Colloquium Series 2015-2016
Here is a list of our upcoming speakers for the 2015 and 2016 academic year:
August 26, 2015 - Steve Furber (University of Manchester)
September 22, 2015 - Mriganka Sur (MIT: Massachusetts Institute of Technology)
October 6, 2015 - Stefan Köhler (Western University)
November 20, 2015 - Paul Cisek (University of Montréal)
December 8, 2015 - Quincy Almeida (Wilfrid Laurier University)
February 16, 2016 - Michael Barnett-Cowan (University of Waterloo)
March 21, 2016 - Zoran Tiganj (Boston University)
March 29, 2016 - Ning Jiang (University of Waterloo)
April 6, 2016 - Waterloo Brain Day (10th annual)
Date:
Wed.,
Aug.
26,
2015
Location:
PAS
2464
Time:
11:00
a.m.
Speaker:
Steve
Furber
(University
of
Manchester)
Title:
The
SpiNNaker
Project
Just two years after the world's first stored program ran its first program at Manchester in 1948, Alan Turing published his seminal paper on "Computing Machinery and Intelligence". The paper opens with the words: I propose to consider the question, "Can machines think?". Turing then goes on to explore this question thought what he calls "The Imitation Game", but which subsequent generations simply call "The Turing Test".
Despite spectacular progress in the performance and efficiency of machines since Turing's time, we have yet to see any convincing demonstration of a machine that can pass his test. This would have surprised Turing - he believed that all that would be required was more memory. Although cognitive systems are beginning to display impressive environmental awareness, they do not come close to the sort of "thinking" that Turing had in mind.
My take on the problems with true artificial intelligence are that we still really haven't worked out what natural intelligence is. Until we do, all discussion of machine intelligence and "the singularity" are specious. Based on this view, we need to return to the source of natural intelligence, the human brain. The SpiNNaker project has been 15 years in conception and 8 years in construction, but is now ready to contribute to the growing global community (exemplified by the EU Human Brain Project) that is aiming to deploy the vast computing resources now available to us to accelerate our understanding of the brain, with the ultimate goal of understanding the information processing principles at work in natural intelligence.
Date:
Tues.,
Sept.
22,
2015
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Mriganka
Sur
(Massachusetts
Institute
of
Technology)
Title: The
Functional
Logic
of
Cortical
Circuits
Cortical circuits transform simpler inputs into complex outputs that underlie behavior and cognition. Neuronal circuits of the cortex involve feedforward and top-down inputs, connections between specific types of excitatory and inhibitory neurons, and neuromodulatory inputs to specific cell classes. We have utilized novel experimental approaches – including two-photon imaging of neurons across multiple cortical regions in behaving mice, targeted recordings from specific neuron types, and activation and suppression of neuron classes - to analyze circuits that mediate visual perception and behavior. I will describe some recent results demonstrating powerful circuit motifs underlying spatial and temporal processing in the cortex. These include inhibitory computations performed via two fundamental operations, division and subtraction, by parvalbumin (PV) and somatostatin (SOM) expressing neurons; modulation of temporal processing by acetylcholine via a unique inhibitory-disinhibitory microcircuit; and information flow and attractor dynamics in neuronal populations of the frontal cortex during memory-guided behavior.
Date:
Tues.,
Oct.
6,
2015
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Stefan
Köhler
(Western
University)
Title:
Mechanisms
of
Familiarity
Assessment
in
Humans
The ability to consciously recognize the prior occurrence of objects, people, or other aspects of the environment is critical to many aspects of adaptive behavior. Recognition memory can succeed in the absence of successful recollection of episodic detail about a pertinent past stimulus encounter. For example, we have all been in situations in which a person we interact with feels familiar, but we cannot conjure up any information as to where or when we met her before. The process that allows for recognition under such circumstances is often referred to as familiarity assessment. In the present talk, I plan to review recent neuropsychological, neuroimaging, and psychophysiological research that aimed to shed light on the neural mechanisms that support familiarity and that shape its phenomenology. The work that I plan to discuss will speak to the following questions: How are memory representations that are tied to impressions of familiarity organized in the human brain? Can familiarity assessment be selectively affected by brain lesions? Does familiarity have an affective component?
Date:
Fri.,
Nov.
20,
2015
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Paul
Cisek
(University
of
Montréal)
Title:
The
Neural
Dynamics
of
Dynamic
Decisions
During
natural
behavior,
animals
must
continuously
make
decisions
in
a
rapidly
changing
environment.
Recent
studies
suggest
that
in
such
conditions,
the
brain
simultaneously
represents
multiple
potential
actions
that
compete
against
each
other
within
the
same
sensorimotor
control
circuits
involved
in
execution.
Here,
I
present
analyses
of
neural
spiking
activity
recorded
from
the
cerebral
cortex
of
monkeys,
while
they
decided
between
two
reaching
movements
based
on
a
changing
stimulus
indicating
which
is
more
likely
to
be
rewarded.
We
represent
the
state
of
the
system
as
an
evolving
trajectory
in
a
very
high-dimensional
space
where
each
axis
corresponds
to
the
activity
of
one
neuron,
and
use
dimensionality
reduction
to
project
this
to
a
9-dimensional
space
capturing
most
of
the
variance
in
the
data.
We
find
that
during
the
process
of
deliberation,
the
neural
state
evolves
upon
a
roughly
two-dimensional
“decision
manifold”
defined
by
orthogonal
components
related
to
sensory
evidence
and
the
growing
urge
to
respond.
The
moment
of
commitment
occurs
when
the
neural
state
falls
off
the
edge
of
this
manifold
into
one
of
two
orthogonal
attractors
that
lead
to
the
initiation
of
the
movements.
We
find
qualitatively
different
decision
manifolds
in
different
brain
regions.
For
example,
the
manifold
in
premotor
cortex
is
significantly
curved
while
in
primary
motor
cortex
it
is
remarkably
linear.
We
conjecture
that
the
premotor
cortex
implements
a
non-linear
recurrent
attractor
system
in
which
the
decision
is
made,
and
this
is
read-out
by
the
primary
motor
cortex
to
initiate
the
chosen
action.
Date:
Tues.,
Dec.
8,
2015
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Quincy
Almeida
(Wilfrid
Laurier
University)
Title:
Parkinson’s
Disease:
Sensorimotor
Transformations
through
a
Degenerating
Circuit
While it is well known that a variety of neural networks are essential for the planning and control of human movement, the interaction between sensory, perceptual and cognitive networks is not easy to disentangle. Vision is critical to all of these processes, and required to accurately detect and make judgments about objects and obstacles that we interact with. Vision is also needed to interpret the progression of successful (or unsuccessful) planned movements relative to environmental obstacles or threats. In addition, higher level cognitive processes (attention, executive function) are required for recognition and semantic processing of environmental stimuli.
Interestingly, while the neural networks for all of these systems are known to anatomically loop through the circuitry of the basal ganglia (BG), little is known about how basal ganglia dysfunction, as in Parkinson’s disease, might mechanistically influence movement control. This talk will consider the controversies related to how movement might be compounded by interactions between these processes.
One example that is highly related to vision is freezing of gait (FOG). It is considered one of the most debilitating symptoms associated with Parkinson’s disease (PD), yet there is considerable debate about the underlying mechanisms of FOG. Research has demonstrated that vision can have a tremendous impact on FOG episodes. In some cases, FOG is improved by enriching the visual environment with step cues, yet in other circumstances the richness of the visual (and sensory) environment is the very trigger of FOG (e.g. narrow doorways, cluttered spaces, visual or auditory dual tasks). Thus, there is a debate as to how visual environments actually influence movement deficits (such as FOG), and many different mechanisms have been suggested including: cognitive-attentional, sensory-perceptual, and anxiety-provoking processes. Yet the inter-relationship between these processes needs to be more carefully and thoroughly investigated.
In order to understand how visual stimuli may be associated with these different theories, a series of experiments will be presented that examines gaze behaviors during gait, in FOG-assisting and/or FOG-triggering environments. Gaze behaviors will then allow us to examine how gait deficits are associated. Results will be discussed in terms of understanding the underlying mechanisms that contribute to FOG.
Date:
Tues.,
Feb.
16,
2016
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Michael
Barnett-Cowan
(University
of
Waterloo)
Title:
Gravity's
Fundamental
Role
for
Perception
and
Action
Orientation is fundamental to perception and action. Knowing where you are in relation to the world and its contents is critical for adaptive behaviour. Orientation requires a reference, which the force of gravity is ideally suited for because it is constant on earth. The senses provide different types of information about the direction of gravity however. Thus a critical step that the brain must achieve to accurately represent the world and its contents is the integration of orientation cues from multiple sensory organs that individually do not provide an accurate representation of the physical world. Under normal circumstances when the head is upright with respect to the direction of gravity’s force and with respect to the orientation of a polarized visual environment, perceived self-orientation within an earth-centric field of reference is reinforced by such sensory cues. Perceptual stability is compromised, however, when orientation information from the senses is discordant. In this talk I will present a number of behavioural studies that demonstrate how fundamental gravity is to perception and action at the individual and group level. I will then briefly introduce a new research project designed to address these individual and group differences using genomics and will conclude by opening a discussion on how developing a working computational model of how the brain integrates multisensory information about orientation could be used to better predict, monitor and treat disorientation as well as sensory processing deficits that come with ageing and disease.
Date:
Mon.,
Mar.
21,
2016
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Zoran
Tiganj
(Boston
University)
Title:
Memory
Across
Scales:
Integrating
Computational
Models
and
Electrophysiological
Data
It is well known that, all things being equal, the accuracy of mammalian memory is better for events that took place at more recent past than at more distant past. I will present a biologically plausible computational framework that can account for this gradual decay of memory over multiple seconds. The framework relies on sequentially activated time cells that constitute an internal timeline. Information about what happened when is dynamically updated and always available in the brain. Having an internal timeline makes various useful computations easily achievable. For instance, if an animal knows its momentary speed, it can compute spatial distance from landmarks and construct place cells. Also, from the internal timeline it is straightforward to construct prediction of the future that is based on the spatiotemporal structure of the input signals. I will discuss the utility of these computations in the context of brain-inspired machine learning and artificial intelligence. Finally, I will present single unit data from electrophysiological recordings in rodents that support the existence of a neural timeline.
Date:
Tues.,
Mar.
29,
2016
Location:
PAS
2464
Time:
3:30
p.m.
Speaker: Ning
Jiang
(University
of
Waterloo)
Title:
Non-invasive
Brain
Computer
Interface
(BCI)
for
Motor
Function
Rehabilitation:
Putting
Patients
on
the
Driver’s
Seat
Neurorehabilitation applications, such as those for motor function rehabilitation of stroke, Parkinson etc, aims to induce neuroplasticity by re-establish damaged sensory-motor control loop. In recently years, Brain Computer Interfaces (BCIs) has been investigated as a promising tool for these applications. This is because patients’ active volition can be incorporated into the rehabilitation process through BCI, as compared to conventional rehabilitation approach where patients are in a passive role. In this talk, I will discuss a series of studies that systematically addressed the following questions: 1) is it possible to detect the motor intention, such as dorsiflexion (lifting up toe), in real time when using non-invasive brain recordings, i.e. electroencephalogram (EEG); 2) Can a system, consisting of such non-invasive detections of motor intent and subsequently triggered peripheral stimulations, be used to induce cortical plasticity in healthy subjects; 3) what are the critical factors and parameters of such a system in inducing cortical plasticity; 4) how well does the system work in a cohort of chronic stroke patients who suffer from ‘drop-foot’ and no longer respond to regular rehabilitation therapy. Our preliminary results suggested that such a patient-centered rehabilitation approach will dramatically increase the efficiency of the rehabilitation program. Our ultimate goal is to develop a viable rehabilitation paradigm that is fully driven by the patients, and can be used outside conventional clinical environment.