Adaptive rewiring: a general principle for neural network development
Frontiers in Network Physiology,
Год журнала:
2024,
Номер
4
Опубликована: Окт. 29, 2024
The
nervous
system,
especially
the
human
brain,
is
characterized
by
its
highly
complex
network
topology.
neurodevelopment
of
some
features
has
been
described
in
terms
dynamic
optimization
rules.
We
discuss
principle
adaptive
rewiring,
i.e.,
reorganization
a
according
to
intensity
internal
signal
communication
as
measured
synchronization
or
diffusion,
and
recent
generalization
for
applications
directed
networks.
These
have
extended
rewiring
from
oversimplified
networks
more
neurally
plausible
ones.
Adaptive
captures
all
key
brain
topology:
it
transforms
initially
random
regular
into
with
modular
small-world
structure
rich-club
core.
This
effect
specific
sense
that
can
be
tailored
computational
needs,
robust
does
not
depend
on
critical
regime,
flexible
parametric
variation
generates
range
variant
configurations.
Extreme
associated
at
macroscopic
level
disorders
such
schizophrenia,
autism,
dyslexia,
suggest
relationship
between
dyslexia
creativity.
cooperates
growth
interacts
constructively
spatial
organization
principles
formation
topographically
distinct
modules
structures
ganglia
chains.
At
mesoscopic
level,
enables
development
functional
architectures,
convergent-divergent
units,
sheds
light
early
divergence
convergence
in,
example,
visual
system.
Finally,
we
future
prospects
rewiring.
Язык: Английский
A statistical analysis method for probability distributions in Erdös–Rényi random networks with preferential cutting–rewiring operation
Frontiers in Network Physiology,
Год журнала:
2024,
Номер
4
Опубликована: Окт. 17, 2024
The
study
of
specific
physiological
processes
from
the
perspective
network
physiology
has
gained
recent
attention.
Modeling
global
information
integration
among
separated
functionalized
modules
in
structural
and
functional
brain
networks
is
a
central
problem.
In
this
article,
preferentially
cutting–rewiring
operation
(PCRO)
introduced
to
approximatively
describe
above
process,
which
consists
cutting
procedure
rewiring
with
preferential
constraints.
By
applying
PCRO
on
classical
Erdös–Rényi
random
(ERRN),
three
types
isolated
nodes
are
generated,
based
common
leaves
(CLs)
formed
between
two
hubs.
This
makes
initially
homogeneous
ERRN
experience
drastic
changes
become
heterogeneous.
Importantly,
statistical
analysis
method
proposed
theoretically
analyze
properties
an
PCRO.
Specifically,
probability
distributions
these
derived,
distribution
CLs
can
be
obtained
easily.
Furthermore,
validity
universality
our
have
been
confirmed
numerical
experiments.
Our
contributions
may
shed
light
new
interdisciplinary
field
complexity
science
biological
would
great
general
interest
physiology.
Язык: Английский