Neocortex of all mammals is built of a common circuit

broadband transmission for local circuits in the gray matter

long-distance, low-bandwidth transmission for white matter

exist only in middle layers of certain sensory cortical areas

axon of most cells does not project out of their own cortical area

axon of most cells does not project out of their own cortical area

smooth cells have a variety of forms; pyramidal cells and spiny stellate cell differ in dendritic arborization and the precision of axonal projections

commonly interpreted as evidence for random connections

however, statistical averaging says nothing about specificity of neuron-to-neuron connections; quantification is needed

there are studies suggested specific connections exist while Peters's rule still holds in overall statistics

distribution of excitatory and inhibitory neurons on their targets are nonrandom

standard cable theory predicts that the most efficient site for inhibitor synapses is he proximal region around cell body, where indeed dominated by inhibitory neurons

inhibitory synapses at distal regions are found to be more efficient to quench distal excitatory signals

space-saving strategy for making most of connections with nearby neurons

number of neurons involved falls exponentially with distance

originate from neurons in superficial layers

there is strong recurrent excitation within and between layers

excitation and inhibition are linked in tandem to remain in balance and to prevent overexciting the circuit

express the functional relationships bw the excitatory and inhibitory neurons in the different cortical layers

show how the inputs to a local region of cortex from the sensory periphery via the thalamus or other cortical areas are integrated

explain how the relative tiny numbers of thalamic synapses are effective if they act in synchrony and are amplified by recurrent excitatory circuits

limits on brain size place constraints on the number and density of neurons

allometric scaling law
white matter volume growing faster than gray matter volume as brain size increases

spatially close neurons tend to be densely and mutually interconnected, conserving wiring cost

connection probabilities are reduced as physical distance increases

metabolic cost in maintaining brain function and its rest state

cause evolutional pressure to increases energy efficiency, neuronal coding strategy that minimize energy use with action potentials, and sparse coding

efficient network allows pairs of network elements to interact along relatively short or direct communication paths

small-world attributes
nodes are more likely to link with the neighboring nodes while having short path length to any nodes

each module corresponds to a set of densely interconnected nodes while connections bw modules are sparse

dense connections within modules and weak interconnections bw then for a structural basis for functional specialization

network hubs
specific subset of nodes dedicates to communication paths bw modules; act as points of convergence and divergence for intermodular signal

Pareto optimality
the construction or operation of a given system simultaneously satisfies multiple objectives

biological fitness or optimal design are derived from trade-off bw cost minimization and efficiency maximization

e.g. derease in GABAergic inhibition and increase spontaneous and driven neural activity

neuroplasticity is responsible for compensation mechanisms during natural aging

fitness is related to modularity of the brain, with highest fitness for low modularity in children, high modularity in young adult, decreased mudularity for elderly

plasiticity within a system exhibits both enhancements and deficits in processing

e.g. visual dorsal pathway exhibits high degree of plasticity compared to ventral path, visual tasks related to dorsal path can be enhanced (as in congenital deaf person) or impaired (as in dyslexic person)