or
maintains
a
potential
functional
reserve
to
facilitate
normal
function.
An
important
property
is
the
long-range
temporal
coherence,
which
can
be
characterized
by
resting
state
fMRI-based
entropy
mapping
(BEN),
relatively
new
method
has
gained
increasing
research
interest.
purpose
study
was
leverage
large
fMRI
and
behavioral
data
publicly
available
from
connectome
project
address
three
but
still
unknown
questions:
stability
rsfMRI-derived
BEN;
relationship
BEN
latent
reserve;
associations
neurocognition.
Our
results
showed
highly
stable
across
time;
in
default
mode
network
(DMN)
executive
control
(ECN)
related
negative
correlation
education
years;
lower
DMN/ECN
corresponds
higher
fluid
intelligence
better
task
performance.
These
suggest
temporally
trait;
may
provide
means
measure
bestows
functionality
enhanced
education.
