Skip to the content | Change text size

Health and wellbeing - iHealth: modelling the virtual patient

Health and Wellbeing

Modelling the virtual patient (iHealth), includes a number of projects that Faculty researchers are leading to address the key research priority of IT enabling Health and Wellbeing.

Projects in this area include: Medical Image processing and retrieval; Cardiac modelling; Knowledge Engineering Bayesian networks for medical applications and a cardiac monitoring system using genetic and clinical knowledge fusion.

Faculty researchers working in this field include Professor David Abramson, Professor Guojun Lu, Associate Professor Ann Nicholson, Dr Iqbal Gondal and Dr Kevin Korb.

 

A High Throughput Grid Based Environment for Real Time Bio-medical Imaging

Researchers: Prof David Abramson (David.Abramson@monash.edu), Blair Bethwaite (Blair.Bethwaite@monash.edu) , Minh Ngoc Dinh, Colin Enticott (Colin.Enticott@monash.edu ) , Slavisa Garic (Slavisa.Garic@monash.edu) , Hoang Nguyen, Tirath Ramdas, A.B.M. Russel; Ian Harper (Ian.Harper@monash.edu ), Stephen Firth (Stephen.Firth@monash.edu ); Martin Lackmann (Martin.Lackmann@monash.edu), Mary Vail (Mary.Vail@monash.edu), Stefan Schek, Thomas Zapf
Partners: Leica Microsystems Inc.
Centre: DSSE
Funding: ARC Linkage 2008-2010
Project website: https://messagelab.monash.edu.au/FundedProjects
Project outline: Since the first microscope, the development and sophistication of imaging technologies have set the pace in life-sciences. The ability to "distinguish" smaller and smaller structures continues to define the foundation for biological and bio-medical research. more...
Over the next 5 to 10 years we expect the resolution of microscopes will challenge the computational and storage capacity of current stand alone instruments.

This project will develop a software architecture that integrates image capturing hardware and data analysis and storage software into a Grid of high performance computers and storage devices. The resulting system will provide a powerful scalable solution for future developments.

Together with Leica Microsystems, we are building a virtual microscope facility that will provide substantial functionality not currently available in Australia. This facility will have major national and international impact on bio-medical imaging. The software solutions and infrastructure, developed as part of this program will have considerable commercial and strategic value in their own right. One guaranteed avenue for exploitation of the software will clearly be through our industry partner, Leica. Importantly, our project consolidates a critical mass of expertise connecting biomedical with computer science, thereby addressing a well-recognised constraint that to date has limited the national and international impact. less...
Key outcomes:  

Medical Image Processing and Retrieval

Researchers: Prof Guojun Lu, Dr Shyh Wei Teng, Dr Dengsheng Zhang, Martin Lackmann (Faculty of Medicine, Nursing and Heath Science) and Mary Vail (Faculty of Medicine, Nursing and Heath Science)
Partners: Leica Microsystems Inc
Funding: Two MGS scholarships (two PhD students are working on this project)
Project website: https://messagelab.monash.edu.au/FundedProjects
Project outline: Huge amounts of images are captured by medical researchers and medical parishioners. There are a number of problems related to the use of medical images. more...
Firstly, many images can be taken for a body part or tissue using different imaging instruments (e.g. different types of microscopes, CT, and X-ray) at different magnifications. These images are then examined and compared to provide diagnosis or determine the effect of a treatment. However, it is hard and time consuming for medical researchers/practitioners to determine the corresponding points in these images. Secondly, many stored images are not being used because it is too hard to search and find images based on certain criteria without any automatic indexing and retrieval systems.

This project will provide solutions to the above problems by developing automatic image registration, indexing and retrieval methods. less...
Key outcomes:  

Cardiac Modelling - Ion Flux in Cardiac Models

Researchers: Prof David Abramson, Andrew McCulloch, Anushka Michailova and Roy Kerchoffs (University of California, San Diego)
Project outline: Electrical activity of the heart is a product of that of the component myocyte muscle cells which in turn is produced by the flow of ions through ion channels in the myocyte membranes. more...
More detailed models of ion flow processes are being developed with the ultimate aim of understanding both normal and dysfunctional hearts. This project uses a new model of rabbit ventricular myocytes, running under Nimrod/O, to determine settings for the metabolic factors that would produce realistic behaviour. less...
Key outcomes:  

Ambient Cardiac Expert: A Cardiac Patient Monitoring System using Genetic and Clinical Knowledge Fusion

Researchers: Dr Iqbal Gondal, Shoaib Sehgal, Mudasser Iqbal, Megan Woods
Centre: DSSE
Funding: Monash small grant
Project outline: Recently, the health care industry has observed a growing demand of continuous monitoring of patients to timely diagnose the diseases. This is attributed to a remarkable growth in uncertain deaths caused by diseases like heart attack and malignant neoplasm. more...
While the existing sensor network technology has been envisaged as a solution to the real time patient monitoring, the focus of such efforts has been to deploy a sensing network and passively deliver the physiological readings from patients to the hospital infrastructures. In this project, we introduce a framework that addresses a drawback in existing experimental test beds: the lack of feedback from the diagnosis infrastructure to the sensor network so that the network can reconfigure itself to monitor the patient in a more useful manner, while concomitantly maximizing the network reliability, throughput and life.

This monitoring can be more effective if in addition to standard clinical parameters genetic information is used because of its ability to predict hereditary diseases like cardiac problems. Current clinical practices, however, only stress on physiological observation to predict heart failure rate which could miss the important information which could lead to fatal consequences. The system uses well established Support Vector Machines (SVM) for class prediction and uses Wrapper Evolutionary Algorithm based on Gaussian Estimation of Distribution Algorithm (EDA) to determine cardiac patient’s criticality. Results suggest that ACE can be successfully applied for cardiac patient monitoring and has ability to integrate the information from both clinical and genetic sources. less...
Key outcomes:  

Knowledge Engineering Bayesian networks for Medical Applications

Researchers: A/Prof A. Nicholson, Dr K Korb, Prof J. McNeil, Prof R. Neopolitan
Centre: CRIS
Funding: ARC Discovery 2004-2006
Project outline: Bayesian networks (BNs) are a leading technology in applied Artificial Intelligence (AI). By combining a graphical representation of the dependencies between variables with probability theory and efficient inference algorithms, BNs provide a powerful and flexible tool for reasoning under uncertainty. more...
Medicine has undoubtedly been the most popular application area for Bayesian networks to date. It is a complex domain where experienced medical practitioners implicitly hold much knowledge, and as such, has long been a target of expert systems. The popularity of BNs for medical application is based on their ability to explicitly model causal interventions, to reason both diagnostically and predictively and the visualisation of the graphical representation, which assists their use in explanation. BNs may be built by eliciting expert knowledge or learned via causal discovery programs such as CaMML, developed at Monash by Chris Wallace, Kevin Korb and others. Both approaches to building BNs have limitations: expert elicitation is expensive, time-consuming and relies on experts having full domain knowledge, while automated learning is often ineffective given small or noisy datasets.

We have used knowledge from the medical literature and longitudinal data (the Busselton data) to hand-craft a Bayesian network model of coronary heart disease, TakeHeart II, which supports risk assessment for individuals with and without treatment interventions.

We are currently combining elicited expert priors about causal relationships with CaMMLs BN structure learning. In collaboration with Dr. Julia Flores at the Universidad de Castilla-La Mancha, and Dr. Andrew Brunskill at the ANZData registry (Queen Elizabeth Hospital, South Australia), our current case study involves building a BN model for predicting heart failure from the IOWA longitudinal dataset. We are also developing a more general methodology for describing the knowledge engineering process. less...
Key outcomes: