04 Small Networks

• rate of firing -> rate code

• timing of firing -> temporal code

• place of firing -> place code

receptive fields: contrast edges by inhibition (triangles)

• receptive fields

• lateral inhibition needed to distinguish stimulus edges/contrast

  • without lateral inhibition, response would just divergently broaden

  • with inhibition edges become even sharper

• stable action potential transmission even through divergence/convergence

• distinct pattern of neural response of a cell

  • firing rate etc

  • representing a certain stimulus

-> changes receptive field

• orientation selective

• increased contrast preserved in neural response

https://www.jneurosci.org/content/32/46/16466.short

• louder -> more frequencies respond

• direction represented in hemispheres

https://www.researchgate.net/publication/278047303_Adaptation_to_Shifted_Interaural_Time_Differences_Changes_Encoding_of_Sound_Location_in_Human_Auditory_Cortex

?

• vector sum from different sensitivities of proximate neurons

• population coding

• systematic variation of feature selectivity across cortical surface

• e.g. homunculus

• place on receptor surface -> place in the neural map

• everywhere in the brain

• within sensory systems: retinotop, tonotop, somatotop

• a cell has a preferred stimulus and fires acoording to stimulus intensitiy

• integration (averaged spike activity) required to be robust against noise

• stimulus information in firing rate

• 
  

• spike activity averaged over space

• robust against noise

• fast but low spatial resolution

• (like rate coding but average not over time but simulatnaousely at one time point with more cells)

• sensor and motor areas

• visual: moving of objects/moving hand

• spatio-temporal spike patterns

  • spike timing, milliseconds

  • also time inbetween spikes

• fast, good, spatial resolution, sensitive to noise

  • in rate coding, these timing fluctuations are noise, here it contains information

    
    

• spatio-temporal patterns

• fast, good spatial resolution, sensitive to noise -> need many cells

• place of receptor on surface -> place in map

• retinotopy, somatotopy, tonotopy

• place cells?

• grandmother cells

• reaction to certain concept

• probably true as well as other concepts

• sparse coding tries to find simpler pattern that already contains enough permustations to infer information

  • strong activation of small set of cells

• dense code can be translated into sparse code

  • sparse code in more central sensory regions

• GABA-A R highly conserved

• experiment: Cl might have been artifact, not in vivo

  • wrong SCF glucose instrad of ketones

• epilepsy

• what now?

• no pattern

• pyramidical cell might be affected by all surrounding inhibitory neurons

• but not entirely random -> inhibitory networks coupled by electrical synapses

• mouse barrell cortex for whiskers

• somatosensory somatotopic paw

• possibility of self-regulatory columns

  • neurons in columns tend to be clones (same progenitor cell)

    
    

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC33979/

• thicker than microcolumns

• ice cube model

  • vertical columns through layers

  • neurons in one column respond to a specific orientation of the stimulus/ contrast edge

https://pubmed.ncbi.nlm.nih.gov/18974855/

• 
  

• Kaschube (2010) Science

• not evolutionary -> self-organization?

• Kaschube

• hypercolumns?

• motor cortex: grabbing, defense, reaching

• segment of cerebral cortex to carry out a certain function