Colloquium Series
Colloquia are generally on Tuesdays at 3:30 p.m., once per month. They are usually held in the new Centre for Theoretical Neuroscience (CTN) seminar room in the Psychology, Anthropology, Sociology building (PAS), room 2464; exceptions will be noted). Abstracts are posted as available.
Here is a link for past events, colloquia, and speakers.
- Sept. 14, 2010 - Sue Becker (McMaster), 3:30 p.m., PAS 2464
- Sept. 24, 2010 - Mark Reimers (VCU), 2:30 p.m., PAS 2464
- Oct. 19, 2010 - Ernst Neibur (Hopkins), 3:30 p.m., PAS 2464
- Nov. 30, 2010 - Eric Shea-Brown (Washington), 3:30 p.m., PAS 2464
- Dec. 10, 2010 - Nathan Insel (Arizona), 10:30 a.m., PAS 2464
- Jan. 25, 2011 - Thilo Womelsdorf (UWO), 3:30 p.m., PAS 2464
- March 22, 2011 - Rob Kass (CMU), 3:30 p.m., PAS 2464
- April 6, 2011 - Waterloo Brain Day, PAS 2083
- May 17, 2011 - Mark Laubach (Yale)
- June 7, 2011 - Bruno Averbeck (NIH/NIMH)
- July 26, 2011 - Tiago Maia (Columbia)
Date:
Tues.,
Sept.
14,
2010
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Sue
Becker,
McMaster
University
Title:
A
Role
for
Hippocampal
Neurogenesis
in
Contextual
Memory
Formation
Abstract: The hippocampus plays a pivotal role in episodic memory formation, and in setting the context for ongoing behaviour. Its unique characteristics, including high plasticity, sparse coding, and the continuous production of new neurons throughout the lifespan, make it suited to both rapid encoding and long-term retention. I will describe a computational model of the hippocampus that postulates a role for neurogenesis in generating distinctive contexts for event memories. In the model, an encoding cost function is used to derive learning equations for optimizing the model parameters. Neurogenesis takes place in the dentate gyrus (DG). Mature DG neurons fire sparsely and contribute to the separation of daily events, whereas young DG neurons fire readily across time, providing common contextual features and contributing to the integration of all events comprising the learning experience into a single episodic memory. The model predicts that manipulations that suppress neurogenesis, such as stress and irradiation, will cause deficits on tasks that have a high potential for generating interference between similar memory episodes. Conversely manipulations that enhance neurogenesis, such as exercise, should improve memory on such tasks. This prediction has been confirmed in our recent empirical studies in both rodents and humans. Rats with irradiation-induced suppression of neurogenesis exhibit increased interference on two different tasks (Smith, Wojtowicz, Becker et al, in preparation; Winocur, Becker, Luu, Rosenzweig and Wojtowicz, submitted). In humans who have undergone six weeks of intense aerobic training, MRI blood volume increases in the DG, a known correlate of neurogenesis, are highly predictive of both increased fitness levels and improved memory on a visual pattern separation task (Dery, Toulouse, Wojtowicz, MacQueen, Noseworthy and Becker, in preparation).
Date:
Fri.,
Sept.
24,
2010
Location:
PAS
2464
Time:
2:30
p.m.
Speaker:
Mark
Reimers,
Virginia
Commonwealth
University
(VCU)
Title:
Neuroimaging
and
the
Architecture
of
Action
Abstract: This talk will develop the thesis that current imaging technologies and emerging strategies for analysis of imaging data will make quantitative data from neuroimaging much more germane to cognitive theories than they seem to be at present. I will illustrate this thesis by arguing that both distributed and localist models capture some part of the truth about brain activity, in that skilled behavior draws on sparsely distributed activity over several regions. This will be illustrated from the literature on functional data obtained by functional magnetic resonance imaging (fMRI), by electroencephalogram/magnetoencephalogram (EEG/MEG), and from direct recordings. It turns out that many regions are useful (improve performance) for a variety of functionally distinct behaviors, while few regions are necessary (i.e. cannot be replaced). I will describe some new technologies and some current work in data analysis strategies that will make neuroimaging data much more accurate, and hence even more relevant, in the near future. Such considerations lead naturally to a discussion of what kinds of modeling strategies can be informed by these kinds of data. I will argue that much of cognitive modeling is implicitly Platonist about cognitive processes, but that the neuroimaging data is more consistent with an ad-hoc developmental and evolutionary view of cognition; such a view in turn suggests a somewhat different set of questions.
Date:
Tues.,
Oct.
19,
2010
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Ernst
Neibur,
John
Hopkins
University
Title:
Up
States
are
Critical,
Down
States
are
Subcritical
Abstract: During sleep, under anesthesia and in vitro, cortical neurons in sensory, motor, association and executive areas fluctuate between Up and Down states (UDS) characterized by distinct membrane potentials and spike rates. While Down states are quiescent, Up state activity resembles that of rapid eye movement (REM) sleep and wakefulness, suggesting similar network processing. Another network phenomenon observed in preparations similar to those that exhibit UDS, such as anesthetized rats, brain slices and cultures devoid of sensory input, as well as awake monkey cortex, is self-organized criticality (SOC). This is characterized by activity "avalanches" whose size distributions obey a power law with critical exponent of about -3/2 and branching parameter near unity. We report an intimate connection between the two phenomena, SOC and UDS. We show analytically that networks of leaky integrate-and-fire neurons with short-term synaptic depression typically have 2 stable activity levels corresponding to UP and Down states, and that Up states are critical and Down states are subcritical. Our analytical results are confirmed by simulations of networks with different connectivity structures which also switch spontaneously between Up and Down states.
Date:
Tues.,
Nov.
30,
2010
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Eric
Shea-Brown,
Washington
University
Title:
Cooperative
Dynamics
in
Simple
Neural
Circuits
Abstract: The brain's networks contain numerous mechanisms that could lead to synchronized -- or correlated -- dynamics among cells. Moreover, information-theoretic calculations tell us that precisely what pattern of correlations emerges can have major impact on the fidelity with which a network encodes incoming signals. We'll explore simplified circuit models in which the link between single-cell dynamics, coupling architecture, and correlation can be understood, and discuss the consequences for neural coding.
Date:
Fri.,
Dec.
10,
2010
Location:
PAS
2464
Time:
10:30
a.m.
Speaker:
Nathan
Insel,
University
of
Arizona
Title:
Network
Dynamics
and
Neural
Coding
in
the
Rat
Medial
Prefrontal
Cortex:
Implications
for
Speed
of
Processing
and
Goal-directed
Action
Selection
Abstract: How does the brain use goals to select the right behaviors? Convergent evidence suggests that the dorsal medial prefrontal cortex (dmPFC) makes an important contribution to goal-directed action selection. The dmPFC is also part of a network of brain regions that becomes compromised in old age. It was hypothesized that during decision-making, some process of comparison takes place in the dmPFC between the representation of available actions and associated values, and that this process is changed with aging. These hypotheses were tested in aged and young adult rats performing a novel 3-choice, 2-cue decision task. Neuron and local field potential activity revealed that the dmPFC experienced different states during decision and outcome phases of the task, with increased local inhibition and oscillatory (gamma and theta) activity during cue presentation, and increased excitatory neuron activity (among regular firing neurons) at goal zones. Although excitatory and inhibitory activity appeared anti-correlated over phases of the decision task, cross-correlations and the prominent gamma oscillation revealed that excitation and inhibition were highly correlated on the millisecond scale. This “micro-scale” coupling between excitation and inhibition was altered in aged rats, and the observed changes were correlated with changes in decision and movement speeds of the aged animals. With respect to decision-making, both aged and young adult rats learned over multiple days to follow the rewarded cue in the 3-choice, 2-cue task. Support for the hypothesis that the dmPFC simultaneously represents alternative actions was not found; however, neuron activity selective for particular goal zones was observed. Interestingly, goal-selective neural activity during the decision period was more likely to take place on error trials, particularly on high-performing sessions and when rats exhibited a preference for a particular feeder. A possible interpretation of these patterns is that goal representations in the dmPFC might have sometimes overruled learned habits, which are likely to be involved in following the correct cue and which are known to be supported by other brain regions. Thus, in investigating the question “How does the brain select the right behaviors,” in the context of an experimentally controlled, slowly-learned, cue-following task, dmPFC activity was found that might sometimes contribute to the brain selecting the "wrong" behaviors.
Date:
Tues.,
Jan.
25,
2011
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Thilo
Womelsdorf,
University
of
Western
Ontario
(UWO)
Title:
Topography
and
Mechanisms
of
Valuation
and
Attentional
Control
Processes
in
Macaque
Fronto-cingulate
Cortex
Abstract:
Attentional
control
describes
network
processes
which
select
sensory
information
most
relevant
in
a
given
task
context.
Controlling
which
stimulus
is
attended
should
thus
originate
from
neurons
encoding
(i)
the
relevance,
i.e.
the
expected
value,
of
a
stimulus,
as
well
as
(ii)
contextual
rules
on
how
to
attentionally
select
target
stimuli.
Both
processes
of
attentional
control,
selective
value
predictions
and
instantiation
of
attentional
rules,
likely
arise
from
separable
nodes
in
fronto-cingulate
cortex:
The
anterior
cingulate
cortex
(ACC)
is
implicated
to
convey
value
predictions
(e.g.
"stimulus
A
promises
reward
outcome
X"),
while
anterior-lateral
prefrontal
cortex
(alPFC)
likely
initiates
conditional
rules
including
cue-target
association
rules
(e.g.
"if
cue
is
red,
then
target
is
stimulus
A").
At
the
neuronal
level,
both
control
processes
need
coordination
ensuring
that
only
those
neurons
become
activated,
which
biases
the
larger
attention-network
to
prioritize
target
processing.
One
neuronal
mechanism
linking
selective
target
encoding
neurons
to
a
functional
assembly
is
neuronal
synchronization.
We
therefore
measured
in
two
macaques
and
across
the
medial-to-lateral
extent
of
the
fronto-cingulate
whether
phase
synchronization
of
neuronal
spiking
activity
to
the
local
field
potential
(LFP)
conveyed
selective
information
about
the
expected
value
of
attentional
targets
and
about
the
cue-to-target
attentional
rule.
We
find
that
for
high-value
targets
ACC
neurons
selectively
synchronized
to
a
theta
rhythm,
suggestive
of
a
limbic
origin.
The
same
neuronal
cluster
engaged
in
beta-rhythmic
synchronization
when
attention
shifted
to
a
low-value
target,
i.e.
when
stronger
control
of
attentional
rule
information
needed
to
dominate
against
motivational
biases.
Consistent
with
this
interpretation,
beta
synchronization
was
particularly
strong
in
alPFC
during
attention
shifts
to
a
low-value
target.
These
results
reveal
a
switch-like
behavior
of
neuronal
synchronization,
which
was
frequency-specific
and
anatomically-confined
and
could
reflect
the
dynamic
instantiation
of
a
target
encoding
neuronal
assembly.
Date:
Tues.,
Mar.
22,
2011
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Rob
Kass,
Carnegie
Mellon
University
(CMU)
Title:
Statistical
Models
of
Neural
Activity
Abstract: One of the most important techniques in learning about the functioning of the brain has involved examining neuronal activity in laboratory animals under differing experimental conditions. Neural information is represented and communicated through series of action potentials, or spike trains, and the central scientific issue in many studies concerns the physiological significance that should be attached to a particular neuron firing pattern in a particular part of the brain. Because repeated presentations of stimuli often produce quite variable neural responses, statistical models have played an important role in advancing neuroscientific knowledge. In my talk I will briefly outline some of the progress made, by many people, over the past 10 years, highlighting work my colleagues and I have contributed. I will also comment on the perspective provided by statistical thinking.
Date:
Tues.,
May
17,
2011
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Mark
Laubach,
Yale
University
Title:
Lost
in
Transition:
Aging-related
Changes
in
Executive
Control
Signals
in
the
Rodent
Medial
Prefrontal
Cortex
Abstract: Aging reduces the ability of the prefrontal cortex (PFC) to exert control over action. We examined neural correlates of executive control in the PFC of younger (6 mo) and older (24 mo) rats during delayed response tasks. We also assessed aging-related differences in proteins related to cAMP signaling that might relate to behavioral changes (e.g., alpha-2A adrenergic receptors, PDE4A, and HCN channels). Older animals performed the tasks at a slow pace and wasted time licking dry spouts and pressing levers despite the presence of stimuli indicating reward elsewhere. These aging-related perseverations were associated with reductions in: 1) PDE4A, 2) neural firing rates, 3) neural sensitivity to behavioral pace, and 4) sensitivity to stimuli indicating reward availability. Our results suggest that aging impairs the ability of prefrontal networks to keep subjects "on task" and that cognitive enhancers that inhibit cAMP signaling may restore this aspect of executive control in the elderly.
Date:
Tues.,
June
7,
2011
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Bruno
Averbeck,
National
Institutes
of
Health
(NIH/NIMH)
Title:
Action
Selection
and
Action
Value
in
Frontal-striatal
Circuits
Abstract: The basal ganglia (BG) make up a substantial fraction of neural tissue in the brain and they have been implicated in most psychiatric and many neurological conditions. There is still, however, little consensus about what they contribute to normal behavior. Two prominent and somewhat overlapping theories of BG function are reinforcement learning and action selection. The reinforcement learning hypothesis suggests that the striatum represents the values of various actions, and that the dopamine input to the striatum is important for updating those value representations. The action selection hypothesis suggests that cortex generates sets of possible actions and the BG subsequently select an action from the set. This selected action is then returned to cortex via a BG-thalamo-cortical loop before the action is forwarded to the output structures. Reinforcement learning is often considered the mechanism which shapes the circuitry that selects the actions. To examine aspects of these hypotheses in the prefrontal cortical-striatal part of this loop, we carried out an experiment in which we asked monkeys to select actions under different behavioral conditions. In both conditions monkeys had to execute a sequence of three binary decisions. In the first condition monkeys had to carry out perceptual inference at each point of the sequence to determine the correct decision. The correct sequence of decisions changed every trial. The second condition was equivalent to the first except the correct spatial sequence of decisions remained fixed until the monkeys executed the sequence without errors 8 times. After that the correct sequence switched to a new sequence. In the second condition, therefore, the animals could use information about which sequence had been correct in the previous trial to determine the correct action at each point in the sequence, even if the perceptual decision was made very difficult. We found that animals were highly effective at using this previous information to improve their accuracy. We carried out simultaneous neurophysiological recordings in the lateral-prefrontal cortex (caudal area 46) and the caudate nucleus while the animals carried out the task. Across both conditions we found that the representation of the movement was much stronger in prefrontal cortex than it was in the striatum, and that the specific movement to be executed was not represented first in the striatum. We also found that the representation of the value of the actions was generally stronger in the striatum. Thus, the striatum appeared to have an enriched representation of the action value, but we did not find evidence that it was involved in selecting actions based upon these representations.
Date:
Tues.,
July
26,
2011
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Tiago
Maia,
Columbia
University