Centre for MEDICAL ENGINEERING RESEARCHCentre for Medical Engineering Research

Research areas

Below is a broad but not exhaustive overview of current and recent projects in this area.

Pablo Moscato

Lead researcher: Professor Pablo Moscato

A team of researchers led by Professor Pablo Moscato have developed sophisticated algorithms and innovative mathematical approaches that allow complex questions posed by 'big data' to be analysed from new perspectives. This research enables diseases such as breast cancer to be scrutinised at previously unseen levels, and results in the identification of disease subtypes.

Current medical procedures involve patients being given a drug for a disease, where the disease categories and definitions are generally quite broad. The identification of disease subtypes will lead to more targeted and effective treatment. Databases now boast thousands of drugs that can be used to treat cancers. Only with sophisticated computer analysis can you screen all of the combinations, and prioritise those that may be relevant for further investigation.

The research has the potential to allow for the analysis of the molecular profiles of affected tumours, as well as normal cells. This could eventually lead to the automatic determination of which drugs are most effective for individual patients, rather than using the lengthy trial-and-error process.

Lead researcher: Professor James Welsh

Collaborating with colleagues in biomedical science, Professor James Welsh is researching calcium release in heart cells to better understand irregularities in heart-beat, (arrhythmia) as well as reactions to the drugs used to treat this condition.

Calcium-induced calcium release is the mechanism that causes cardiac muscles to contract. Associate Professor Welsh is working to find an algorithm to enable simulation of calcium release in a complete heart cell. An accurate simulation has the potential to facilitate more efficient testing processes.

Below is a Final Year Project Video of one of Professor Welsh's students, Jeremy Stoddard, who worked on Calcium Release in Heart Cells.

Lead researcher: Professor James Welsh
Partners: Professor Alan Brichta, Dr Rebecca Lim and Professor Derek Laver

The vestibular system is an intrinsic sensory system that relays information to the brain about where a person is in space, how they relate to gravity; whether they are moving or still, if they are moving, how quickly, and in what direction. It is critical for normal human mobility and coordination.

Verifying the accuracy of non-live vestibular system testing

Research by biomedical scientists who seek a better understanding of the vestibular system is constrained by the difficulty of delicate experimental procedures that rely on using a live mouse with an electrode inserted into a single cell in its head, whilst moving the animal to disrupt its vestibular system.

In collaboration with colleagues from the discipline of biomedicine, Professor Welsh is working on a project to verify the accuracy of non-live vestibular system testing in mince. The team aim to replicate experiments done in a live mouse by using a vestibular system that has been removed from the animal. Associate Professor Welsh is assessing the accuracy of the experiments with the removed system by comparing the results with data from live experiments in published literature.

Classifying and detecting neurotransmitter responses

Experiments that measure inhibitory responses are important for developing a better understanding of how the vestibular system responds to drugs. Biomedical scientists are interested in measuring inhibitory responses from two types of neurotransmitters in the vestibular system of mice; gaba and glycine. However, in experiments in a live animal it is difficult to distinguish between gaba and glycine responses. Professor Welsh is analysing experimental data from a time series record to extract ‘features’ to distinguish between gaba and glycine responses.

Measuring / determining surface area of the cell by patch clamp methods

Professor Welsh is also working on measuring data from patch clamp methods to determine the surface area of vestibular and brain cells. By measuring surface area, researchers can determine whether cells are ‘open’ or ‘closed’, and whether they are able to transmit signals.

Lead researchers: Professor Frini Karayanidis and Dr Aaron Wong

Dr Aaron Wong is collaborating with colleagues from psychology on an ARC-funded project led by Professor Frini Karayanidis. Known as ‘Age-ility,’ the project examines how individual variability in the neural systems that enable executive functions or cognitive control across the lifespan, impact on adaptive functioning at different stages of life. Executive functions refer to those processes that orchestrate complex behavior into well-established and flexible routines.

For example, the team is seeking to determine the links between cognitive control and risk-taking behavior in the adolescent population. The project involves multiple massive sets of related data – including EEG and MRI scans, as well as survey data. Dr Wong’s role is to apply signal processing to computationally analyse these data sets using statistics and dimensionality reduction techniques in a high-performance computing facility. The aim is to reduce the complexity of the data to identify important features – such as amplitudes in the EEG signals or functional areas of blood flow in the brain from MRI data - that could lead to a better understanding of the links between cognitive control and risk-taking behavior.

Theta frontoparietal connectivity associated with proactive and reactive cognitive control processes, read more.

See also:

Lead researcher: Associate Professor James Welsh

Collaborating with colleagues from Singapore – an orthopaedic surgeon and a biomedical engineer – Associate Professor Welsh is working towards developing an intelligent implant to help lengthen leg bones in patients who are born with one leg longer than the other or who have had an accident that resulted in the shortening of a leg bone.

The project team is working towards an efficient way of operating the implant as well as the development of a new kind of intramedullary intelligent implant that includes a transmitter that can provide feedback on the progress of the bone growth by reporting back on whether it has extended and by how much.

Australian Artificial Pancreas Program

Lead researchers: Emeritus Laureate Professor Graham Goodwin and Dr Adrian Medioli
Associated research group: Australian Artificial Pancreas Program
Partners: John Hunter Children’s Hospital – Conjoint Associate Professor Bruce King, Dr Carmel Smart

8% of the world’s population is affected by diabetes. The disease costs 1.6 billion dollars per year in Australia. Diabetics cannot produce their own insulin. Treatment requires the delivery of exogenous insulin via a pump or syringe to moderate their blood glucose level (BGL). This involves measuring their current BGL via monitor or glucometer and determining how much insulin to administer and when. It is very difficult to achieve an optimal treatment scenario. There is significant scope for improvement in the treatment of diabetes to achieve a more normal BGL distribution.

At the Australian Artificial Pancreas Program (AAPP) our researchers are developing a closed loop algorithm to autonomously adjust insulin delivery as BGL rise and fall. Known as ‘SCEN1C’, the software uses complex control theory to calculate a patient’s insulin requirements in real time and initiates appropriate delivery. In its finalised form the SCENIC algorithm will offer patients interactive, automated closed loop insulin delivery.

The AAPP algorithm has being designed around a predictive capability that uses the patient’s own history to assess the likelihood of upcoming food and exercise activities. This closely mimics the decision-making process that patients currently contribute to their therapy. SCEN1C will make decision-making intelligent and pro-active, but it will still offer an element of consultation to give the patient complete freedom to deviate from the usual routine. Uniquely, the system will have the ability to learn as the patient’s eating and exercise habits change. This is particularly appealing in the context of growing children and adolescents.

The AAPP algorithm is being developed for type 1 diabetes and may be adapted for type 2 and other types of diabetes. The model has already achieved outstanding results in FDA approved pre-clinical trials and clinical trials are now being carried out.

Lead researchers: Professor Rick Middleton, Associate Professor Rohan Walker and Mr John Welsh

Electrical engineer Professor Middleton and bio-scientist Associate Professor Rohan Walker are jointly supervising a research project by PhD student John Welsh, which aims to reduce or eliminate the impacts of secondary neurodegeneration from stroke.

The primary and immediate neurological damage caused by a stroke is widely understood; a clot or bleed results in sudden necrotic brain cell death, then as cells die they disassemble and release breakdown products that damage neighbouring neurons. Initial stroke Treatment aims to deliver the right drugs to halt this chain reaction as quickly as possible.

However, there is increasing evidence that stroke can also cause progressive secondary neurodegeneration. This occurs because any nerves that are connected to the affected neurons become over or under-stimulated and either of these conditions can lead to the further cell death. This collaborative project is exploring the use of electrical stimulation to correct the activity in these neighbouring neurons and so preserve them and halt secondary degeneration.

Lead researcher: Professor Andrew Fleming
Associated research group: Precision Mechatronics Laboratory
Partners: Mr Ben Routley and Dr Ferdinand Miteff.

Guidewires facilitate the delivery of catheters, stents and other interventional devices in a wide range of diagnostic and therapeutic medical procedures. For example, in a vascular intervention to treat an aneurism or stroke, a guidewire is manipulated through the femoral artery into the brain. A catheter is then slid along the guidewire to deliver a device to the critical location. This delicate procedure involves the manipulation of the guidewire around many tight corners with only manual force and twisting to control the movement. It is common for the guidewire to get stuck during this process.

Researchers from the Precision Mechatronics Laboratory are exploring ways to make the device easier to manipulate. They have ruled out electrostatic levitation of the guidewire but are testing the possibility of magnetic levitation. They are also working on the development of a hydraulic flexible catheter that is able to bend, crawl, and sense force.

Lead researchers: Dr Aaron Wong and Dr Patrick Cooper

Dr Aaron Wong is collaborating with psychology researcher Dr Patrick Cooper to explore the use of a single electrode EEG device integrated with computer games and custom software to deliver a unique interactive mindfulness coaching experience. The team have modified an off-the-shelf EEG device and written interface to software to process the EEG data to allow a participant to play the computer game ‘Mario’.

From the EEG device’s single electrode, they have succeeded in pulling out three types of signals and have equated these with controls for the computer game. If the participant blinks Mario will jump, if they relax Mario will move forward, if they are stressed Mario will go backwards – the last two are based on time frequency components in the data. The team are currently exploring how the technology can be extended to allow a participant to control a Darwin robot.

Professor Sarah Johnson

Lead researcher: Professor Sarah Johnson
Associated research centre: Priority Research Centre for Stroke and Brain Injury
Partners: Associate Professor Rohan Walker and Professor Michael Nilsson

The aim of physiotherapy exercises for a stroke patient is to shift brain function from the ‘dead’ part of the brain to another area, thus allowing the patient to regain lost brain function. The standard level of physiotherapy treatment recommended for a stroke patient is a minimum of 1 hour of active practice per day However, due to limited access, the typical level of treatment patients actually receive is significantly less than this. As a result, their recovery is often compromised.

Collaborating with colleagues from biomedical sciences, neuroscience and rehabilitation, Professor Sarah Johnson is applying systems engineering to develop a sensor enhanced motor learning approach to augment physiotherapy treatment and improve rehabilitation outcomes for stroke patients. The project aims to support patients to optimize their physiotherapy exercises without expert supervision and in a non-clinical environment, such as their own home.

Neuro-feedback driven stroke rehabilitation

Firstly, a visual model of the ideal movements for each exercise is modelled through readings from a movement tracking suit, which uses gyroscopes that track range motion (the suit is worn by a physiotherapist who demonstrates the exercises). This expert model is then incorporated into software that the patient can use in their home. The patient then does their exercises whilst wearing a movement tracking suit and the software provides real-time guided feedback on how to adjust their actions. The system also records the patient’s performance for review by an expert.

Secondly, non-invasive NIR (near-infrared) imaging is used to shine light on the brain to measure activity in different areas. As a particular part of brain becomes active blood flow increases oxygen levels. Because the oxygenated areas reflect and refract the light differently, the infrared image can show which areas of the brain are active. If the infrared imaging is captured during the physiotherapy exercises, feedback can be provided on which movements are working to stimulate brain function.

Radiotherapy - Varian

Researchers: Professor Rick Middleton, Professor Peter Greer and Dr Todsaporn Fuangrod

Collaborating with medical physicist Professor Peter Greer from the Mater Hospital and others, Professor Middleton has jointly supervised a PhD project by Dr Todsaporn Fuangrod, to develop a new system to help identify and eliminate errors in radiotherapy administration.

Although quite rare, errors in radiotherapy dosages can have serious consequences. The objective when delivering radiotherapy is to accurately administer the radiotherapy to the tumour, shaping the collimator beam very precisely to minimize damage to the surrounding tissue. However, errors can occur due to human procedural error, for example when the wrong patient treatment plan is delivered. Or, they can occur due to technical problems, such as an equipment malfunction that results in the collimator not focusing the beam correctly.

Greer, Middleton and Fuangrod have developed a system called WatchDog that takes an image of the radiotherapy beam after it passes through the patient, from which the shape and intensity of the beam can be measured and compared with the patient’s treatment plan to identify any major deviations. WatchDog is a completely separate, add-on system that will not interfere with dose delivery but can provide a warning if it detects something out of the ordinary.

Images: Varian.com

Lead researcher: Professor Andrew Fleming
Associated research group: Precision Mechatronics Laboratory
Partners: Mr Ben Routley, Professor Mark Parsons and Dr Ferdinand Miteff.

Gas embolism and stroke can result from the introduction of a catheter or device during endovascular procedures. These procedures are required to correct restrictions or failure of vessels in the heart and brain.

Researchers from the Precision Mechatronics Laboratory analysed bubble dynamics to predict the severity of gas embolisms and to determine conditions that can lead to dissolution before brain function is affected.

They found that by eliminating Nitrogen from the patient’s breathing gas, the dissolution rate of a bubble can be increased by factor of three. For example, a 1mm bubble can be eliminated in 50 minutes rather than three hours. Furthermore, if the ambient pressure can be increased to three atmospheres, which is equivalent to being 20 meters underwater, the dissolution time is reduced to less 7 minutes. With this reduction, permanent injury may be completely avoided from an otherwise catastrophic stroke.

Though this protocol is yet to be implemented in clinical practice, it could significantly reduce the impact of gas embolism during endovascular procedures.

Lead researcher: Dr Shamus Smith
Partners: Professor Bronwyn Hemsley and Professor Clare Collins

Dr Shamus Smith is working on applying virtual and augmented reality technologies in a range of health-related projects, including:

  • Working with nutritionists (Professor Clare Collins) on image analysis and augmented reality apps to support portion estimation and healthy food choices. Evaluating haptic interaction – developing skills that you need to touch things.
  • Collaborating with speech pathologists (Professor Bronwyn Hemsley) to develop an app that aids communication for people whose speech is impeded. The particular aim of the app is to support these people in communicating with hospital staff and health professionals. The device will allow them to personalize the communications by adding recordings of their own voice.
  • Working with built environment experts to apply virtual reality in the design of a care home environment where you can control a staged evacuation.
  • Collaborating with speech pathologists (Professor Bronwyn Hemsley) to develop a haptic device to analyse food to determine whether it’s of the right consistency for someone with swallowing devices.

Disaster Risk Reduction in Residential Aged Care Facilities: Using the Unreal Engine to model real-world care facilities to support staff training for emergency evacuations. (Stereo views of virtual health-care building)

Disaster Risk Reduction in Residential Aged Care Facilities: Using the Unreal Engine to model real-world care facilities to support staff training for emergency evacuations. (Stereo views of virtual health-care building).

Body Scanning and Segmentation with Gaming Technology: Developing a low-cost, portable system to measure body size and shape using an Xbox Kinect.

Body Scanning and Segmentation with Gaming Technology: Developing a low-cost, portable system to measure body size and shape using an Xbox Kinect.

Haptic Food: Exploring the use of touch-based interaction and haptic devices to support the training of carers in food preparation for meal viscosity and texture.

Haptic Food: Exploring the use of touch-based interaction and haptic devices to support the training of carers in food preparation for meal viscosity and texture.

Hospital Talk: An Augmentative/Alternative Communication app to help hospital patients with communication difficulties to get their message across to staff and carers.

Hospital Talk: An Augmentative/Alternative Communication app to help hospital patients with communication difficulties to get their message across to staff and carers.

Free download on the iTunes App Store

3D electromagnetic localization pill cam

Lead researcher: Professor Andrew Fleming
Associated research group: Precision Mechatronics Laboratory
Partner: Mr Mohd Noor Islam

Researchers in the Precision Mechatronics Laboratory are developing a 3D electromagnetic localization device that can be swallowed by a patient. This method can be combined with endoscope capsules to identify the exact location of lesions in the gastrointestinal tract.

Endoscopy is a medical procedure that allows a doctor to observe the inside of the body. An endoscope is a long flexible tube with a lens at one end and a video camera at the other. Whilst an endoscopy generates a large amount of video footage, it does not identify the precise location of target object. The uncertain location of video footage can complicate the diagnosis and treatment of gastrointestinal diseases.