Development of simple cell spatiotemporal structure and direction selectivity through Hebbian modification

Alan B Saul, J. C. Feidler, A. Murthy, A. L. Humphrey

Research output: Contribution to journalArticle

Abstract

Purpose. In order to make direction-selective simple cells, inputs that differ in spatial and temporal response properties must converge. Hebbian models rely on correlated firing to associate afferents to a given cell, but simple cells receive inputs that appear not to be correlated. Assuming Hebbian modification of the synapses onto cortical cells, how can temporal competition within the stimulus cycle lead to associations between afferents that respond at different times? We undertook a theoretical study that suggests how experience with moving stimuli might lead to mature simple cell responses. Methods. Simulated stimuli corresponded to drifting gratings of variable temporal frequency and direction. LGN units were assigned response properties based on published data, and consisted of lagged and nonlagged ON- and OFF-center cells with a distribution of response timing and spatial position. Rectified outputs of the LGN units excited cortical units through modifiable synapses. Inhibitory and excitatory intracortical connections were integrated into some simulations as well. Initial connectivity was random, so that initial cortical responses were poorly organized. Results. Cortical units showed a wide range of behaviors but, in particular, simple-like cells with modulated responses to sinusoidal stimuli were found. Cortical units received input from multiple LGN units with different timings. Many cortical units acquired inputs that were in approximate spatiotemporal quadrature, and were thereby direction selective. Direction preference varied across the population and direction selectivity could be altered by experience. Conclusions. Direction-selective simple cells arose through temporal competition. That is, the only variation in activity was temporal modulation across the stimulus cycle. LGN cells with different timings could connect to common cortical cells if their spatial locations compensated for their timing differences, so that they fired synchronously for the cortical cell's preferred direction of motion.

Original languageEnglish (US)
JournalInvestigative Ophthalmology and Visual Science
Volume37
Issue number3
StatePublished - Feb 15 1996

Fingerprint

Synapses
Direction compound
Theoretical Models
Population

ASJC Scopus subject areas

  • Ophthalmology
  • Sensory Systems
  • Cellular and Molecular Neuroscience

Cite this

Development of simple cell spatiotemporal structure and direction selectivity through Hebbian modification. / Saul, Alan B; Feidler, J. C.; Murthy, A.; Humphrey, A. L.

In: Investigative Ophthalmology and Visual Science, Vol. 37, No. 3, 15.02.1996.

Research output: Contribution to journalArticle

@article{340ce26e412a4ed6bc8cafb567313197,
title = "Development of simple cell spatiotemporal structure and direction selectivity through Hebbian modification",
abstract = "Purpose. In order to make direction-selective simple cells, inputs that differ in spatial and temporal response properties must converge. Hebbian models rely on correlated firing to associate afferents to a given cell, but simple cells receive inputs that appear not to be correlated. Assuming Hebbian modification of the synapses onto cortical cells, how can temporal competition within the stimulus cycle lead to associations between afferents that respond at different times? We undertook a theoretical study that suggests how experience with moving stimuli might lead to mature simple cell responses. Methods. Simulated stimuli corresponded to drifting gratings of variable temporal frequency and direction. LGN units were assigned response properties based on published data, and consisted of lagged and nonlagged ON- and OFF-center cells with a distribution of response timing and spatial position. Rectified outputs of the LGN units excited cortical units through modifiable synapses. Inhibitory and excitatory intracortical connections were integrated into some simulations as well. Initial connectivity was random, so that initial cortical responses were poorly organized. Results. Cortical units showed a wide range of behaviors but, in particular, simple-like cells with modulated responses to sinusoidal stimuli were found. Cortical units received input from multiple LGN units with different timings. Many cortical units acquired inputs that were in approximate spatiotemporal quadrature, and were thereby direction selective. Direction preference varied across the population and direction selectivity could be altered by experience. Conclusions. Direction-selective simple cells arose through temporal competition. That is, the only variation in activity was temporal modulation across the stimulus cycle. LGN cells with different timings could connect to common cortical cells if their spatial locations compensated for their timing differences, so that they fired synchronously for the cortical cell's preferred direction of motion.",
author = "Saul, {Alan B} and Feidler, {J. C.} and A. Murthy and Humphrey, {A. L.}",
year = "1996",
month = "2",
day = "15",
language = "English (US)",
volume = "37",
journal = "Investigative Ophthalmology and Visual Science",
issn = "0146-0404",
publisher = "Association for Research in Vision and Ophthalmology Inc.",
number = "3",

}

TY - JOUR

T1 - Development of simple cell spatiotemporal structure and direction selectivity through Hebbian modification

AU - Saul, Alan B

AU - Feidler, J. C.

AU - Murthy, A.

AU - Humphrey, A. L.

PY - 1996/2/15

Y1 - 1996/2/15

N2 - Purpose. In order to make direction-selective simple cells, inputs that differ in spatial and temporal response properties must converge. Hebbian models rely on correlated firing to associate afferents to a given cell, but simple cells receive inputs that appear not to be correlated. Assuming Hebbian modification of the synapses onto cortical cells, how can temporal competition within the stimulus cycle lead to associations between afferents that respond at different times? We undertook a theoretical study that suggests how experience with moving stimuli might lead to mature simple cell responses. Methods. Simulated stimuli corresponded to drifting gratings of variable temporal frequency and direction. LGN units were assigned response properties based on published data, and consisted of lagged and nonlagged ON- and OFF-center cells with a distribution of response timing and spatial position. Rectified outputs of the LGN units excited cortical units through modifiable synapses. Inhibitory and excitatory intracortical connections were integrated into some simulations as well. Initial connectivity was random, so that initial cortical responses were poorly organized. Results. Cortical units showed a wide range of behaviors but, in particular, simple-like cells with modulated responses to sinusoidal stimuli were found. Cortical units received input from multiple LGN units with different timings. Many cortical units acquired inputs that were in approximate spatiotemporal quadrature, and were thereby direction selective. Direction preference varied across the population and direction selectivity could be altered by experience. Conclusions. Direction-selective simple cells arose through temporal competition. That is, the only variation in activity was temporal modulation across the stimulus cycle. LGN cells with different timings could connect to common cortical cells if their spatial locations compensated for their timing differences, so that they fired synchronously for the cortical cell's preferred direction of motion.

AB - Purpose. In order to make direction-selective simple cells, inputs that differ in spatial and temporal response properties must converge. Hebbian models rely on correlated firing to associate afferents to a given cell, but simple cells receive inputs that appear not to be correlated. Assuming Hebbian modification of the synapses onto cortical cells, how can temporal competition within the stimulus cycle lead to associations between afferents that respond at different times? We undertook a theoretical study that suggests how experience with moving stimuli might lead to mature simple cell responses. Methods. Simulated stimuli corresponded to drifting gratings of variable temporal frequency and direction. LGN units were assigned response properties based on published data, and consisted of lagged and nonlagged ON- and OFF-center cells with a distribution of response timing and spatial position. Rectified outputs of the LGN units excited cortical units through modifiable synapses. Inhibitory and excitatory intracortical connections were integrated into some simulations as well. Initial connectivity was random, so that initial cortical responses were poorly organized. Results. Cortical units showed a wide range of behaviors but, in particular, simple-like cells with modulated responses to sinusoidal stimuli were found. Cortical units received input from multiple LGN units with different timings. Many cortical units acquired inputs that were in approximate spatiotemporal quadrature, and were thereby direction selective. Direction preference varied across the population and direction selectivity could be altered by experience. Conclusions. Direction-selective simple cells arose through temporal competition. That is, the only variation in activity was temporal modulation across the stimulus cycle. LGN cells with different timings could connect to common cortical cells if their spatial locations compensated for their timing differences, so that they fired synchronously for the cortical cell's preferred direction of motion.

UR - http://www.scopus.com/inward/record.url?scp=33750192461&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750192461&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:33750192461

VL - 37

JO - Investigative Ophthalmology and Visual Science

JF - Investigative Ophthalmology and Visual Science

SN - 0146-0404

IS - 3

ER -