NeuroEngineering Research Lab
Technology for reverse-engineering cortical circuits
Date: Monday 8 December
Speaker: Dr Simon Schultz, Department of Bioengineering, Imperial College London
Recent years have seen a step up in efforts to develop technology for reverse-engineering brain circuitry, with for instance the recent Obama BRAIN initiative in the USA. In this talk, I will describe efforts in my laboratory to develop, and scale up, technology for studying cortical circuits. We are currently focusing on (i) improving scanning technology for multiphoton microscopy in order to acquire high signal to noise ratio measurements of activity from large numbers of neurons simultaneously, (ii) the development of improved algorithms for action potential detection from calcium imaging using GCamp6 and OGB-1 AM indicators, (iii) the development of algorithms for decoding neural ensemble activity patterns that scale well to hundreds or thousands of neurons, and (iv) the development of a novel multi-photon targeted robotic in vivo patch-clamping platform. We are applying these tools to study information processing in the cerebellum and in mammalian neocortical circuits. A major driver of research into cortical circuits is the increasing prevalence and social cost of dementias such as Alzheimer’s Disease, which disrupt information processing in the cortical circuit. An important outcome of our research is thus the development of technologies which allow information processing deficiencies to be detected at earlier stages in Alzheimer Disease models, and for therapeutic strategies to be directly assessed in terms of their effect on cortical circuit function.
SimonSchultz is Royal Society Industry Fellow, Reader in Neurotechnology, and Director of the Centre for Neurotechnology at Imperial College London. He received Bachelor degrees in electrical engineering and physics at Monash University, and a Masters degree in electrical engineering at Sydney University, before completing a DPhil in computational neuroscience at Oxford University in 1998. This was followed by postdoctoral stints in experimental neuroscience with Tony Movshon at New York University, and Michael Häusser at UCL. He joined Imperial College in 2004, and has led the development of Imperial’s critical mass in the area of Neurotechnology. He is widely known for work on neural coding. He has been amongst the pioneers in the use of two-photon imaging to study neural coding and has also worked on large-scale computational models of cortical circuits. He has been the PI or Co-PI of grants totalling over £20M, including being on the executive board of the EU FP7 Marie-Curie Training Network “NETT - Neural Engineering Transformative Technologies”, and PI of the £10M EPSRC Centre for Doctoral Training in Neurotechnology for Life and Health. He acts as Associate Editor for the Journal of Computational Neuroscience.
Development of a high-dimensional brain machine interface
Date: Monday 24 November
Speaker: Dr Yan Wong, The University of Melbourne
Over the last few decades neural prosthetics such as the cochlear implants and deep brain stimulators have greatly improved the lives of tens of thousands of patients. In recent years, there has been a push to translate this early success into new therapeutic devices to help those suffering from a broader range of sensory and motor deficits. One such device is the brain machine interface targeted towards upper limb amputees. Current state-of-the-art brain machine interface devices suffer limitations due to an inability to extract enough information from neural signals as well as instability in the neural signals under examination. In this talk I will outline work towards the utilization of movement synergies as well as the incorporation of the local field potential into decoding algorithms. I will present a novel test system that allows recording from multiple depths of the frontal motor cortices simultaneously with all the movements of the arm and hand. Finally, I will present results showing the successful decoding of high-dimensional upper limb movements both offline and online.
Dr Yan Wong received his PhD for work towards the design and development of a vision prosthetic microchip and novel electrode organizations for current focusing from the University of New South Wales in 2009. For his postdoctoral work, he joined the Center for Neural Science at New York University studying the role of spike-LFP interactions in the Parietal cortex on movement planning, as well as developing a Brain Machine Interface for high-dimensional upper limb control. He has recently joined the Biomedical Engineering department at the University of Melbourne. During his career, Dr Wong has received grants from the NIH and DARPA, as well as many prestigious accolades such as twice being awarded the best student paper at the annual IEEE International EMBC meeting and the AusBiotech National Student Excellence award.
Improving our understanding of cardiovascular disease through innovations in coronary artery imaging and fluid dynamics research
Date: Monday 17 November
Speaker: Associate Professor of Medicine Peter Barlis, Melbourne Medical School, University of Melbourne
Cardiovascular disease remains the leading cause of morbidity and mortality in our community. Several advances in the diagnosis and treatment of heart disease have played an instrumental role in improving outcomes for patients. Nevertheless, a number of challenges still exist. Over the last 5 years, a novel intracoronary imaging technology called optical coherence tomography (OCT) has been increasingly used. With a super-high resolution of 15 microns, this light-based imaging modality permits detailed interrogation of within the coronary arteries to look at the features that go on to cause build up of plaque. This imaging modality is also best suited to visualize coronary stents, the devices used to prop open blocked arteries and used to identify predictors of why stents may fail once they have been implanted. This seminar will examine the utility of OCT in contemporary clinical practice and showcase some of the innovative applications and studies currently under way within the Faculty of Medicine, Dentistry & Health Sciences and the School of Engineering at the University of Melbourne.
Peter Barlis, MBBS, MPH, PhD, FESC, FRSA, FACC, FCSANZ, FRACP. Peter is an internationally recognised Interventional Cardiologist and Associate Professor of Medicine with the Melbourne Medical School, University of Melbourne and holds an Honorary Principal Fellow appointment with the Department of Mechanical Engineering, the University of Melbourne. After graduating from the University of Melbourne (MBBS), he completed his advanced cardiology training at Austin Health and then completed a Master of Public Health with the Department of Epidemiology and Preventative Medicine, Monash University. He then undertook his interventional fellowship at the National Heart & Lung Institute, Royal Brompton Hospital, UK supervised by Professor Carlo Di Mario and then went on to complete his PhD at the Thoraxcentre in the Netherlands on the use of optical coherence tomography (OCT) in interventional cardiology, supervised by Professor Patrick W. Serruys.
Since returning back to Australia in 2008, he has introduced this novel imaging modality into the country and has pioneered its use and uptake across the region. He is an investigator on numerous clinical trials and is a reviewer for multiple cardiovascular journals. Peter is frequently invited to give lecturers nationally and internationally on the application of light based imaging for the assessment of atherosclerotic plaque and coronary stents. He has published over 90 peer-reviewed manuscripts, many of which he is first or senior author and in 2011 was co-Editor of the ‘Textbook of interventional cardiology: Principles & Practice’ (Wiley-Blackwell Publishing). He is holder of a prestigious NHMRC Health Professional Fellowship allowing him to undertake in-depth analytical research using OCT to better understand the reasons behind coronary stent failures and in 2011, with the Department of Mechanical Engineering group, was awarded a $1M ARC Linkage Grant examining cutting-edge imaging and fluid mechanics techniques to help improve future coronary stent designs. In addition to being a Fellow of the Cardiac Society of Australia and New Zealand and the European Society of Cardiology, in February 2014, Peter was awarded Fellowship status of the American College of Cardiology in recognition of his clinical and research impact.
Information theoretically estimating the performance of cochlear implants
Date: Monday 10 November
Speaker: Xiao (Demi) Gao, The Institute for Telecommunications Research, University of South Australia
Information theory has been used in estimating the performance of cochlear implants. Previously, a channel model of cochlear implant stimulation was developed and the performance of cochlear implants was numerically estimated from an information theoretic perspective. Here, we propose an improvement to the biological accuracy of a crucial component of the overall modelling framework, a revised model of the channel output. We estimate the stochastic information transfer from cochlear implant electrodes to auditory nerve fibres, and infer the optimal number of electrodes by calculating the mutual information between channel input (choice of electrode) and channel output (defined as a function of the active nerve fibres in response to an electrode choice). We also investigate to what extent the positions and the usage probabilities of electrodes could impact on the performance of cochlear implants based on this modelling framework.
Xiao (Demi) Gao is currently a PhD student at the Institute for Telecommunications Research, University of South Australia under the supervision of Dr Mark D. McDonnell and Assoc Prof David B. Grayden. Her PhD research is focused on modelling the cochlear implant system and theoretically estimating the performance of cochlear implants. Demi received a Bachelor of Science in Computer Science in Central China Normal University in 2006, and a Master of Science in Biology in Huazhong Agricultural University in 2009. Since graduating she worked at the Institute of Hydrobiology, Chinese Academy of Science as a research assistant from 2007 to 2010. Then she worked as a research engineer in Chinese Maritime Academy from 2011 to 2012.
Biomechanics of knee joint injuries using cutting-edge musculoskeletal models and gaming technologies
Date: Monday 20 October
Speaker: Dr Hossein Mokhtarzadeh, the University of Melbourne and Australian Institute for Musculoskeletal Science (AIMSS)
The aim of this presentation is to address the biomechanics of anterior cruciate ligament injury (ACL) and its consequences. The ACL is one of the main spring-like ligaments of the knee joint that plays a major role in stabilizing the knee joint. ACL rupture is a comMonday sporting injury that may lead to an early knee osteoarthritis (OA). OA following ACL injury is a debilitating cartilage disease which has no cure yet. ACL reconstruction (ACLR) surgery is a costly method that is performed in order to help patients to return to sport and reduce the likelihood of OA. However, it has been shown that ACLR may not prevent OA. This talk will present some our recent work on the role of lower limb muscle function on the knee joint loading that is believed to contribute to initiation of OA. In addition, some novel methodologies using gaming technologies to enhance our understanding of the long term effect of ACL injury will be presented.
Dr Hossein Mokhtarzadeh has recently completed a PhD in Mechanical Engineering specialising in biomechanics at the University of Melbourne under the supervision of A/Prof. Peter Lee and Dr Denny Oetomo. Dr Mokhtarzadeh is currently doing his postdoctoral research fellow at the University of Melbourne and Australian Institute for Musculoskeletal Science (AIMSS) and leading musculoskeletal modelling group at AIMSS. He directly works with A/Prof. Peter Pivonka who is an internationally expert in computational bone biology. Hossein’s main area of research is in mechanobiology and particularly neuromusculoskeletal modelling , computational biomechanics, drug delivery in bone, and in-vitro and animal studies. Throughout his PhD, he received a number of awards including one of the top 12 early career researchers in Australia in 2013 as a Fresh Scientist, an Inspiring Scientist to present his work in Museum Victoria, and first place in 3 minute thesis competition in 2010. He received the Outstanding Researcher Award from the NIH center at Stanford University in 2014. Dr Mokhtarzadeh received a huge amount of media attention for his PhD related research, including articles in Herald Sun, The Age, ABC science show to name a few. He presented his work on biomechanics of knee injury at the World Congress of Biomechanics, 2014 as an invited speaker in the ANZSB Young Investigator session. He has recently come back from visiting major Universities in the US including Harvard and MIT Universities, Cornell University, Boston University, Northeastern University, and Stanford University.
Neuromorphic approaches to computing acoustic information
Date: Monday 13 October
Speaker: Associate Professor Neil McLachlan, Melbourne School of Psychological Sciences
During the 1890’s Ivan Pavlov observed that dogs could be conditioned to salivate at the sound of a bell. The association of conditioned stimuli to behaviours (or unconditioned stimuli) has been studied in a wide range of animals for over a century, however practically no research has been undertaken on how animals learn to recognize the conditioned stimuli in the first place. This is important because sound recognition likely occurs early in auditory processing, and underpins most other auditory functions. More specifically, previous research has shown that conditioned reflexive responses to sound involve ponto-cerebellar pathways, and so these pathways likely underpin sound recognition more generally. High level computational models of these pathways have been used to recognize human speech, music, environmental sounds and animal calls, and to act as adaptive filters for integrating pitch information. This paper will outline a new neurocognitive account of the auditory pathways and provide examples of computational algorithms based on this model. More broadly, it will discuss the possibility that neuro-cognition based on memory processes may provide the operating systems for future generations neuromorphic computers based on memsistors. These computers will learn and adapt to natural environments just like animals, but can “inherit” (or share) their memories from other computers at any time.
Dr McLachlan is an Associate Professor in Psychological Sciences at The University of Melbourne and has broad professional experience in music, acoustic design, engineering, and auditory neuroscience. In 2000 he designed the World’s first harmonic bells, and more recently has designed a new harmonic percussion ensemble for use in educational and a range of community contexts. To establish better design criteria for musical instrument design he has developed the first end-end neurobiological model of auditory processing. He has computationally implemented aspects of this model leading to the development of new sound segregation and recognition algorithms for hearing prosthetics and automated sensing systems.
Diamonds are Electronics’ Best Friend: An all-diamond packaging for a bionic vision prosthesis
Date: Monday 22 September
Speaker: Samantha Lichter, The University of Melbourne
Bionic vision through electrical stimulation of the retina is fast becoming a reality. To date, clinical trials have allowed blind patients to see a lover’s smile and navigate night scenes. This data has encouraged an abundance of research activity. Bionic Vision Australia, among others, is developing a retinal prosthesis to restore high visual acuity. One of its flagship technologies is a diamond electrode array, which will form part of the encapsulation for the implanted electronics. The remainder of the encapsulation also needs to be constructed from leak-proof, or hermetic, materials. The PhD’s aim was to design and test the feasibility of a hermetic encapsulation that incorporated the diamond electrode array. An all-diamond hermetic encapsulation design was proposed, in which a diamond box-shaped capsule was bonded to the diamond array, with the electronics contained inside.
Diamond capsules were made from polycrystalline diamond. Laser micromachining was found to be the optimal fabrication method. Hermetic joints were made in diamond using vacuum brazing with precious metal braze alloys. Bond interfaces were studied for morphology, chemical composition and hermeticity. Brazed diamond capsules were sealed at room temperature using laser microwelding. Welds were optimised for smooth surfaces and hermeticity.
The results demonstrated a hermetic all-diamond encapsulation. Combining the hermetic capsule, the brazing technique, and the welding technique with the diamond electrode array formed a retinal prosthesis technology that can protect against degradation for the lifetime of the patient.
Samantha Lichter is doing her PhD with the Bionic Vision Australia team at the University of Melbourne. Her thesis is entitled “An All-Diamond Hermetic Encapsulation for a High-Acuity Retinal Prosthesis”. Samantha completed a Bachelor of Science and Bachelor of Materials Engineering (Hons) from Monash University in 2007. She has designed, prototyped and proved in concept a novel leak-proof encapsulation technology for implantable electronics, to prevent inflammatory response due to CMOS materials and moisture damage to the device components.
The Revolutions of Scientific Structure
Date: Monday 8 September
Speaker: Colin Hales, The University of Melbourne
This is the debut seminar detailing the first of the two main scientific outcomes of the new book “The Revolutions of Scientific Structure”. The book resulted from an examination of what it would take to build an artificial scientist. In presenting this seminar the audience will be introduced to what might be the first formal act of science self-governance in the modern era. Until this book it can be claimed that self-governance is absent from science, whereas science self-regulation is brilliantly integrated into the daily life of all of us. The first scientific outcome in the book is measurement and documentation of a “Law of Scientific Behaviour”. It is a law of nature about the natural world of construction of laws of nature by their constructors: us (scientists). In the process a formal framework for the kind of science we currently do is created and named ‘appearance-aspect’ science.
That done, the fundamental limitation of appearance aspect science is identified. Science, as it is currently practised, will permanently fail to deliver an adequate science of the scientific observer. This science is currently operating under the name ‘science of consciousness’. The character of the failure is that unlike anywhere else in science, when it is done, we will be unable to hand to engineers the usual science deliverables. As a result, engineers will be unable to design and create an artificial scientist and then prove it. This is a procedural failure due to science operating at a scientific evidence boundary condition intrinsic to what we all assume is the only way to do science. The book then sets about an act of science self governance that adds a new kind of scientific behaviour that does not suffer the same problem. It is the only kind that will deliver a the necessary principled scientific account of the scientific observer. As a result engineers will be able to build an artificial scientist/scientific observer and prove it. The practical methods and interpretation of the new kind of science are outlined. Some scientists are already using it in other areas and don’t realise it. The new framework for science is called DUAL-ASPECT science.
The book is available here: http://www.worldscientific.com/worldscibooks/10.1142/9211.The Front-Matter (preface) and preamble (Chapter 1) are already accessible free from the publisher. Press release here: http://www.worldscientific.com/page/pressroom/2014-07-11-01
Nanites, drug delivery and microfluidics: simple complexity
Date: Monday 1 September
Speaker: Mattias Björnmalm, Department of Chemical and Biomolecular Engineering, The University of Melbourne
Through nanotechnology, synthetic structures can be engineered at the nanometer-level, thus creating materials that can interact with, and influence, biological systems at their very core. Intelligent ‘nanobots’ or ‘nanites’ are still far from reality, but in recent years sophisticated nano-systems have started to emerge. A prominent example is the use of responsive nanoparticles for drug delivery to increase the efficacy of current therapies, as well as to enable new ones. But despite the great promise of biomedical nanoparticles, only a few have made it to the clinic. In this talk, we discuss how new technologies can be used to overcome or circumvent key challenges facing particle-based drug delivery. We focus on microfluidics and lithography-based methods and present recent results demonstrating the potential of these techniques for the engineering of new and improved particles, as well as for evaluating their biological performance using biomimetic models. The ability of these methods to simplify both engineering and evaluation of increasingly complex particle-systems has helped to start the slow process of scratching away the latter half of the label ‘science fiction’ from concepts such as ‘nanites’.
Mattias Björnmalm obtained his MSc in Engineering within Nanotechnology from Lund University, Sweden, in 2012. Before coming to Australia he was a lab engineer within biotechnology at the KTH Royal Institute of Technology, Sweden, working within protein engineering of antibody-like molecules. Currently, he is a PhD student in Prof. Frank Caruso’s group at the Department of Chemical and Biomolecular Engineering, The University of Melbourne, and his research is primarily focused on microfluidic methods for engineering and evaluating drug delivery particles.
Biomaterials for Musculoskeletal Functional Tissue Regeneration
Date: Monday 25 August
Speaker: James C.-H. Goh, Department of Biomedical Engineering, National University of Singapore
Biomaterials including polymers, hydrogels, ceramics, metals and self-assembled materials have been widely used for bone, cartilage, ligament, tendon and dental tissue regeneration. Most importantly, nano scaled biomaterials provide some important and interesting properties to the traditional biomaterials, including high ratio of surface area to volume, enhanced mechanical properties, high purity and homogeneity, and outstanding magnetic, optical and electrical properties. Taking advantages of these novel properties, the nano-biomaterials can be manipulated to stimulate the chemical and structural similarities to natural musculoskeletal tissues. For instance, cartilage is hard to be regenerated due to lack of an efficient vascular system, limited progenitor cells and chondrocyte mobility in the dense cartilage extracellular matrix. However, nano-biomaterials can provide a biomimetic interface (or environment) to improve functions of chondrocyte, inhabiting and differentiation of progenitor cells. Moreover, the nano-biomaterials will enhance the lifetime and performance of tissue engineered scaffolds for cartilage regeneration (such as improved mechanical properties and hierarchical structures). On the other hand, the development of nano-biomaterials on bone regeneration is now focusing on two different directions: scaffolds for treatment of segmental defects (with structural functions) and cavity defects (such as fillers and injectable scaffolds). The utilization of nano-biomaterials in orthopedic applications has been proved to be an effective and innovative technique to enhance osseointegration and bone regeneration. Further, the nano-biomaterials have also been applied in dental applications, including dental implants and dental adhesives. Similar to bone implants, such dental implants will enhance osseointegration. Further, the nano-biomaterials based dental adhesives can also provide significant properties such as strong durability of the bond between teeth and adhesive, leading to new choices for dental restoration. The nano-biomaterials also have significant impacts on ligament/tendon regeneration. For example, nanostructured scaffolds are extremely helpful that can serve as structural and mechanical support for cellular function and tissue regeneration. In conclusion, countless opportunities and approaches have been brought by the development of nano-biomaterials to solve the current difficulties of implant materials, providing lots of success for nano-biomaterials in a variety of musculoskeletal tissue regeneration applications.
James C.-H. Goh is Professor and Head of Department, Department of Biomedical Engineering, National University of Singapore (NUS), Singapore. He was awarded BSc (1st Class) in Mechanical Engineering and PhD in Bioengineering by University of Strathclyde, UK. He has been conducting cutting edge research at NUS. He has published widely in musculoskeletal biomechanics and tissue engineering, and has given numerous invited/plenary talks in international conferences. He has organized many international conferences and was International Vice-President of the World Congress on Medical Physics and Biomedical Engineering (Seoul, Korea, 2006), Chairman of the Organizing Committee of the 3rd WACBE World Congress on Bioengineering (Bangkok, Thailand, 2007), Chairman of the Organizing Committee of the 6th World Congress of Biomechanics (Singapore, 2010), and Chairman of the Organizing Committee of TERMIS-AP Conference (Singapore, 2011). He currently serves as President of the Biomedical Engineering Society (Singapore), Vice-President of the International Federation of Medical and Biological Engineering, Member of the World Council of Biomechanics, and Secretary General of the Asia-Pacific Association for Biomechanics.
The role of macroscopic brain networks in seizure initiation
Date: Monday 4 August
Speaker: John Terry, College of Engineering, Mathematics and Physical Sciences, University of Exeter
In this talk we introduce the mathematical language of graph theory, which has evolved into a useful tool for studying complex brain networks inferred from a variety of measures of neural activity such as fMRI, DTI, MEG and EEG. In the context of neurological disorders, recent work has discovered differences in the structure of graphs inferred from patient and control cohorts. However, most of these studies pursue a purely observational approach; identifying correlations between properties of graphs and the cohort which they describe, without consideration of the underlying mechanisms. To move beyond this necessitates the development of mathematical modelling approaches to appropriately interpret network interactions and the alterations in brain dynamics they permit. In the talk we introduce some of the mathematical and computational modelling approaches we have taken to study epilepsy, exploring how differences in the properties of functional networks inferred from resting state EEG recordings of people with idiopathic generalised epilepsies can lead to a heighten probability of seizures. Our findings demonstrate the potential for a mathematical model based analysis of routine clinical EEG to provide significant additional information beyond standard clinical interpretation, which should ultimately enable a more appropriate mechanistic stratification of people with epilepsy leading to improved diagnostics and therapeutics.
Multiscale modelling of epileptic seizures, from macro- to micro-scopic dynamics.
Date: Monday 21 July
Speaker: Sebastian Naze, Institute of Systems Neurosciences, Marseille
Epileptic seizure dynamics span multiple scales in space and time. Understanding seizure mechanisms requires identifying the relations between seizure components within and across these scales, together with the analysis of their dynamical repertoire. Mathematical models have been developed to reproduce seizure dynamics from the micro-scale of a single neuron to a neural mass macro-scale, as either driven or autonomous processes. In this study, we start from a phenomenological model displaying features of electrical epileptic neural activity (i.e. macroscopic), and we move towards a more biologically realistic network of neurons (i.e. microscopic), while keeping track of the dynamical repertoire exhibited by the macroscopic system. Our model is composed of two neuronal populations, characterized by fast excitatory bursting neurons and regular spiking inhibitory neurons, embedded in a comMonday extracellular environment represented by a slow variable. By systematically analyzing the parameter landscape offered by the simulation framework, we reproduce typical sequences of neural activity observed during status epilepticus. We find that exogenous fluctuations from extracellular environment and electro-tonic couplings play a major role in the progress of the seizure, which supports previous studies and further validates our model. Simulated traces are compared with in vivo experimental data from rodents at different stages of the disorder. We discuss potential mechanisms underlying such machinery and the relevance of our approach within a multi-scale paradigm, supporting previous detailed modelling studies and reflecting on the limitations of our methodology.
Sebastian Naze studied Networks and Telecommunications at the University of Tours, France, followed by a Bachelor degree in Computer Science at the Heriot-Watt University, Edinburgh, Scotland, and a Master degree in Information Science at the Vrije University, Amsterdam, The Netherlands. Following courses in Artificial Intelligence, Sebastian gained interests in modelling cognitive processes in the context of psychiatric disorder so he completed Master thesis on Computational modelling of post-traumatic stress (PTSD), under the supervision on Jan Treur at the Agent systems department. Then further motivated in modelling biological processes, Sebastian Naze started a PhD in 2012 under the supervision of Viktor Jirsa and Christophe Bernard at the Institute of Systems Neurosciences in Marseille, working on multi-scale modelling of epileptic seizures.
A conductance based model of intrinsic sensory neurons of the gastrointestinal tract
Date: Monday 21 July
Speaker: Jordan Chambers, University of Melbourne
The enteric nervous system regulates function of the gastrointestinal tract. Intrinsic sensory neurons (ISNs) of the enteric nervous system play a crucial role in this regulation because they respond to stimuli and drive many intestinal motor patterns and reflexes. ISNs express a large number of voltage and calcium gated ion channels, but how interactions between different ionic currents in ISNs produce both normal and pathological behaviours in the intestine remains unclear. A conductance based model of ISNs was constructed based on electrophysiological recordings and data from the literature. The model included voltage-gated sodium and potassium channels, N-type calcium channels, big conductance calcium dependent potassium channels, calcium dependent non-specific cation channels, intermediate conductance calcium dependent potassium channels, hyperpolarisation activated cation channels and internal calcium dynamics. The model reproduced several key physiological observations. A sensitivity analysis indicated which conductances have the largest influence on the excitability of ISNs. In conclusion, the model identifies how interactions between different iconic currents influence the excitability of ISNs and highlights an important role for Ih in enteric neuroplasticity resulting from disease.
Jordan completed his Bachelor of Science (Hons.) and PhD in the Department of Physiology, University of Melbourne. He then joined the Ion Channels and Disease Group at Howard Florey in 2010 before returning to the Department of Physiology 2011. Recently, Jordan has joined BME to work on attention in auditory perception.
Towards closed-loop stimulation strategies in bionic devices
Date: Monday 14 July
Speaker: Matias Maturana, University of Melbourne
Currently, retinal prosthesis and many other medical bionics devices use open-loop stimulation strategies. That is, the level of stimulation does not depend on any continuous measurements of neural activity. Closed-loop stimulation strategies have been implemented in some medical bionics devices, such as functional electrical stimulation, with great success, and have shown the benefits of closed-loop systems. We propose that a closed-loop stimulation strategy in a retinal prosthesis could improve reliability of retinal ganglion cell responses, mitigate neural fatigue, require lower power consumption, and lead to a better visual perception. We present a model for predicting neural responses to electrical stimulation derived using in vitro data. We show how this type of generalised model can be used to design a model-based controller for the control of neural responses, and propose that this method could also be implemented in other medical bionic devices.
Matias completed Bachelor of Arts and Science at the University of Melbourne in 2006 and Master of Engineering (Electrical) in 2012. During his Masters, Matias worked part-time at Bionic Vision Australia doing computational modelling of the intrinsic properties of retinal ganglion cells. His interest in visual neuroscience led him to commence his PhD in 2013, where he is looking at improving stimulation strategies for the bionic eye.
Mapping the human connectome with MRI: Promise, progress and pitfalls
Date: Monday 7 July
Speaker: Assoc Prof Alex Fornito, Monash University
The human brain is an extraordinarily complex network, comprising billions of neurons connected by trillions of synapses. Generating a comprehensive map of these connections‹a so-called human connectome‹across multiple spatial resolution scales has become a central goal for neuroscientists.
Magnetic resonance imaging (MRI) has featured prominently in such attempts, representing the only technique allowing in vivo mapping of whole-brain connectivity in human volunteers. The application of graph theory and complex network science to such data has driven rapid advances in our capacity to map and model brain structure and function in both health and disease. This talk will provide an overview of the basic principles underlying a connectomic approach to the human brain, discuss recent progress and highlight limitations that must be addressed if we are to make further gains in our understanding of the brain.
Alex completed his Masters (Clinical Neuropsychology) and PhD in 2007 in the Departments of Psychiatry and Psychology at The University of Melbourne, followed by Post-Doctoral training in the Brain Mapping Unit at the University of Cambridge, UK. He joined the School of Psychological Sciences at Monash University in 2013 and is currently Deputy Director of Monash Clinical and Imaging Neuroscience and an ARC Future Fellow.
Behavioural Performance and Neural Activation Patterns in Partial Hearing Cochlear Implanted Cat Model
Date: Monday 23rd June
Speaker: Yuri Benovitski, Bionics Institute
Animal behavioural studies make a significant contribution to hearing research and provide vital information which is not available from human subjects. Animal psychoacoustics are usually extremely time consuming and labour intensive; in addition, animals may become stressed, especially if restraints or negative reinforcers such as electric shocks are used. To address these issues, a novel psychoacoustic experiment was designed and vigorously tested. Measured frequency discrimination thresholds were stable, repeatable and comparable with previously published results. The same psychoacoustic experimental setup was used to study chronic performance of partially deafened cats using a cochlear implant. Four cats were systematically tested on different reference frequencies with intra-cochlear electrical stimulation turned on and off. This study was able to show, for the first time, that cats can utilize information provided by a CI in performing a behavioural frequency discrimination task. Moving up the auditory processing pathway, behavioural performance was compared to acute neural activation patterns in primary auditory cortices of the same animals. A new data analysis technique allowing direct comparison of psychoacoustical and electrophysiological recordings was developed. This behaving animal model allows studying how electric and acoustic stimulation of the cochlea are combined. This is particularly important as more people with significant amount of residual hearing receive cochlear implants.
Yuri Benovitski is currently submitting his PhD thesis with the Bionic Institute and the department of Engineering at La-Trobe University. He holds a bachelor of Electronic Engineering (Hons) from RMIT University. His research objectives are combining Biomedical Engineering and Neuroscience by using novel experimental setups and data analysis techniques.
Optimal Control of an Epileptic Neural Population Model
Date: Monday 16th June
Speaker: Assistant Professor Justin Ruth, Singapore University of Technology and Design
Neural population models describe the macroscopic neural activity that can be clinically recorded by an electroencephalogram (EEG). Such models are relevant for the investigation of many pathological neurological phenomena including epilepsy and Parkinson's disease because the models operate on the same scale as the recorded data. Although several models exist in the neuroscience literature, none have leveraged the systematic approach of optimal control theory to design stimuli to treat such neurological conditions. In this talk, I will present the model and a formulation of the seizure abatement goal expressed as an optimal control problem. I will show several results including a realistic, noise-driven simulation where the control is applied as needed in a moving window.
Justin Ruths is an assistant professor at the Singapore University of Technology and Design with the faculty of Engineering Systems and Design. Justin holds degrees in Physics (BS, Rice University), Mechanical Engineering (MS, Columbia University), Electrical Engineering (MS, Washington University in Saint Louis), and Systems Science and Applied Math (PhD, Washington University in Saint Louis). His research themes include casting problems in the natural sciences and medicine as optimal control problems and investigating the control of large-scale complex systems. Towards this latter goal, some of his recent work is at the interface of control and network science.
Modelling and identification by Cellular Nonlinear Networks: seizure prediction in epilepsy?
Date: Monday 2 June
Speaker: Prof Ronald Tetzlaff, Technische Universität Dresden, Germany
Approximately 1% of the world’s population is affected by epilepsy, which is the most comMonday chronical neurological disorder worldwide. The problem of detecting a pre-seizure state in epilepsy using EEG signals has been addressed in many contributions by various authors over the past decades. Recently, Chua has shown that the emergence of complex behavior (e.g. pattern formation) in reaction-diffusion networks is based on local activity and especially on a subset called the “edge of chaos” in the parameter space of these networks. In the presentation, the theory of local activity will be introduced. Furthermore, an identification procedure for reaction-diffusion networks will be proposed based on the theory Cellular Nonlinear Networks which are characterized by local couplings of dynamical systems of comparably low complexity. By using Electroencephalogram (EEG) signal segments in epilepsy, Reaction-Diffusion Cellular Nonlinear Networks (RD-CNN) have been determined and analyzed. Results will be given discussed in the presentation.
Ronald Tetzlaff is a Full Professor of Fundamentals of Electrical Engineering at the Technische Universität Dresden, Germany. His scientific interests include problems in the theory of signals and systems, stochastic processes, physical fluctuation phenomena, system modelling, system identification, Volterra systems, Cellular Nonlinear Networks, and Memristive Systems. From 1999 to 2003 Ronald Tetzlaff was Associate Editor of the IEEE, Transactions on Circuits and Systems: part I. He was "Distinguished Lecturer" of the IEEE CAS Society (2001-2002). He is a member of the scientific committee of different international conferences. He was the chair of the 7th IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA 2002), of the 18th IEEE Workshop on Nonlinear Dynamics of Electronic Systems (NDES 2010), of the 5th International Workshop on Seizure Prediction (IWSP 2011) and of the 21st European Conference on Circuit Theory and Design (ECCTD 2013). Ronald Tetzlaff is in the Editorial Board of the International Journal of Circuit Theory and Applications since 2007 and he is also in the Editorial Board of the AEÜ —International Journal of Electronics and Communications since 2008. He serves as a reviewer for several journals and for the European Commission. From 2005 to 2007 he was the chair of the IEEE Technical Committee Cellular Neural Networks & Array Computing. He is a member of the Informationstechnische Gesellschaft (ITG) and the German Society of Electrical Engineers and of the German URSI Committee.
Circuits and Systems for Electroceuticals
Date: Friday 30 May
Speaker: Prof Wouter Serdijn, TU Delft
In the design process of electroceuticals, such as hearing instruments, pacemakers, cochlear implants and neurostimulators, the tradeoff between performance and power consumption is a delicate balancing act. In this presentation I will cover techniques to deal with the acquisition and generation of electrophysiological signals and to provide reliable communication with and through the body. We will discuss signal-specific analog-to-digital converters, morphological filters, arbitrary-waveform neurostimulators, energy harvesting and ultra wideband wireless communication from a low-power circuits and system perspective. Design examples and their performance will be discussed and an avenue sketched for treatment of various neurological disorders, such as tinnitus and addiction.
Wouter A. Serdijn (M'98, SM'08, F'11) was born in Zoetermeer ('Sweet Lake City'), the Netherlands, in 1966. He received the M.Sc. (cum laude) and Ph.D. degrees from Delft University of Technology, Delft, The Netherlands, in 1989 and 1994, respectively.
His research interests include low-voltage, ultra-low-power and ultra wideband integrated circuits and systems for biosignal conditioning and detection, neuroprosthetics, transcutaneous wireless communication, power management and energy harvesting as applied in, e.g., hearing instruments, cardiac pacemakers, cochlear implants, neurostimulators, portable, wearable, implantable and injectable medical devices and electroceuticals.
He is co-editor and co-author of the books EMI-Resilient Amplifier Circuits (Springer 2013), Ultra Low-Power Biomedical Signal Processing: an analog wavelet filter approach for pacemakers (Springer, 2009), Circuits and Systems for Future Generations of Wireless Communications (Springer, 2009), Power Aware Architecting for data dominated applications (Springer, 2007), Adaptive Low-Power Circuits for Wireless Communications (Springer, 2006), Research Perspectives on Dynamic Translinear and Log-Domain Circuits (Kluwer, 2000), Dynamic Translinear and Log-Domain Circuits (Kluwer, 1998) and Low-Voltage Low-Power Analog Integrated Circuits (Kluwer, 1995). He authored and co-authored 7 book chapters and more than 250 scientific publications and presentations. He teaches Circuit Theory, Analog Signal Processing, Micropower Analog IC Design and Bioelectronics. He received the Electrical Engineering Best Teacher Award in 2001 and 2004.
He has served, a.o., as General Chair for IEEE BioCAS 2013, Technical Program Chair for IEEE BioCAS 2010 and as Technical Program Chair for IEEE ISCAS 2010 and 2012, as a member of the Board of Governors (BoG) of the IEEE Circuits and Systems Society (2006—2011), as chair of the Analog Signal Processing Technical Committee of the IEEE Circuits and Systems society, as a member of the Steering Committee of the IEEE Transactions on Biomedical Circuits and Systems (T-BioCAS) and as Editor-in-Chief for IEEE Transactions on Circuits and Systems—I: Regular Papers (2010—2011). He currently is TPC Co-Chair for IEEE ISCAS 2014 and General Co-Chair for IEEE ISCAS 2015.
Wouter A. Serdijn is an IEEE Fellow, an IEEE Distinguished Lecturer and a mentor of the IEEE.
Capturing Light-Fields on Chip: lens-less 3-D imaging in standard CMOS
Date: Wednesday 21 May
Speaker: Alyosha Molnar, Cornell University
Whereas traditional image sensors map the intensity of light at a particular plane, significantly more information is present in a field of light rays. In particular, by mapping the distribution of incident angle in a scene, “light-field” imaging permits passive extraction of 3-D structure from a single frame. I will present a new class of pixel, the “angle-sensitive pixel” (ASP) built in a standard CMOS manufacturing process. ASPs use pixel-scale diffraction gratings built from metal interconnect layers to generate a strongly angle-sensitive light response.
An appropriately chosen mosaic of ASPs provides a much richer description of incoming light and does so in a computationally compact format, similar to the Gabor filters used in many image-processing applications. I will discuss several applications for arrays of ASPs, including digital light-field photography, lensless far-field imaging, and near-field lensless 3-D imaging of fluorescent microscale sources. I will also discuss several recently developed variants on the ASP, especially applicable to problems in biomedical imaging.
Alyosha Molnar received his BS from Swarthmore College in 1997, and after spending a season as a deck-hand on a commercial Tuna fishing boat, worked for Conexant Systems for 3 years as an RFIC design engineer. He was co-responsible engineer developing their first-generation direct-conversion receiver for the GSM cellular standard. Starting graduate school at U.C. Berkeley in 2001, Molnar worked on an early, ultra-low-power radio transceiver for wireless sensor networks, and then joined a retinal neurophysiology group where he worked on dissecting the structure and function of neural circuits in the mammalian retina, using a combination of electrophysiology, pharmacology and anatomy. He joined the Faculty at Cornell University in 2007, and presently works on low-power software-defined radios, neural interface circuits, and new integrated imaging techniques. He is recipient of the DARPA Young Faculty Award in 2010, NSF CAREER Award in 2012, and Lewis Winner outstanding paper award at ISSCC in 2012. He is presently associated with the University of Melbourne as a visiting professor.
Bionic Voice — Restoring Natural Voice for the Severely Speech Impaired
Date: Monday 19 May
Speaker: Farzaneh Ahmadi, University of Sydney
The most important mechanism of human communications is speech and the larynx is the only source of voice production in the human speech generation mechanism. When a person speaks, the sound waves generated at his larynx pass through the vocal tract and are shaped into different phonemes by the changes in the shape of the vocal tract. If for any reason the larynx fails to function (e.g. as the result of larynx cancer or aphonia), the person will lose the ability to generate voice and will remain merely with the possibility of producing limited whispers. These “voice loss” patients suffer greatly from unintelligibility issues and have difficulty communicating in daily life and over the phone. During the process of speech generation, motor neural signals are generated in the speech and larynx motor cortex of the brain and are transferred to the face and larynx muscles to control their movements. Despite the vast amount of research in modelling human speech production mechanism, limited effort has been made in discovering how the underlying neural activity controls the face, neck and larynx muscles in the speech production. This research aims to discover the neuromuscular mechanism of larynx movement generation and control and use this understanding to implement a Bionic Voice (voice prosthesis with neural interface) prototype which can be tested in-situ in a variety of users. As the first stage, the research correlates the human larynx movements with neuromascular activity of face and neck muscles during voice production. The project aims to be the first in Australia to apply these findings to develop a Bionic Voice Prosthesis that will significantly enhance voice generation and intelligibility of speech for the voice loss patients.
Dr Farzaneh Ahmadi has a background in Electrical engineering. Her area of expertise is signal processing in general and her areas of interest are bio-signals processing, computational neuroscience, brain dynamics and bionic systems. She has a main interest in biomedical research and especially bionic technologies to aid the disabled. She has started working on Bionic Voice research since her PhD at Nanyang Technological University of Singapore and subsequently has decided a research career to develop a working bionic voice solution for voice loss patients. Farzaneh Ahmadi has been the highest ranked awardee of two research grants including the GPRW research fellowships at the University of Sydney in 2013, She is currently working on the question of how to control a bionic voice solution using a combination of physiological attributes. Bionic voice will act as an artificial larynx for patients who lose their larynx functionality.
Opportunities and Pitfalls in Biology for Engineers
Date: Monday 19 May
Speaker: Tristan Croll, Queensland University of Technology
With apologies to Albert Einstein:
Biology without engineering is lame; bioengineering without biology is blind
For much of its history, biological science has held a reputation as a field of study for people who can’t handle physics, chemistry or math. For a great deal of time this was, in a sense, true – but not because these fields are irrelevant to biology. Far from it: rather, the “big problems” in biology were so far beyond the scope of the tools available as to make them effectively intractable. Thus, biology was relegated to a “soft” observational science, while the “hard” scientists honed their skills on easier problems.
The status quo has, of course, changed dramatically in the past few decades, as computational and theoretical resources have grown to the point where we can begin to seriously tackle what is, in reality, perhaps the hardest of all hard sciences – and engineers are well placed to excel in this space. After all, what is protein folding if not the world’s most challenging structural engineering challenge? What is a cell if not the world’s most complex chemical processing plant? What is a protein interaction network if not the most tangled spaghetti code, generated not by the work of an incompetent programmer but by the better part of 4 billion years of blind evolution? Such challenges are well within the scope of engineering principles and, in fact, must be met if the nascent field of bioengineering is to truly succeed and flourish.
However, there are pitfalls here for the unwary engineer. In particular, in the world of biology where the sole guiding principle is, in effect, “what works, works” the engineer’s training in making simplifying assumptions can prove disastrously counterproductive. Conversely, important problems in biology may go unsolved for decades simply because those active in the field lack the necessary background to understand concepts that to a suitably qualified engineer would seem trivial.
In a deliberately provocative talk cast through the lens of my various research projects spanning much of the past decade, I will discuss examples of each of these cases, and expand on the challenges and potential opportunities they represent, the need for both biologists and engineers to work together with mutual respect, and the need for a new generation of research scientists to explicitly treat biology with a hard-science approach. Particular scientific topics to discuss include:
- Why are most solid polymers (the primary focus of a few decades of tissue engineering research) probably dead ends for most tissue engineering applications?
- Why is surface immobilization of normally cell-bound ligands such as stem cell factor probably insufficient to improve hematopoietic stem cell culture?
- Why are transglutaminases probably some of the most important enzymes you’ve never heard of?
- What can be done to accelerate both teaching and research in the challenging field of structural biology?
- What bioengineering problems can we most successfully attack right now while we gather the necessary understanding for the more complex challenges?
Dr Tristan Croll began his academic career with a first-class Honours degree in Chemical Engineering at the University of Queensland. Having developed a growing appreciation of the potential benefits of applying engineering techniques to biological and medical problems, he leaped at the opportunity to become one of the inaugural PhD students of the Tissue Engineering group in the University of Melbourne Department of Chemical Engineering. After graduating in mid-2006, his subsequent postdoctoral experience in the Australian Institute of Bioengineering and Nanotechnology at UQ added to his growing appreciation of the many fundamental biological unknowns hampering progress in bioengineering. This ultimately led to the difficult choice to make a substantial change in the direction of his career. With this in mind, he moved to QUT’s Institute of Health and Biomedical Innovation in 2008, and since that time has focused his research primarily on questions regarding the fundamental biology of the wound healing process, while maintaining an interest in the development of enabling technologies for study of processes relevant to biomedical engineering. He has active research projects in the structural biology of cell surface receptors, understanding of novel interactions modulating the signalling of insulin-like growth factor I, and probing the in vivo biological roles of transglutaminases, a critically important yet poorly understood family of enzymes. In addition, he has spent the past few years developing a low-cost, large-volume 3D bioreactor design to eventually be released on an “open source” basis to facilitate study in the challenging field of 3D cell culture.
Dr Croll is currently a teaching/research lecturer at QUT, where he spends his non-research time teaching students concepts in molecular structure, enzymology, cell signalling pathways and general cell biology.
Threshold Concepts and Disciplinarity in Biomedical Engineering Education
Date: Monday 12 May
Speaker: Paul Junor, Senior Lecturer, Electronics & Biomedical Engineering La Trobe University
The task of consolidating and streamlining the multidisciplinary strands of Biomedical Engineering poses a challenge for course design. A relatively contemporary development in education, the Threshold Concepts Framework (Meyer and Land, 2003) offers some potential assistance to this ongoing process: “It has been discovered that, not only can threshold concept theory help in focusing students’ and teachers’ attention, it can also be a tool for curriculum development where there is a tendency to overcrowd the curriculum” (Male & Baillie, 2011).
The threshold concepts notion has attracted widespread interest and has been used in a number of disciplines such as mainstream engineering (Entwistle et al, 2005; Flanagan et al, 2010; Male & Baillie, 2011; Scott & Harlow, 2012). There are also number of publications emerging relating to applications in healthcare education, but an initial survey revealed no existing literature applying threshold concepts to Biomedical Engineering.
Tempered by considerations of disciplinarity (Shulman, 2005; Donald, 2011), this presentation considers the appropriation of suitable proxies from acknowledged threshold concepts of related fields such as Physiology, Biology, Medicine and Surgery in addition to those of the established engineering disciplines, to inform and guide further planning for specific threshold concept identification for Biomedical Engineering.
Paul Junor is a Senior Lecturer in Electronics & Biomedical Engineering at La Trobe University. Immediately prior to joining the Department of Electronic Engineering there in 1993, he had been a Biomedical Engineer at Peter MacCallum Cancer Institute from 1982, before which he had worked for a private radiology company, and had held teaching and research positions (RMIT and Monash Medical School). He is immediate past-president of the Society for Medical & Biological Engineering (Vic.) (2005-2010); a Senior Member of IEEE (currently chair of the Victorian IEEE Education Society Chapter) and a Fellow of Engineers Australia (currently Victorian representative to the EA Biomedical Engineering College Board).
The Human Connectome in Health and Disease
Date: Monday 5 May
Speaker: Andrew Zalesky, University of Melbourne
A central goal in neuroscience is to comprehensively map the network architecture of the human brain, known as the human connectome. Mapping and deciphering this amazingly complex neural wiring diagram will reveal much about what makes us uniquely human and provide novel insights into disorders of the brain and mind. From the perspective of an engineer that has been working in human connectomics soon after the field’s inception in 2005, I will present some of the unique computational and modelling challenges associated with mapping brain networks and give examples of how these challenges have been solved using traditional computer science and engineering approaches. In this talk, I will specifically focus on mapping the large-scale connectome with magnetic resonance imaging (MRI) techniques. I will discuss shortest path approaches that I have developed for tracking the trajectories of axonal fibre bundles in a computational process known as tractography. I will also discuss the network-based statistic and applications in neuropsychiatric disorders, including schizophrenia and addiction.
Andrew Zalesky is a senior research fellow at the University of Melbourne. He currently holds an NHMRC research fellowship and previously held numerous ARC fellowships. Dr Zalesky is known for publishing some of the first neuroimaging evidence of connectome pathology in schizophrenia. He is also recognised for developing the network-based statistic, a statistical method that is now widely implemented in brain image analysis software. He has published more than 70 peer reviewed articles; including two articles in the top 0.01% cited papers in the same year and field. Dr Zalesky is an editor of Brain Topography and currently serves on the editorial board of NeuroImage.
Salience and a task-based objective measure of image quality
Date: Monday 28 April
Speaker: Prof Murray Loew, Director of the Biomedical Engineering Program in the Department of Electrical and Computer Engineering at George Washington University
We present an objective image-quality measure that is correlated with perceived image quality as a function of the most conspicuous features contained within an image. Those salient features are determined by combining aspects of multiple disciplines (including psychology, vision science, and image engineering) to define a new measure that emphasizes the importance of contrast-based features as a function of spatial frequency, i.e., scale. Visual search models and visual psychophysics provide the framework for our channel model. We leverage previous work on just-noticeable discrimination models to develop a salience model for single images. Signal detection and estimation theory is used to develop a statistical basis for our scale-based salience measure.
The new measure is applied to medical images; salient features within mammograms are studied extensively. Our new salience measure has multiple potential applications. The development of a perceptually-correlated metric that is useful for quantifying the conspicuity of local, low-level or bottom-up visual cues can be used to improve reader training. This could be accomplished by presenting technicians, residents, etc., with images that are processed to enhance or degrade perceptually important frequency bands, thereby increasing or decreasing the perceptual “pop-out” within regions-of-interest. Further, images that can be grouped by known perceptual difficulty can be used as training and testing aids because selection is based upon perceptually-correlated image similarities or differences. This would complement the selection of cases based upon anatomical, histological or clinical requirements by providing a means of selecting cases using an image (contrast and scale) similarity measure. The salience metric was evaluated using a set of 40 mammograms and registered eye-position data from nine observers.
Murray Loew is a professor and director of the Biomedical Engineering Program in the Department of Electrical and Computer Engineering at George Washington University (Washington, D.C. USA). He is the inaugural recipient of the Fulbright Distinguished Chair in Advanced Science and Technology, sponsored by the Defence Science and Technology Organisation (DSTO). His research is in the area of medical imaging and image analysis, image registration, and image compression. Through his Fulbright, Prof. Loew will come to the DSTO laboratories in Adelaide for five months to work on object tracking and image and data fusion. He is a Fellow of the IEEE and of AIMBE.
Bistable Dynamics of Perceiving Ambiguous Stimuli
Date: Monday 14 April
Speaker: John Rinzel, Center for Neural Science and Courant Institute of Mathematical Sciences, New York University
When experiencing an ambiguous sensory stimulus (e.g., the vase-faces image), subjects may report haphazard alternations (time scale, seconds) between the possible interpretations. I will describe dynamical models for neuronal populations that compete through mutual inhibition for dominance - showing alternations, behaving as noisy oscillators or as multistable systems subject to noise-driven switching. In highly idealized formulations networks are percept specific without direct representation of stimulus features. Our recent work involves perception of ambiguous auditory stimuli (e.g., http://auditoryneuroscience.com/topics/streaming-galloping-rhythm-paradigm ); the models incorporate feature specificity, tonotopy, so that perceptual selectivity is emergent rather than built-in.
Reduced-order modelling of perinatal cardiovascular dynamics and congenital heart disease
Date: Monday 7 April
Speaker: Dr Jonathan Mynard, University of Melbourne
In newborns with congenital heart disease, the cardiovascular system often displays 1) heart and vascular abnormalities that disturb normal blood flow patterns, and 2) a survival-dependent persistence of some features of the fetal circulation. For example, in pulmonary atresia with intact ventricular septum (PAIVS), the right ventricular outflow path fails to develop and pulmonary flow is dependent on a patent ductus arteriosus; in addition, profound coronary abnormalities may be present. Reduced order numerical modelling allows complex cardiovascular dynamics to be investigated with more flexibility and less ethical concern than in clinical or experimental settings. This talk will provide an overview of the development of numerical models of the entire circulation of the normal fetus and neonate, these being derived from a reference model of the adult circulation. A model of PAIVS is also developed to elucidate the determinants of coronary blood flow before and after surgical opening of the right-ventricular outflow path.
Jonathan Mynard completed undergraduate degrees in Medical Biophysics and Electronic Engineering at Swinburne University in 2005. He then earned a Master of Research in Computer Modelling at Swansea University, Wales, where he developed a one-dimensional model of the adult systemic arterial circulation. Jonathan completed his PhD at Melbourne University in conjuction with the Murdoch Children’s Research Institute, with his thesis focusing on cardiovascular dynamics in the perinatal period. He currently holds a CJ Martin Overseas Biomedical Fellowship from NHMRC, and he recently returned from two years at the University of Toronto, Canada where he worked in the Biomedical Simulation Laboratory gaining skills in image-based computational fluid dynamics.
Directions to the venue: The Mechanical Engineering building is located at the south end of the Parkville campus, in block E fronting Grattan street, please see 18K in the attached map. As you enter the building from Grattan street, go through the glass doors on your right. Take the lift to the third level. The lecture theatre E311 is down the corridor from the elevator (on the left, mid-way). You may find these links useful to find the venue:
Neural Network Model of Visual Cortex for Perception of Motion Transparency
Date: Monday 31 March
Speaker: Parvin Zarei Eskikand, University of Melbourne
Motion transparency relates to the situation where there are multiple motions perceived in only one spatial location. One of the main challenges of motion processing systems is that local motion signals do not represent accurately direction of the whole object. This is known as the “aperture problem”. The small receptive fields of individual neurons can only measure motion components orthogonal to their preferred direction. This problem makes the processing of motion transparency very challenging, because solving the aperture problem requires uniqueness in the motion domain, but perception of a transparent motion is based on the representation of multiple objects. The aim of this project is to model different layers of the visual cortex to demonstrate the neural processing that underlies the perception of motion transparency. In this talk, I will introduce a model that is able to detect the motion signals and discriminate these signals resulting from different objects in the input. The initial local motion components will be extracted at the first stage by a motion detector and then these components will be analyzed in higher layers to form an accurate estimation of motion of the individual objects.
Low power wireless transceivers for implanted medical devices and neural prostheses
Date: Monday 17 March
Speaker: Farhad Goodarzy, University of Melbourne
In this session a wireless transceiver for implantable medical devices (IMD), Neural prostheses (NP) and embedded neural systems is discussed. We'll talk about modulation schemes and a new modified technique, called saturated amplified signal (SAS). We analyze the system based on the underlying circuitry and the modulation scheme which provides a theoretical basis to compare and produce optimal low power wireless transceivers for biomedical applications. We also present a low power wireless transceiver in the MICS frequency band based on the modified modulation and present the results. The design is capable of being fully integrated on single chip solutions for surgically implanted bionic systems, wearable devices and neural embedded systems.
Towards a biophysically inspired model of the cerebellum
Date: Wednesday 5 February
Speaker: Tom Close
Despite much of the distinct regular structure and neurophysiology of the cerebellum being known from the time of Eccles fifty years ago, we still do not understand how the elegant motor control systems found in nature could arise from this neural circuitry. The most well known theory, proposed by Marr and Albus in the late 60's, draws an analogy between an adaptive filter and the arrangement Purkinje cells and parallel fibres. However, this explanation doesn't agree with many recent biophysical observations, such as the relatively short delays that occur in spike transmission along the parallel fibres and the timing of the "error signal" from the climbing fibre axon. Therefore, to work towards a theory of cerebellar function that can explain the neurological basis of highly skilful motor actions that currently cannot be reproduced by artificial systems, we are building a biophysically detailed model of the cerebellum to capture key biophysical characteristics with potential functional implications. However, highly detailed network simulations quickly become unmanageable using standard simulation software practices. Therefore, we have developed a software framework that separates the biophysically derived parameters from the algorithmic code required to initialise the simulation on large computer clusters. We then demonstrate how this framework can be used to quickly identify the effect of parallel fibre feedback on the input layer of the cerebellum.
Tom Close was awarded a PhD on "Advanced techniques in diffusion MRI tractography" from the Department of Electronic and Electrical Engineering at the University of Melbourne in 2011, under the wise tutelage of Leigh Johnston and Iven Mareels amongst many other supervisors. He is currently a post-doctoral researcher in the Computational Neuroscience Unit of the Okinawa Institute of Science and Technology, Japan, where he is following his passion of studying the behaviour of the cerebella network in between staring out his window at blue coral-filled waters.
Mathematical Modelling of Brain Networks: From Synaptic Plasticity to Behaviour
Date: Monday 20 January
Speaker: Rob Kerr
Professor David Grayden
T: +61 3 8344 5234