Current projects in Intelligent Systems Research Centre
At ISRC we use world class facilities and ideas to make out work pioneering
Within the Intelligent Systems and Research Centre we have a variety of research areas that are being studied.
Functional Brain Mapping
The specialist equipment, which will be located at the Intelligent Systems Research Centre (ISRC), has the potential to position Northern Ireland as the location for advanced neurological drug evaluation and testing for global pharmaceutical companies.
The multidisciplinary brain imaging research project will be carried out by the ISRC, in conjunction with the Biomedical Sciences Research Institute and Institute of Nursing and Health Research at the University of Ulster and C-TRIC.
ISRC Director Professor Martin McGinnity said the Functional Brain Mapping Facility is a major boost for the ISRC and its ongoing efforts to understand the complexities of the human brain and research into neurological disorders, including depression and Alzheimer's disease.
“There is a clear need for a better understanding of how these begin and evolve. Such knowledge could lead to effective diagnostic tools for early diagnosis and more effective prevention or intervention measures.
“If we can understand better how the brain works, the impact on society will be enormous, both in terms of our health and employment in health related industries.
“Today we are closer than ever to understanding the operation and complexities of the human brain, in health and illness. As a result of this initiative, our researchers will be able to make a greater contribution to this great challenge, alongside the EU (via its Human Brain Project) and the USA (via its BRAIN - Brain Research Through Advancing Innovative Neurotechnologies) project.”
Professor McGinnity continued:
“Our vision is to exploit the capabilities of functional brain mapping to perform excellent research in neurological research and commercialisation.”
Announcing the investment Minister Foster said:
“The new brain mapping facility will strengthen the R&D infrastructure in Northern Ireland and has the potential to move us closer to being able to diagnose and treat some forms of mental illness, which is a truly worthwhile objective.
“Our Programme for Government recognises the importance of innovation and R&D in rebuilding and rebalancing the local economy. However, innovation on its own is only part of the picture. Commercialisation of research is what drives business and economic success.
“This project is an excellent example of an academic research project that has sound commercial application. Over the next 10 years it is anticipated that it will generate significant revenue streams.”
Mitochondrial G Protein signaling in astrocytes: a new player in the tripartite synapse
Collaborators: Intelligent Systems Research Centre (CNET group) at Ulster University, NeuroCentre Magendie. AVENIR Group "EndoCannabinoids and NeuroAdaptation at the University of Bordeaux in France, Laboratory for Neuron-Glia Circuitry at the RIKEN Brain Science Institute in Japan and the Department of Neuroscience at the University of Minnesota, USA
Astrocytes, classically considered as simply supportive cells for neurons, are emerging as relevant elements in brain information processing through their ability to regulate synaptic activity. Indeed, the tripartite synapse formed by pre- and post-synaptic neurons and surrounding astrocyte structures has been proposed as a functional unit of brain processes. These novel and important functions of astrocytes are under control of G protein signaling-dependent processes, which trigger astrocyte Ca2+ signals eventually leading to the release of gliotransmitters and other mediators regulating synaptic functions.
Recent evidence indicates that the roles of mitochondria in the brain may go beyond the mere needs of energy supply for cell survival and maintenance, being possibly involved in the regulation of synaptic functions. It is conceivable that mitochondrial G protein signaling participates in these processes. Various subtypes of G protein-coupled receptors (GPCRs) and the associated signaling molecular elements are present within mitochondria, suggesting the existence of mitochondrial G (mtG) protein signaling pathways. Mitochondrial type-1 cannabinoid receptors (mtCB1) are an example of GPCRs regulating mtG protein signaling in brain astrocytes. Thus, astroglial mtG signaling potentially plays an important role in the regulation of tripartite synapse and hence brain functions. However, no studies have addressed this issue so far. The present project proposes to investigate the consequences of the activation of astroglial mtG signaling pathways in brain physiology, identifying the underlying mechanisms at cellular, network, behavioral and theoretical modeling levels. We propose to generate and develop novel pharmacogenetic tools (DREADDs specifically expressed by astroglial mitochondria, mtDREADDs) that will allow the specific control of astroglial mtG protein activity. We will experimentally and theoretically analyse the consequences of activation of different mtG proteins (via mtDREADDs or mtCB1 activation) on neuronal, synaptic, and network activity as well as brain functions in living animals. The expected results will reveal novel processes of cellular signaling in the CNS, and will identify new regulatory mechanisms mediated by astroglial cells in brain function.
SPANNER: Self-repairing Hardware Paradigms based on Astrocyte-neuron Models
Collaborators: Intelligent Systems Research Centre (CNET group) at Ulster University and the Department of Electronic Engineering at the University of York
Living organisms are complex systems, and yet they possess an extremely high degree of reliability. Failure mechanisms in nature are often local and their repair is also undertaken at this local level. In engineering however, we have traditionally approached the problem of unreliability from the system or sub-system level. That is, we have incorporated redundancies by replicating entire systems or sub-systems, in the hope that at least one would still function faultlessly when the others fail. It has been suggested recently that interaction between neurons and astrocytes may hold the key to repair in large networks of neurons. We could therefore, justifiably ask the question whether it might not be more effective, efficient and less costly to draw inspiration from nature, seeking to learn how it deals with the complexity vs. unreliability issue in such a remarkable way. This project is an inter-disciplinary collaboration that arose naturally from the combined expertise of the Intelligent Systems Group (ISG) at the University of York (UY) and the Intelligent Systems Research Centre (ISRC) at Ulster University (UU). The fundamental astrocyte-neuron computer model for self-repair proposed in this project was proven in an earlier EPSRC eFutures funded project (EFXD12011) where it was demonstrated that the co-existence of astrocytes with spiking neurons in a network can yield a fault diagnostic and repair capability at the cellular level, which addresses current hardware reliability challenges [1-6]. This proposed project will demonstrate that the Self-rePairing spiking Neural NEtwoRk (SPANNER) is capable of diagnosing faults and subsequently performing repairs beyond existing levels, where the repair capability is showcased in hardware using real-time robotic applications.
Self-Repairing Neural Controllers for Autonomous Chemical Identification
Collaborators: Intelligent Systems Research Centre (CNET group) at Ulster University and the Department of Electronic Engineering at the University of York
This project will demonstrate a fault tolerant autonomous robotic system that is able to continually detect, in real-time, changes in the air environment, and construct a hazard map of potential threats. At the same time the system will have an autonomous control and navigation system implemented using a state of the art approach based on self-repairing neural networks. These networks go way beyond traditional paradigms by including astrocyte cells which have recently been modelled to capture the repair capability exhibited in the human. This will afford a resilience to various types of potential failure within the controller; resilience of the controller is initially proposed as way of demonstrating capability. Human operators will be able to provide real-time feedback to the performance of the system to allow for machine learning to improve the overall performance of detection and navigation through a reinforcement based approach to learning. This proposal will deliver a proof of principle demonstrator of a fault-tolerant autonomous robotic system, capable of mapping hazards chemical environments and identifying key hazards of interest. Human users will be able to have real-time feedback of the hazard map and items identified, as well as perform limited control of the unit to compliment the units autonomy. The work is based on previous research, by the investigators, on self-repairing neural networks, real-time anomaly detection and robotic deployment.
BIONICS: BIO-iNspired survellianCe System
Automated artificial vision technology is becoming more widely adopted for monitoring and surveillance applications to reduce the need for physical resources and human intervention. These systems suffer from weaknesses in artificial vision technologies resulting in imprecise learning and false or missed detections. Biological visual systems are vastly superior in terms of performance for real-time and low-power applications when compared with conventional artificial vision technologies. To enable automated surveillance technology to cope with the challenges of dynamic visual conditions, existing artificial vision technology must be improved. Biologically inspired artificial vision algorithms have already demonstrated their effectiveness in dealing with these challenges.
BIONICS is an InvestNI Proof of Concept project building on the retinal ganglion cell models developed in the VISUALISE project. BIONICS aims to take the VISUALISE ganglion cell model technology towards a marketable product of a biologically inspired hardware image encoder and feature detector for security cameras.
Computational Approaches to real-time energy trading
Collaborators: Intelligent Systems Research Centre (Cognitive Robotics group) at Ulster University and ClickEnergy
The ability to collect, manage, analyse and report on massive volumes of market data is vital to long term market success. This project will focus on the development of computational intelligence algorithms to analyse and interpret the wide variety of data that have a direct impact. This project is funded by Innovate UK under the Knowledge Transfer Partnerships scheme.
It is well known that various terrorist and criminal organisations are communicating undetected via digital images using advanced steganography techniques through the Internet. Secret messages encrypted and encoded in various ways can be sent over the internet or posted somewhere on a website. Encrypting coded messages in digital images, such as pornographic pictures, is the least detectable way for terrorists and criminals to communicate. It is extremely difficult for investigators to track such steganography across the internet because, by its very nature, it is difficult to detect. The art of discovering the existence of steganographic data or secret messages in a digital media is called steganalysis.
FLAME (Forensic image anaLyzer frAMEwork) detects the presence of hidden messages in images sent over the internet or incorporated into a website page. The FLAME core technology consists of machine learning based steganalysis algorithms and an image tracking algorithm. FLAME will therefore be at the core of a system capable of locating, collecting, and analysing images placed or exchanged over the Internet by terrorists and cyber criminals. FLAME is funded by Ulster Research Impact Award fund.
Key technical & innovative features:
FLAME consists of four components: a crawler module, a decision module, a steganalysis module and a database. The crawler module will be fed with a series of start URLs (website addresses) to begin crawling. Based on the seed URLs subsequent URLs will be discovered by parsing the contents of the starting URLs. The crawler will attempt to locate image files at these URLs and pass them to the decision module.
The decision module decides if the image has already been encountered by calculating the cryptographic hash of the image (SHA1) and looking it up in the database. If the image exists, the image is discarded from the temporary store. If not, the image will be stored in the database.
The steganalysis module obtains images from the database and tests them for steganographic content using open source image steganalysis sources such as Outguess, novel high level intelligent machine learning algorithms and statistical tests. If the image is found to contain steganographic content, it will be tested further to extract the hidden content from the image if possible. If not, the image is marked as analysed.
The images (both analysed and tested for hidden content extraction) are preserved until all results have been consolidated. Each of the four components can work independently of each other in order to allow for separation of concerns, as well as allowing for the possibility of running them on separate machines.
FLAME early interest and endorsement:
A white paper based on the FLAME technology was submitted to the U.S. Army CyberQuest 2016 competition (http://intelligencecommunitynews.com/army-announces-cyber-quest-2017/) and FLAME was selected for the second stage of the competition. This initial acceptance shows the potential of our technology given the US Military endorsement.
Legal Innovation Centre
Collaborators: Allen & Ovary, Baker McKenzie, HookTangaza, Caselines, Clio, Invest NI
The legal innovation centre will create a bridge between Law and Computer Science to the end of fostering and creating legal service provision innovation and advancing educational provision in Legal Technology. See https://www.ulster.ac.uk/legalinnovation for more information.
Royal Academy of Engineering - Ingenious Award
Collaborators: STEM Aware, Bespoke Communications
This Public engagement award is to publicise the need for cyber security specialists in addition to creating a cost-effective model to train degree-qualified engineers in the art of inspiring students to follow an engineering career.
RAEng/Leverhulme Trust Senior Fellowship
This Fellowship concentrates on reusing the existing WiFi AP deployment in homes for location-oriented activity recognition without the need for additional static Wireless infrastructure or wearable sensors. It also integrates device free passive localisation techniques.
HSC Commissioned Research - Dementia Care
The aim of this feasibility study is to investigate the effects of individual specific reminiscence, facilitated using bespoke software, on people with dementia and their carers.
Knowledge Transfer Partnership
Collaborators: Sentel Ltd, Belfast
This KTP aims to incorporate machine learning intelligent techniques within their call monitoring platform enabling identification of call fraud in real-time.
Fluid Software Interfaces (InterTrade Ireland)
Collaborators: Verify Technologies, Limerick
This Fusion project aims to create an automated customer portal 'configuration engine' to replicate the process of selecting a customer’s products features.
Management Information Systems (InterTrade Ireland)
Collaborators: Reprographic Systems Ltd, Dublin
This Fusion project will design and develop a bespoke, innovative, and integrated Management Information Systems (MIS) to facilitate change and enhanced efficiency.
Invest NI Proof of Concept
Collaborators: BBC, Vicomtech
This Proof of Concept (PoC) project titled “Broadcast Language Identification & Subtitling System (BLISS) investigates language identification and captioning and subtitling within the entertainment industry.
Predicting short-term wholesale prices on the Irish Single Electricity Market with Artificial Neural Networks (InterTrade Ireland)
Collaborators: Ark Energy Ltd, Longford
This Fusion project is to develop forecasting algorithms in order to predict short-term (72 hours ahead) wholesale prices on the Irish Single Electricity Market so that customers can make more informed trading decisions.
A BCI Operated Hand Exoskeleton based Neurorehabilitation System for Movement Restoration in Paralysis
This project is funded by UK India Education and Research Initiative (UKIERI) phase-II and the Department of Science and Technology (DST) Government of India under the DST-UKIERI Thematic Partnership program (DST-2013-14/126). This international collaborative project has three main components brain-computer interface (BCI), exoskeleton, and rehabilitation which are being undertaken under respective leaders Professor Girijesh Prasad of Neural Systems and Neuro-technology Research Team, Professor Ashish Dutta of Indian Institute of Technology Kanpur (IITK) India, and Professor Suzanne McDonough of Institute of Nursing and Health Research under the co-ordination of the PI Professor Prasad. Main project objectives are as follows:
- Develop a lightweight three-finger exoskeleton with embedded sensors, capable of replicating human motion for physical practice. It will be controlled by users’ EMG and EEG signals in assist-as-needed mode;
- Develop a novel brain-computer interface (BCI) that facilitates EMG and EEG for controlling the exoskeleton and provides visual neurofeedback to ensure focused physical and MI practices;
- Conduct pilot trials to evaluate the effectiveness of the exoskeleton along with BCI in movement restoration
It is known that much enhanced upper limb recovery can be gained if stroke sufferers with limb impairments, perform intensive active physical practice (PP) in conjunction with motor imagery (MI) practice (or mental practice) of activities of daily living.
Although a PP can be performed with the help of a therapist, it is expensive and limited and dependence on the therapist may lead to a passive practice. To this end, it is proposed to investigate development of a lightweight neuro-rehabilitation system for people with stroke that facilitates intensive active PP as well as MI practice with the help of a robotic exoskeleton and neuro-feedback from a novel non-invasive brain-computer interface (BCI). It will consist of a three-finger exoskeleton that can be worn by the subjects and will be controlled through the users’ Electromyography (EMG) and Electroencephalography (EEG) based BCI commands in assist-as-needed mode.
Additionally visual neuro-feedback from BCI will help ensure highly focused performance of PP as well as MI practice. The exoskeleton will be superior to existing ones as it will be able to replicate natural human finger motion with more degrees-of-freedom and be directly aimed at restoring critical hand functions, for grasping and manipulation of objects. The system will undergo pilot trial on a set of healthy individuals as well as people with impairments.