The random neural network:
theory and applications
This talk will introduce the random neural network model (RNN) and some
of its extensions, for instance to synchronised interactions. We will
also discuss applications both in biology and engineering. These
include cortico-thalamic oscillations, protein alignment, video
compression, image texture generation and combinatorial optimization.
Sashi Marella, NJIT
Neuromodulation of cortical interneurons in epileptic animal models
In this talk, I will present some recent experimental findings on
putative modulation of cortical interneurons activity in animal
(vertebrate) models of seizure and epilepsy. Changes in receptor
density for neuropeptide Y (NPY) in these animal models implicated in
experimentally observed modulation of excitatory and inhibitory inputs
onto cortical interneurons will be discussed. Theoretical
questions regarding the functional significance of these
observations to the putative cortical circuit consisting of
pyramidal neurons and fast-spiking interneurons will be motivated.
Allen
Tannenbaum, University of
Alabama-Birmingham
Mathematical Methods in Medical Image Computing
In this talk, we will
describe some
theory and practice of controlled active vision. The
applications
range from visual tracking (e.g., laser tracking in turbulence, flying
in formation of UAVs, etc.), nanoparticle flow control, and
sedation control in the intensive care unit. Our emphasis will be on
the medical side,especially image guided therapy
and surgery. This includes projectssuch as radiation planning
in
cancer therapy, traumatic brain injury, and left atrial fibrillation.
We concentrate on two key areas: segmentation and registration.
For segmentation, we will describe several models of active
contours for which both local and global information may be included.
We will indicate how statistical estimation and prediction
ideas
(e.g., particle filtering) may be naturally combined with this
methodology. For registration, we propose using ideas from
optimal mass transport (Monge-Kantorovich) and uniformization theory
from Riemann surfaces. The techniques we invoke come from curvature
driven flows,
differential geometry, control theory, and the calculus of variations;
they will be demonstrated on a wide variety of data sets from various
medical imaging modalities.
Markus
A. Dahlem, Humboldt University Berlin
Migraine: A Dynamical
Disease
The key to the genesis of
migraine is an
emergent phenomenon called cortical spreading depression (SD).
SD
is a transient state that during its course massively perturbs ionic
gradients across cells membranes. A generic reaction-diffusion
mechanism with global inhibition is presented by which localized,
long-lasting but transient SD wave patterns are formed. The distinct
transient waves are caused by a ghost of a saddle-node bifurcation,
i.e., a bottle-neck passage in phase space associated with a
characteristic form (shape, size) that depends critically on the
curvature of the cortex. Similar patterns have been observed with fMRI.
We correlate the patterns with biological pathways for the pain
formation in migraine and investigate means by which neuromodulation
techniques
may affect these pathways.
Katherine
Newhall, New York University
Synchronous Firing Events in Stochastic Neuronal Network Models
Synchrony manifests itself in a variety of forms in noisy biological
systems. Within computational neuronal models, even with the
desynchronizing effect of noisy input, the excitatory coupling between
neurons can cause the network to synchronize, or oscillate, over a
large range of model parameters. I will discuss synchrony in
the
context of pulse-coupled Integrate-and-Fire models where cascading
firing events cause multiple neurons to fire at the exact same instance
in time. I will describe the synchronizing mechanism and
present
methods for computing the probability distribution for the number of
neurons firing together, as well as the probability the system
maintains a synchronous firing state. These results can
potentially be combined with population based simulation methods to
efficiently simulate interacting excitatory and inhibitory neuron
populations in biologically relevant regimes.
Analyzing conductance correlations involved in recovery of bursting
after neuromodulator deprivation in lobster stomatogastric neuron models