powerpoint slides - Gestalt ReVision

From receptive fields to dynamic
activity patterns
Cees van Leeuwen
Laboratory for Perceptual Dynamics
Receptive fields in V1
Patterned activity in V1
Patterned activity in the whole brain
2 Roles for whole-brain activity
Classical Receptive field:
Body surface where stimulation
could elicit a reflex.
Sherrington (1906)
© http://www.scholarpedia.org/article/Receptive_field
Classical Receptive field:
portion of sensory space that can
elicit neuronal responses when
© http://www.scholarpedia.org/article/Receptive_field
Receptive fields in V1
Classical Receptive field:
region of visual space where a
luminous stimulus could drive
electrical responses in a retinal
ganglion cell
Hartline (1938)
© http://www.scholarpedia.org/article/Receptive_field
Classical Receptive field
in vision:
a two-dimensional region in visual
space whose size can range from a few
minutes of arc to tens of degrees
On-center and Off-center receptive
fields. The receptive fields of retinal
ganglion cells and thalamic neurons
are organized as two concentric circles
with different contrast polarities. Oncenter neurons respond to the
presentation of a light spot on a dark
background and off-center neurons to
the presentation of a dark spot on a
light background.
© http://www.scholarpedia.org/article/Receptive_field
Shapley et al. 2007
Classical Receptive Field in V1
Hubel and Wiesel (1962): First characterization of receptive fields in primary visual
cortex and classification based on receptive field structure. Some cortical cells
respond to light and dark spots in different subregions of the receptive field These
are called simple cells. All other cells (the majority) are called complex cells
(Martinez & Alonso, 2003)
© http://www.scholarpedia.org/article/Receptive_field
Receptive fields of four
primate V1 neurons (9o-20o
eccentricity). The receptive
field of each neuron was
mapped with light spots
(continuous lines, top
panels) and dark spots
(dotted lines, bottom
panels). Unlike complex cells
(c,d), simple cells (a,b)
respond to light and dark
spots in different regions of
the receptive field (Figure
taken from Chen et al.,
Receptive Fields as the cradle (and the
cage) of our thinking about perception
• Perception begins with a mosaic of features
and proceeds in a bottom-up fashion
• Segregation (functional specialization at the
level of cells and regions)
• Convergence
• Cannot explain visual experience
• More of an ideology than a science
Layers of V1
The primary visual cortex has distinct anatomical layers, each with
characteristic synaptic connections. (Adapted from Lund 1988)
Context-specific responses in V1
Yeh et al., 2009 PNAS
Example spatial maps of V1
cells in layer 4C and layer 2/3.
(A) Two simple cells and one
complex cell in layer 4C. (B)
Two simple cells and one
complex cell in layer 2/3. For
each example, the Hartley
subspace maps are drawn at
the top and the sparse-noise
maps at the bottom. Spatial
maps are shown as color
maps (grid size: 0.2°) in which
on subregions are
represented in red and off
subregions are in blue.
Spatiotemporal receptive field
Shapley et al. 2007
Spatiotemporal receptive field
A spatial receptive field plotted at different time delays between stimulus and
neuronal response. The response function will be influenced by:
•The state of the neuron prior to stimulation (ranging from habituation to
•The state of surrounding neurons in the same layer
•The state of surrounding layers
•The state of surrounding areas of the visual system
•The state of the whole brain
•Oculomotor-related effects (responses depend on whether the stimulus is
flashed or the result of a saccade; MacEvoy et al. 2007).
Patterned activity in V1
Single-cell vs Population responses
• Weak selectivity at the single-cell level can still
lead to strong responses at the population
• Complex selectivity + nonlinearity can lead to
flexibility in read-out.
Rentzeperis et al. (in press).
Spatiotemporal activation pattern of the VSDI signal.
Ayzenshtat I et al. J. Neurosci. 2012;32:13971-13986
©2012 by Society for Neuroscience
Population Response in v1
Macaque: faces/scrambled face discrimination task
The early response was
highly correlated with local luminance of the stimulus
The late response
showed a much lower correlation to the local luminance, was confined to
central parts of the face images, and was highly correlated with the animal’s
perceptual report.
“Our study reveals a continuous spatial encoding of low- and high-level
features of natural images in V1. The low level is directly linked to the
stimulus basic local attributes and the high level is correlated with the
perceptual outcome of the stimulus processing.”
Ayzenshtat, et al., 2012
Traveling Waves in V1
(A) Voltage-Sensitive Dye time
courses of an anesthetized monkey
n a 6 X 6 mm array following a small
grating presented for 250 ms.
(B) Time courses at the retinal
location of the stimulus and the
other 4.5mm away (Grinvald et al.,
(C) Spread of activity in area V1 and
V2 of awake monkey, after onset of
a small visual stimulus. The large
response is in V1 the smaller in V2
as delineated by the ocular
dominance map at bottom right.
(D) Spatial profiles measured
through axis parallel to the V1/V2
border, at different time points
(Slovin et al., 2002).
Review by Sato et al., 2012 Neuron
Traveling Waves in V1
LFP (L) and spike activity (R) in V1 (Busse et al., 2009) Cats shown rapid sequence of bars, each for 32 ms at a random position,
orientation, and spatial phase. LFP and spike responses to bars having the optimal orientation for the site.
(A) Average time course of the LFP for multiple stimulus distances from the receptive field center.
Heat map of the LFP responses
(C) Amplitude of the traces in (A) as a function of distance.
(D–F) Same as (A)–(C) but for the multiunit spike responses. Spike trains were smoothed with a Gaussian window
Review by Sato et al., 2012 Neuron
The stimulus-evoked population response in visual cortex of awake monkey is a
propagating wave
Muller, Reynaud, Chavane & Destexhe (2014) Nature Communications
(a) Single-trial phase-latency maps for V1 ROI in the 50 ms stimulus presentation condition, from trials 1, 3 and
10. These maps are calculated at +72.7 ms after stimulus onset. Note black box in top panel corresponds to V1
ROI in (a). (b) Phase-latency maps for spontaneous waves observed during no-stimulus, blank conditions. Note
the varying color axis and temporal points for the spontaneous maps.
Origin of Traveling Waves
• LGN transmission delays? Neuh! (No such systematicity)
• Top-down? Neuh! (also when animal is anesthetized)
• Horizontal connections within V1 (Bosking et al., 1997; Creutzfeldt et al., 1977; Fisken et al., 1975;
Gilbert & Wiesel, 1979; Rockland & Lund, 1982).
Propagate with the same speed as the waves do .2-.3 m/s.
Connect like with like (just like the waves do!)
Interesting Features of Traveling
Waves in V1
• They are observed in the Layers 2-3
• Facilitory
• They occur in spontaneous activity (non-rem
sleep; quiet wakefulness) and weak
• They cover large regions of space
• The more intense stimulation, the shorterrange the waves
Most interesting Feature of Traveling
Waves in V1
Sites with similar orientation preference are more strongly linked than sites with dissimilar orientation preference.
The abscissa represents the absolute difference in orientation between the reference site and the site of the spiketriggered LFP. The ordinate corresponds to residual z-score values of the amplitude after subtracting the distance
dependence predicted by the exponential fits. Plotted are the mean and standard error bars for three different bins
of orientation difference. Nauhaus et al. (2009) Nature Neurosc
Spontaneous vs Evoked
Visual stimulation
modifies the effective
lateral connectivity in the
cortex. Each row shows
the dependence of the
spike-triggered LFP
amplitudes as a pseudocolor image in
spontaneous and driven
conditions and as scatter
plots. The top two
examples correspond to
two different monkeys.
The bottom two are from
two different cats.
Nauhaus et al. 2009 Nat.
Proposed roles of Traveling Waves
• Facilitory interaction amongst stimuli
– Integration (Gilbert, 1992; Kapadia et al., 1999;
Polat et al., 1998)
– Receptive field tuning (Angelucci & Bressloff,
2006; Cavanaugh et al., 2002)
– Normalization (Carandini & Heeger, 2012).
Patterned Activity in the Whole Brain
Scalp EEG
f2 + f3
f1 +
• originates from cortical
pyramidal cells
• reflects the postsynaptic
dendritic potentials
• aggregated from 6 cm2 of
cortical gyri tissue
• frequencies undistorted
by head tissues
EEG generator
Adapted from Ivanitsky, Nikolaev, Ivanitsky 1999
Dynamics of EEG phase
• Phase contains essential
information about temporal
structure of signals
• Advantages:
– relative phase captures
spatiotemporal ordering of
cortical areas
– Meta-stable dynamics can be
observed through abrupt
changes in relative phase
Varela et al. 2001 Nature Rev
Spontaneous Phase Waves
Ito et al. 2007
Traveling and Standing waves
Ito et al. 2007
Instability index of relative phase
First Role of Whole Brain Activity
Brain Dynamics
• Waves of Phase constitute local and global
• Spontaneous transitions between global
• Variability of these modes in regularity,
velocity, duration, etc.
• These modes are recruited for information
Evoked Phase Waves
Alexander et al. 2013.
Information Processing Maxim
• States with low synchrony reflect local
information processing
• States with high synchrony reflect global
• Alternation of states is functionally
Phase synchrony is information-Specific
electrode chain
Phase synchroni za tio n index
Range of
SD thresholds
Nikolaev, Gong & van Leeuwen, Clin Neurophysiol 2005
unambiguous (left) and ambiguous (right) dot lattices
Patterned Activity
• Spontaneous patterns of synchronization and
• Desynchronization following stimulation
• (Regional) resynchronization reflects the
information communicated
Wave duration reflect stimulus information
Duration, ms
Aspect Ratio
information content increase
Nikolaev, et al., Cereb Cortex 2010
Aspect ratio AR = |b | / |a |
Second Role of Whole Brain Activity
Modular Small Worlds
• Optimal local and global connectivity
• The brain is a modular small world
• Structure emerges following spontaneous
large-scale wave activity (GDP in prenatal rats)
and re-emerges in functional architecture
following non-REM sleep
• Brain diseases linked to disturbance of
modular small-world functional architecture:
Schizophrenia, Alzheimer, Autism(?)
Symbiosis of Structure and Function:
a theoretical model
• Wave sequences help shape the architecture
• The architecture should sustain wave
Neural Mass Model
Breakspear et al. 2003
Return plot in three dimensions. Potential of pyramidal (V) and inhibitory (Z) neurons, average number of
open potassium ion channels (W)
Poincaré section of the Mass Model
Logistic Map
Coupled Logistic Maps
Note: the Network structure is a Small World (Watts & Strogatz, 1997)
Coupled Maps: From Random to
Small-world Organization
Gong & van Leeuwen, 2003; 2004; Kwok et al, 2007; Rubinov et al, 2009;
van den Berg & van Leeuwen, 2004; van den Berg et al., 2012
Adaptive Rewiring
Network Evolution
Initial (row 1), evolving (row 2) and
asymptotic (row 3) network
configurations for structural (column 1),
fast (column 2) and slow time scale
functional (column 3) networks. Fast time
scale networks represent the
instantaneous patterns of dynamical
synchrony. Slow time scale networks
based on the correlation coefficient of
100 consecutive functional states. Nodes
in all networks were reordered to
maximize the appearance of modules,
Rubinov, et al. (2009).
Macaque Cortex
(Young, 1993; Sporns & Zwi, 2004)
Path Length Cluster Index
2.0310 (0.0051)*
3.8262 (0.0099)*
0.6593 (0.0002)*
2.1159 (0.0133)*
0.2409 (0.0047)*
2.8901 (0.1173)*
0.8992 (0.0211)*
Dynamic activity shows global modes
In V1
In the whole cortex
Functional coordination in the service of local
processing and global communication
• Dynamic activity helps create optimal
architecture to sustain this type of activity
Thanks to:
Previous PDL:
Sergei Gepshtein (SALK); Pulin Gong (U Sydney); Junji Ito (FZ
Juelich); Hironori Nakatani (Tokyo U),Gijs Plomp (Geneva); Ivan
Tyukin (Leicester), and technical staff.
Current PDL:
David Alexander, Chie Nakatani, Tomer Fekete, Andrey
Nikolaev, Erik Steur, Chris Trengove, graduate students and
Daan van den Berg, Michael Breakspear, Michael Kubovy,
Thomas Lachmann, Michael Rubinov, Johan Wagemans, and
many others.
Thank You!
(senior) postdocs: David Alexander, Tomer Fekete, Chie
Nakatani, Andrey Nikolaev, Erik Steur, Chris Trengove; Ph.D.
students: Mojtaba Chehelcheraghi, Nicholas Jarman, Radha
Nila Meghanathan, Alessandro Solfo, Steffen Theobald,
Aleksandra Zharikova (and alumni)
Sergei Gepshtein (SALK), Thomas Lachmann (Kaiserslautern),
Antonino Raffone (Rome), Narayanan Srinivasan (Allahabad),
Ivan Tyukin (Leicester), Johan Wagemans (and their teams)

similar documents