NeuroEngineering Research Lab

2012 Seminars

Brain Networks and Epilepsy

Date: 11:00am Wednesday 12 December 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building #261

The Melbourne School of Engineering presents a special NeuroEngineering Laboratory Seminar, with a guest speaker from the University of Exeter, UK, Professor John Terry. Professor Terry is an excellent speaker and his work is of the highest quality. ProfessorTerry is a world leader in computational and theoretical neuroscience, and nonlinear dynamics. His work focuses on studying oscillatory systems in the brain, with strong clinical implications.

In this seminar Professor Terry explores two classes of models (biophysically inspired and phenomenologically derived) to study network mechanisms of seizure initiation and evolution. He first introduces a biophysically inspired model, based on the principles of mass action, and demonstrate that transitions in the model precisely correspond to those observed in the EEG recordings of patients with absences. He describes …

Techniques for the processing and analysis of Magnetic Resonance Imaging phase data

Date: 11:00am Wednesday 24 October 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building #261

The NeuroEngineering Laboratory Seminar Series continues with a PhD Completion Seminar by Amanda Ng.

Magnetic Resonance Imaging phase data contains information about the magnetic properties and chemical composition of tissue and demonstrates novel contrast compared to the magnitude data more commonly used in MRI studies. Current methods for processing phase data and their limitations will be discussed, and new methods for artifact removal and analysis of phase data will be presented. Particular focus will be placed on Quantitative Susceptibility Mapping, in which localised magnetic susceptibilities are estimated from the phase data. This technique holds great promise in clinical studies of neurodegenerative diseases in which iron accumulates in the brain.

Amanda is completing her PhD at the University of Melbourne in collaboration with the Florey …

Overview of Research Activities in the NeuroEngineering Laboratory

Date: 11:00am Tuesday 11 September 2012

Venue: Brown Theatre, Electrical and Electronic Engineering Building, Building number 193

NeuroEngineering is about using scientific methods to understand and model the nervous system, and to use this knowledge to engineer systems that interact with, augment, or mimic nervous system functionality. This seminar will provide an overview of the research activity of the NeuroEngineering Laboratory in the Department of Electrical and Electronic Engineering at the University of Melbourne. The areas of research are: Audition, Speech and Bionic Ear Design, which aims to develop improved technology for the hearing impaired. Bionic Eye Design and Vision, which aims to develop a bionic eye to aid the vision impaired. Computational Neuroscience, which aims to develop mathematical models and computational analyses of neural systems. Epilepsy, which aims to develop technology that can detect, predict and stop epileptic seizures. Neuroimaging, which aims to develop MRI acquisition …

High-frequency neuronal oscillations and the role of intrinsic neuronal properties

Date: 11:00am Wednesday 22 August 2012

Venue: Main Conference Room, Ground Floor, Centre for Neural Engineering Building, Building Number 261

Neurons in the brain engage in collective oscillations in a range of frequencies that span several orders of magnitude. In particular, gamma and high-gamma oscillations (40-100 Hz and above) have been associated with neuronal activation in several brain regions, and are altered in cognitive disorders such as schizophrenia. First proposed as a psychophysiological mechanism for perceptual binding, gamma oscillations are now acknowledged as a general and versatile mechanism of neuronal processing. Gamma oscillations have been shown to be critically dependent on the activity of inhibitory interneurons. In contrast to excitatory neurons, inhibitory interneurons are present in many different subtypes, which differ in their molecular, electrophysiological and anatomical properties. In particular, interneuronal types that are responsible for gamma generation often exhibit …

System Identification in Neural Circuitry. How big is the problem, how do you make it feasible?

Date: 11:00am Wednesday 15 August 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering, Building Number 261

Every aspect of modern science relies upon creating representations of things. And when we do, we pick the signals that interest us and the behavior that interests us. From that, we determine how to interpret the way input is converted into output in a system. Our description of that process is our understanding of the system. The same is true for mental processes and reverse engineering their implementation in neural circuitry.

The feasible approach to this is called (whole) brain emulation and relies on determining precisely which signals we care about and then breaking the problem down into a collection of smaller system identification problems. To tackle those, there is a roadmap that includes structural scanning (connectomics) as well as new tools for functional recording – some of which are now in development in collaboration with laboratories at …

Intermittent Control: a Computational Theory of Human Control Systems

Date: 11:00am Wednesday 1 August 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building, Building Number 261

Professor Peter Gawthrop, visiting researcher to National ICT Australia (NICTA) in the Department of Electrical & Electronic Engineering will be presenting a seminar as part of the NeuroEngineering Laboratory seminar series, titled "Intermittent Control: a Computational Theory of Human Control Systems". The term "Intermittent" has been used since the 1940s to describe human motion control as a sequence of open-loop trajectories punctuated by intermittent feedback. This seminar formulates a computational model to capture this behaviour whilst also explaining why human motion often appears to be continuous in nature. The use of the model to explain experimental data will be discussed.

Stem Cell Translation: Tools & Therapies

Date: 11:00am Wednesday 25 July 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building, Building No. 261

The NeuroEngineering Laboratory Seminar Series continues with a talk by Associate Professor Jeremy Crook on Stem Cell Translation: Tools & Therapies.

Neural Network Model of Auditory Perception

Date: 11:00am Wednesday 30 May 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building, Building number 261

In this seminar, PhD student, Nina Saeedi will report on her research project which creates a physiologically-inspired neural network model of auditory perception and evaluates its performance in perceiving different aspects of sound. In this network, acoustic and electric models of the auditory system are applied as models of normal hearing and hearing with cochlear implants.

The project assesses model performance in different tasks like pitch perception, vowel recognition and consonant recognition to ensure that it is similar to human performance reported in the literature. It uses this model as a test platform to investigate the effect of the limiting factors on the auditory performance of cochlear implants. These include insufficient temporal resolution, limited number of electrodes and current spread, as well as the inability of the brain …

Smartphone - The Centrepiece of mHealth Reality

Date: 11:00am Wednesday 16 May 2012

Venue: Main Conference Room (Ground Floor), Centre for Neural Engineering Building, Building number 261

The capabilities of Smartphone platforms will play major roles in driving the practicality of mHealth applications, which are expected to impact prevention and treatment methods in diverse medical fields. This talk opens the concept of Mobile Health (mHealth) from the biomedical applications, Quality of Service (QoS), security, and cloud computing perspectives. The number of operational mobile subscribers is expected to pass the 6 billion mark by mid 2012 and the number of wearable wireless sensors is expected to grow to 400 million by the year 2014. Such a tremendous growth in the mobile space will result in an increasing number of mobile-based applications deployments, such as: Machine-to-Machine (M2M) communications, Electronic-Health (eHealth), and Mobile-Health (mHealth). These emerging mobile applications will require 3G and 4G mobile networks for biomedical …

Algorithms for current steering in retinal prostheses

Presenter: Evgeni Sergeev, The NeuroEngineering Laboratory, Department of Electrical and Electronic Engineering, University of Melbourne.
Date: Wednesday 13 June

Abstract

The current emitted by each electrode of a retinal prosthesis can be independently controlled. The problem of current steering is to select appropriate currents in order to shape the firing pattern of the target neurons as required. Before this is possible, a link between the tissue current and the response of a retinal neuron must be established. Several in vitro results from the literature with be discussed, which will help us develop a relevant model. Then preliminary results will be presented.

Neural Network Model of Auditory Perception

Speaker: Nina Saeedi, The NeuroEngineering Laboratory, Department of Electrical and Electronic Engineering, University of Melbourne.
Date: Wednesday 30 May 2012

Abstract

There are different factors limiting the auditory performance of cochlear implant recipients. From a signal processing point of view, some of these factors are: insufficient temporal resolution, limited number of electrodes and current spread. Also, poor performance might be a result of inability of the brain to adapt to the information provided by the implant.

Investigating the effects of any of the above factors by observing patients' performance is very expensive and time-consuming. The goal of this project is to create a physiologically- inspired neural network model of auditory perception and evaluate its performance in perceiving different aspects of sound. In this network, acoustic and electric models of the auditory system are applied as models of normal hearing and hearing with cochlear implants. Model performance will be assessed in different tasks like pitch perception, vowel recognition and consonant recognition to ensure that it is similar to human performance reported in the literature. After the model is validated, it will be used as a test platform to investigate the effect of the limiting factors mentioned above.

Functional changes of the ganglion cells following photoreceptor degeneration in a mouse model of retinitis pigmentosa

Speaker: Susmita Saha
Date: Wednesday 9 May

Abstract

In order to use a retinal implant to restore vision in Retinitis Pigmentosa (RP), the ganglion cells need to be intact since they are the output cells of the retina. However, few studies have investigated how photoreceptor loss in retinitis pigmentosa affects the synaptic activity and output of the ganglion cells during different stages of degeneration. Also, very little is known about the synaptic activity of the ganglion cells in response to electrical stimulation in a RP-degenerated retina. The objective of this PhD project is to investigate the changes in the spontaneous excitatory and inhibitory synaptic activity of the alpha-ON and alpha-OFF ganglion cells in the degenerate retina at different stages of the disease. This involves investigating the anatomical distribution and electrophysiological properties of synapses of the ganglion cells in an animal model.

Adult neurogenesis and mental health: A summary of recent results

Speaker: Prof. Joachim Diederich
Date: Wednesday 2 May 2012

Abstract

Adult neurogenesis is one of the most active and vibrant areas in the neurosciences. While it was believed for decades that “no new neurons” are being formed in the adult mammalian brain, it has been confirmed now that the generation of new neurons and synapses does occur in both the hippocampus and the olfactory bulb. Even more importantly, new neurons are being recruited into functional circuits and the process appears to be highly relevant for learning and memory. Impaired neurogenesis has been associated with mental health problems such as depression and schizophrenia. This informal presentation will provide an overview of recent results on adult neurogenesis and the link to mental health problems.

The Neural Tissue Simulator: An Ultrascalable Solution to Large-Scale Structural and Electrophysiological Simulations of Neuronal Circuit Function

Speaker: Dr. James Kozloski, Research Staff Member, Master Inventor, Biometaphorical Computing Thomas J. Watson Research Center, Yorktown Heights, NY USA
Date: Wednesday 18 April 2012

Abstract

We have developed a novel tissue volume decomposition, and a hybrid branched cable equation solver for performing large-scale simulations of neural tissue (2011). The decomposition divides the simulation into regular tissue blocks and distributes them on a parallel multithreaded machine. The solver computes neurons that have been divided arbitrarily across blocks and can be considered a tunable hybrid of Hines’ fully implicit method (1984), and the explicit predictor-corrector method of Rempe and Chopp (2006). We demonstrate thread, strong, and weak scaling of our approach on a machine of 4,096 nodes with 4 threads per node. Scaling synapses to physiological numbers had little effect on performance, since our decomposition approach generates synapses that are almost always computed locally. The largest simulation included in our scaling results comprised 1 million neurons, 1 billion compartments, and 10 billion conductance based synapses and gap junctions. Based on results from the ultra-scalable Neural Tissue Simulator, we estimate requirements for a simulation at the scale of a human brain and derive an approach to scaling computational neuroscience together with expected supercomputational resources over the next decade.

Biomolecular Structure Determination without Crystallography: Bright Prospects for X-ray Free-Electron Laser Imaging

Speaker: Dr. Harry Quiney, ARC Centre of Excellence for Coherent X-ray Science, School of Physics, The University of Melbourne
Date: Wednesday 4 April 2012

Abstract

The structures of biomolecules are currently determined crystallographically using coherent X-ray or electron sources. This process naturally requires the availability of high quality crystals, the production of which frequently presents formidable difficulties. Membrane proteins, which mediate communication between the interior and exterior environments of cells, typically form two-dimensional structures and have proven resistant to the methods that are used to grow samples suitable for crystallography. Periodic structures amplify high-resolution information about the unit cell structures, so any structure determination scheme using aperiodic samples must trade off this amplification against an increased incident flux of photons or electrons. The X-ray Free-Electron Laser (XFEL) sources in the US, Japan and Germany are being developed to determine biomolecular structures by interacting intense femtosecond X-ray pulses with nanocrystals or single-molecule targets. The pulses are so bright that the electronic damage caused by the interaction inevitably leads to the Coulomb explosion of the target. If the pulses are short enough, however, sufficient structural information can be gathered before this disintegration, an approach that is now generally known as “diffract and destroy” imaging. This presentation will survey the potential uses of emerging XFEL source technologies in the determination of structure and dynamics in biomolecular systems.

Exploring the Human Default Mode Network with Intracranial Electrophysiology

Speaker: Dr. Brett Foster, Stanford Human Intracranial Cognitive Electrophysiology Program & Laboratory of Behavioral and Cognitive Neurology. Department of Neurology and Neurological Sciences, School of Medicine, Stanford University
Date: Tuesday 27 March 2012

Abstract

Since its discovery over a decade ago, the human “Default-Mode Network” (DMN) has received an incredible amount of research interest and skepticism in the field of neuroimaging. With many early criticisms being addressed, the scientific debate now focuses on adjudicating the specific functional role(s) of this network and its unique contribution to cognition. While neuroimaging studies have made progress in this respect, more refined methods are required for anatomically and temporally resolving the functional dynamics of DMN activity. Indeed the electrophysiological correlates of DMN function remain relatively unknown. Our group has tackled this specific question using human intracranial recordings from the posteromedial cortex (PMC), a core node of the DMN. This talk will summaries our progress so far in studying the cognitive electrophysiology of the PMC and the novel insights obtained by using intracranial recordings.

Modelling the impact of complex network structure on balanced excitation and inhibition in large networks of cortical neurons

Speaker: Mark McDonnell
Date: Wednesday 21 March 2012

Abstract

The developing interdisciplinary field of network science is proving highly fruitful in revealing complex structure in the connections between the components of many complex systems. Examples of “complex networks” include small-world networks and scale-free networks, and example systems where they have been observed include genomes, social groups, financial markets, and the brain. Network science also seeks to develop mathematical analysis of the influence of complex structure on system dynamics, by combining empirical connectivity data with theory from statistical and nonlinear physics.

Physiology and Modelling of complex cells in primate Visual cortex

Speaker: Shaun Cloherty
Date: Wednesday 7 March 2012

Abstract

One of the enduring dichotomies in visual neuroscience is the classification of visual cortical neurons into simple and complex types. These cell types are distinguished by the way they spatially and temporally integrate visual stimuli, i.e., their receptive field properties. Simple cells exhibit spatially distinct subregions within their receptive fields, which respond selectively to either luminance increments (bright stimuli) or decrements (dark stimuli). In contrast, complex cells exhibit bright and dark receptive fields that are largely overlapping. As a result, complex cells are often considered to be phase invariant energy detectors. I will present results from single unit recordings in primate visual cortex demonstrating that complex cell receptive fields are dynamic and, in fact, complex cells can encode phase information. In light of these data I will review and assess the prevailing models of cortical complex cells.

Fluorescence guided resection of glioma using 5-aminolevulinic acid-Quantitative Methods

Speaker: Neda Haj-Hosseini
Date: Wednesday 29 February 2012

Abstract

Total tumor resection in patients with glioblastoma multiforme (GBM) is difficult to achieve due to the tumor's infiltrative way of growing and morphological similarity to the surrounding functioning brain tissue. The diagnosis is usually subjectively performed using a surgical microscope. A hand-held optical touch pointer using a fluorescence spectroscopy system is developed and evaluated for distinguishing healthy from malignant brain tissue intraoperatively.

Inferring neuronal network connectivity from multiunit tetrode recordings in live animals

Speaker: Dr. Patrick O’Brien
Date: Wednesday 8 February 2012

Safety of Wide-Field Retinal Prosthesis

Speaker: Joel Villalobos
Date: Wednesday 25 January 2012

Abstract

Within the research effort to develop a clinically viable retinal implant, there is a pre-clinical program being conducted. This session will present a summary of the PhD research efforts, which involved design of a spherically contoured implant substrate and passive chronic suprachoroidal implantation in a cat model. Results of the histopathology assessment indicated the implant was well tolerated in the eye's suprachoroidal space. Electrode impedances and electrical stimulation thresholds showed that this technology is viable and could continue progression towards a clinical program.

Representation of input signals in recurrently connected neuronal networks

Speaker: Matthieu Gilson
Date: Wednesday 18 January 2012

Abstract

Spike-timing-dependent plasticity (STDP) has been observed in many brain areas, such as sensory cortices, where it is hypothesised to develop the structure of synaptic connections between neurons. Previous studies have demonstrated how STDP can capture spiking information at a short timescale, such as coincident spiking, spike patterns and oscillatory spike trains. However, it is not yet clear how STDP stores this information in detail in the synaptic structure of recurrently connected neuronal networks. We use a theoretical framework to investigate the learning dynamics induced by STDP based on the temporal cross-correlograms between input spike trains. The key to understand the computation scheme is the interplay between the properties of STDP, neuronal response and input correlograms. In the case of a single neuron, STDP can extract the strongest spectral components of the input correlation structure in a similar fashion to principal component analysis (PCA). In a recurrent network, STDP can encode different features of input stimuli relying on partial connectivity and heterogeneities in the synaptic connections (e.g., conduction delays). Our mathematical study of the STDP dynamics aims to bridge the gap between physiology and machine learning, thus shedding light on possible neuronal encoding schemes. On the application side, such spiking neuronal networks that behave as self-adapting filters can be used for the representation, categorisation and detection of spatio-temporal stimuli.

The frequency of oscillatory inputs is encoded in the connection strengths of networks of spiking neurons by STDP

Speaker: Rob Kerr
Date: Wednesday 7 December 2011

Abstract

Spike-timing-dependent plasticity (STDP) is a learning rule that updates synaptic strengths based on the relative timing of pre- and post-synaptic spikes. Unlike classical, rate-based Hebbian learning, STDP can potentially encode fast temporal correlations in neuronal activity, such as oscillations, in the functional structure of networks of neurons that have axonal and dendritic propagation delays. In this study, the changes made by STDP to synaptic strengths in recurrent networks with axonal delays receiving oscillatory inputs were investigated analytically with Poisson neuron models and verified through simulations with leaky integrate-and-fire neurons. The motivation behind this was to understand how general spatiotemporal patterns can be learnt by a network of neurons with STDP present. An application of this specific model might be in explaining how the brain can perceive the pitch of complex sounds up to 300Hz, even when the fundamental frequency is missing. Both the analytical description and simulations found that connections were selectively potentiated and depressed based on their axonal delay in such a way that the delays of the strong connections in the network “resonated” with the input frequency.


Contact Us

Professor David Grayden

Director, NeuroEngineering

T: +61 3 8344 5234
E: