Current Opinion in Neurobiology,
Год журнала:
2020,
Номер
65, С. 59 - 69
Опубликована: Ноя. 1, 2020
The
brain
is
composed
of
many
functionally
distinct
areas.
This
organization
supports
distributed
processing,
and
requires
the
coordination
signals
across
Our
understanding
how
populations
neurons
in
different
areas
interact
with
each
other
still
its
infancy.
As
availability
recordings
from
large
multiple
increases,
so
does
need
for
statistical
methods
that
are
well
suited
dissecting
interrogating
these
recordings.
Here
we
review
multivariate
have
been,
or
could
be,
applied
to
this
class
By
leveraging
population
responses,
can
provide
a
rich
description
inter-areal
interactions.
At
same
time,
introduce
interpretational
challenges.
We
thus
conclude
by
discussing
interpret
outputs
further
our
Complex
cognitive
functions
such
as
working
memory
and
decision-making
require
information
maintenance
over
seconds
to
years,
from
transient
sensory
stimuli
long-term
contextual
cues.
While
theoretical
accounts
predict
the
emergence
of
a
corresponding
hierarchy
neuronal
timescales,
direct
electrophysiological
evidence
across
human
cortex
is
lacking.
Here,
we
infer
timescales
invasive
intracranial
recordings.
Timescales
increase
along
principal
sensorimotor-to-association
axis
entire
cortex,
scale
with
single-unit
within
macaques.
Cortex-wide
transcriptomic
analysis
shows
alignment
between
expression
excitation-
inhibition-related
genes,
well
genes
specific
voltage-gated
transmembrane
ion
transporters.
Finally,
are
functionally
dynamic:
prefrontal
expand
during
individual
performance,
while
cortex-wide
compress
aging.
Thus,
follow
cytoarchitectonic
gradients
relevant
for
cognition
in
both
short
long
terms,
bridging
microcircuit
physiology
macroscale
dynamics
behavior.
Nature Communications,
Год журнала:
2022,
Номер
13(1)
Опубликована: Март 1, 2022
Brain
function
relies
on
the
coordination
of
activity
across
multiple,
recurrently
connected
brain
areas.
For
instance,
sensory
information
encoded
in
early
areas
is
relayed
to,
and
further
processed
by,
higher
cortical
then
fed
back.
However,
way
which
feedforward
feedback
signaling
interact
with
one
another
incompletely
understood.
Here
we
investigate
this
question
by
leveraging
simultaneous
neuronal
population
recordings
midlevel
visual
(V1-V2
V1-V4).
Using
a
dimensionality
reduction
approach,
find
that
interactions
are
feedforward-dominated
shortly
after
stimulus
onset
feedback-dominated
during
spontaneous
activity.
The
patterns
most
correlated
were
distinct
feedforward-
periods.
These
results
suggest
rely
separate
"channels",
allows
signals
to
not
directly
affect
forward.
Trends in Neurosciences,
Год журнала:
2023,
Номер
46(7), С. 508 - 524
Опубликована: Май 8, 2023
The
rapid
and
coordinated
propagation
of
neural
activity
across
the
brain
provides
foundation
for
complex
behavior
cognition.
Technical
advances
neuroscience
subfields
have
advanced
understanding
these
dynamics,
but
points
convergence
are
often
obscured
by
semantic
differences,
creating
silos
subfield-specific
findings.
In
this
review
we
describe
how
a
parsimonious
conceptualization
state
as
fundamental
building
block
whole-brain
offers
common
framework
to
relate
findings
scales
species.
We
present
examples
diverse
techniques
commonly
used
study
states
associated
with
physiology
higher-order
cognitive
processes,
discuss
integration
them
will
enable
more
comprehensive
mechanistic
characterization
dynamics
that
crucial
survival
disrupted
in
disease.
Nature Neuroscience,
Год журнала:
2024,
Номер
27(7), С. 1349 - 1363
Опубликована: Июль 1, 2024
Abstract
Flexible
computation
is
a
hallmark
of
intelligent
behavior.
However,
little
known
about
how
neural
networks
contextually
reconfigure
for
different
computations.
In
the
present
work,
we
identified
an
algorithmic
substrate
modular
through
study
multitasking
artificial
recurrent
networks.
Dynamical
systems
analyses
revealed
learned
computational
strategies
mirroring
subtask
structure
training
task
set.
motifs,
which
are
recurring
patterns
activity
that
implement
specific
computations
dynamics,
such
as
attractors,
decision
boundaries
and
rotations,
were
reused
across
tasks.
For
example,
tasks
requiring
memory
continuous
circular
variable
repurposed
same
ring
attractor.
We
showed
dynamical
motifs
implemented
by
clusters
units
when
unit
activation
function
was
restricted
to
be
positive.
Cluster
lesions
caused
performance
deficits.
Motifs
reconfigured
fast
transfer
learning
after
initial
phase
learning.
This
work
establishes
fundamental
compositional
computation,
intermediate
between
neuron
network.
As
whole-brain
studies
simultaneously
record
from
multiple
specialized
systems,
motif
framework
will
guide
questions
specialization
generalization.
