Systems Neuroscience Group

Michael Breakspear

Group Leader

Computational psychiatry and translational neuroimaging

I am the group leader of the Systems Neuroscience Group with interests in computational neuroscience and translational neuroimaging. My contributions to the former focus on dynamic models of large-scale brain activity, toolbox development and the detection of nonlinear dynamics in empirical data. My work in translational imaging encompasses healthy ageing, dementia, bipolar disorder and schizophrenia, with a focus on connectomics and risk prediction.

I grew up in Sydney and studied medicine, philosophy and mathematics. I undertook early career research training in the School of Physics at the University of Sydney before moving to the School of Psychiatry at UNSW as a mid-career researcher.

I formed the Systems Neuroscience Group at UNSW in Sydney in 2004, then moved to QIMR Berghofer Medical Research Institute from 2009. I relocated to Newcastle in 2019 and established the Systems Neuroscience Group, Newcastle (SNG-Newy) with aspirations to integrate basic methods, bioinformatics and clinical translation with a unique regional Australian character. Our imaging centre is in a beautiful bushland setting on Awabakal country.

In addition to basic research training, I also completed training in psychiatry and nowadays combine my research career with clinical sessions in adult psychiatry. I have an interest in recovery-focussed treatment of mood disorders, psychosis, and addiction. In the past I have worked in Prison Mental Health and Inner City community psychiatry.

I have a passion for climate science, being rather social, and surfing.

Johan van der Meer


Real-time approaches to monitor or change brain signals

Research Interests

Our brains produces electrical patterns and rhythms that occur during walking, concentration in solving a difficult arithmatical question, or noticing a car that is approaching too fast. How to relate these signals to the different types of behaviors is usually done with experiments where the important analyses are done after all data has been acquired. These types of analyses can then discover assocations. However, causal relationships are harder - these require a more dynamic approach where you change the conditions of the experiment based on the current state of the brain signal.

Coupled with this thought is Neurofeedback training, an often used (and touted) approach whereby a (prolonged) training regimen of particular aspects of your brain signal is hypothesized to result in a concomitant improvement in some behavior. No-one really knows how this works, though, and the one theory of how this is supposed to work - neurons that fire together wire together (Hebbian Learning) does not seem to hold up when delays in the signal processing stream in NF training might prevent the ‘fire together’ part of this equation.

Finally, I am interested in deployment of these complicated real-time experimental and analysis approaches outside of the confines of the Lab, with their specialized (and expensive) machinery. I’d like to make an effort to remove some of these barriers and make it more easy to perform experimental acquisitions anywhere. Not by removing necessary complexity (i.e. make a one-push-button approach), but by trying to explain all necessary components as clearly as possible.

Léonie Borne


Machine Learning & Neuroimaging

Having always been passionate about artificial intelligence and medical imaging, my research work has naturally focused on the development of “intelligent” tools for the study of an “intelligent” organ: the brain.

Deep learning for the study of cortical folds

After a master’s degree in artificial intelligence, I did my PhD with Prof. Jean-François Mangin at Neurospin in France. My PhD work focused on the development of several structural MRI analysis pipelines for the study of cortical folds using machine learning algorithms, including deep learning. The tools I developed are now available in the BrainVISA toolbox (

Brain-behaviour association in dementia

Following my PhD, I started a postdoc in 2020 with Prof. Michael Breakspear in Australia, where I applied my skills to address the challenges of dementia. Using advanced statistical analysis, I have notably identified a robust association between brain atrophy and cognitive decline during the preclinical stage of Alzheimer’s disease. My work is based on the databases ADNI, AIBL and PISA.

Renate Thienel

Research Associate & Research Manager

Schizophrenia, Cognitive Neuroscience


I am a Mid-Career Researcher with a passion for improving health outcomes; I hold a PhD (Science), Postgraduate Diploma in Psychology, and a BSc (Hons);
My research activities span across clinical, cognitive and neuroscientific aspects of health and mental health;
My research focuses on work with clinical and non-clinical participants and spans across methodologies such as self-report surveys, clinical and cognitive assessments, magnetic resonance imaging, pharmacological agents, transcranial direct current stimulation, electroencephalography etc.;
I have published 56 HERDC research publications including 30 (C1) journal articles, 5 book chapters (1 B1, 1 D1), and 25 conference papers, 1 white paper and several governmental reports reflecting my academic efforts and contributions, with a h-index of 17, and 1111 citations;
I have over 10 years of experience working as a mentor and supervisor for domestic and international students including HDR students;
My career grant income is $520,853 (excluding $1,500,000 from AI positions) across category 1, category 3 and internal grants including a prestigious 4 year UoN post-doctoral fellowship;
Since 2019 I am the Research Manager of Michael Breakspear’s Systems Neuroscience Group at HMRI’s Imaging Centre where I am coordinating and am involved in the groups various research studies which includes the Newcastle site for the Australian Dementia Network (ADNeT).

Megan Campbell


Emotional inference; perception & action; fMRI

Research Interests

My previous work focused on the overlapping sensorimotor processes underlying interpersonal interactions, within healthy adult cohorts. By using Bayesian models to explain behavioural and neural datasets I described mechanisms for the context-dependent modulation of automatic imitation responses, a feature of how humans prepare responses during interactions with others. Now my focus has shifted to applying this experience in functional neuroimaging, behavioural paradigms and computational approaches to social-emotional cognition and extending this to clinical populations. In particular, I aim to model dynamic brain networks involved in emotion in order to understand, differentiate and improve diagnosis of mood disorders. ORCID


I recently took up my first postdoctoral position here in the Systems Neuroscience Group, Newcastle. This is a continuation of my slow migration south, having grown up in Far North Queensland where I completed a BPsychHons at JCU Cairns, before moving to Brisbane. There I enrolled in a Masters of Neuroscience at the Queensland Brain Institute fell head-over-heals for the field of cognitive neuroscience, and continued on to obtain my PhD in 2019 at UQ.

Nikitas Koussis

PhD Student

Active inference in cognitive deficits in psychosis

Research Interests

I am very interested in the cognitive symptoms of psychosis and more generally in the functioning and structure of the human brain in pathology and mental illness. My work focuses on approaching cognitive dysfunction in psychosis using a Bayesian approach. I hope to describe mechanisms of gain and attention within an interference framework. I utilise functional, diffusion and structural neuroimaging, as well as behavioural and cognitive paradigms, and computational approaches to cognition.


Upon completing a Bachelor of Science at Queensland University of Technology in 2017, I worked at QIMR Berghofer Medical Research Institute as assistant to PI on the Prospective Imaging Study of Ageing; an AQIP grant and several other projects. I am currently completing a PhD at University of Newcastle under Michael Breakspear. In my spare time I write fantasy books and seem to be perpetually fixing up my house.

Jayson Jeganathan

PhD student

Brain connectivity, Emotional inference, Psychosis


I completed his undergraduate studies and MBBS (Hons Class I) at the University of Sydney. Since 2018, he has been an RANZCP accredited psychiatry trainee. Concurrently, his research interests have included network analysis and neuroimaging in bipolar disorder. Currently, I am a PhD student in the lab, investigating emotional inference in psychosis.

Decoding the brain basis of emotion

Is emotion really as simple as smiling when you’re happy, and frowning when you’re angry? I am investigating how the mind combines multimodal information, including seeing, hearing and feeling the outside world, the inner sensation of organs such as the heart and gut, and our previous experiences, to construct the moment-to-moment emotions that we feel. I am using state-of-the-art brain imaging, emotional analysis, and physiological measurements of our inner organs, to uncover the brain basis of emotion.

Affective symptoms of early psychosis

In psychosis, impairments in motivation, enjoyment, and emotional facial expressiveness are the most debilitating symptoms for patients, yet we have no effective treatments for them. I aim to integrate recent advances in neuroimaging and machine learning to find out how the brain basis of emotion goes awry in early psychosis

Bryan Paton

Data Scientist & Technical Support (Wizard)

Consciousness, Attention, Learning under uncertainty and computational models of this

Research Interests

I have a diverse background in computer engineering, philosophy and cognitive neuroscience. I am a nerd at heart and having this broad experience lets me combine all of my interests.

Consciousness & Attention

I am deeply interested in the brain and how perceptual experiences transition from non-conscious to conscious and what are the factors, cognitive, behavioural and neurophysiological involved. I have some experience with binocular rivalry, continuous flash suppression and related paradigms.

Learning under conditions of uncertainty or volatility

I am also interested in learning, in general, especially models of the process where there are sources of noise and/or uncertainty and when these themselves change over time e.g. volatility. I have also been trying to develop novel paradigms to study these processes in a dyadic, social setting.

Computational models

Finally I have in interest in how the brain works, and the methods used to study it, having a long background in EEG, MRI and many other related technologies. By combining these methods of data collection and using computational models of the behavioural and neurophysiological mechanisms involved I hope to gain more insight in these phenomena in both the healthy brain and its pathophysiology including Schizophrenia, Autism Spectrum Disorder and Dementia.