The mission of our lab is to identify ways to improve recovery after brain injury, through an understanding of the neuroscience of resilience to injury and restoration of function. To achieve this, we investigate networks in the brain that support functions such as memory, using cutting-edge techniques including diffusion MRI, tractography and tractometry, positron emission tomography (PET), EEG and cognitive paradigms.

We study clinical populations with stroke, mild traumatic brain injury (“concussion”) and hypoxic-ischaemic injury. Through investigation of cross-cutting areas of cognitive neuroscience we aim to develop a detailed understanding of deficits and the potential for restoration of function.

  • Professor Michael O'Sullivan

    UQ Centre for Clinical Research
    Affiliate Professor
    Institute for Molecular Bioscience
  • Dr Lena Oestreich

    Postdoctoral Research Fellow
    UQ Centre for Clinical Research
  • Dr Delphine Levy-Bencheton

    Postdoctoral Research Fellow
    UQ Centre for Clinical Research
    Centre for Restorative Neuroscience, Queensland Brain Institute
    Affiliate Research Fellow
    Queensland Brain Institute
  • Ms Elizabeth Arnold

    Research Nurse & Research Nurse
    UQ Centre for Clinical Research
    Clinical Operations Consultant
    Queensland Brain Institute


Approximately half a million Australians live with the effects of previous stroke. Even with recent breakthroughs in the treatment of acute stroke, over 40% of patients end up with at least a moderate degree of disability. Problems with thinking and mental skills (cognition) are common and an area of unmet need.

Investigation of memory following stroke is currently an active area of research in the O’Sullivan lab. This is made possible by the STRATEGIC cohort – a group of approximately 200 patients recruited early after a first mild to moderate stroke and followed for one year.

Brain Networks for Memory

FornizThe hippocampus, a 3-4 cubic centimetre structure on the inner surface of each hemisphere of the brain, has been at the centre of memory research for several decades. However, the hippocampus is part of a wider network, which we now refer to as the extended hippocampal system. We have used advanced imaging methods, especially diffusion MRI and tractography, to reconstruct connections in this system and determine how they are involved in ageing or disease. For example, the fornix – a bidirectional tract that connects hippocampus to the diencephalon, basal forebrain and frontal lobe – is the main correlate of memory performance in both young adult and older adult healthy volunteers. The fornix is also damaged early in the course of Alzheimer’s disease. (The fornix is shown in red in the animation below; the other tracts are the uncinate fasciculus, blue, and the parahippocampal cingulum, green.)

In the STRATEGIC cohort (>200 participants recruited in South London, UK), we are investigating memory disturbance and recovery after stroke. Understanding of the main predictors of prognosis is likely to shed light on potential mechanisms of recovery and decline. For example, if structural features remote from the infarct predict outcome, that might suggest that the coexistence of other disease processes, for example early Alzheimer’s disease, is important.

We have found that the location of injury is an important predictor of memory function and that this is due to disconnection of memory networks through damage to white matter projections (see below).

White matter projections
Five patients with infarcts in the territory of the left posterior cerebral artery. The area of damage (infarction) is shown in red and the parahippocampal white matter projections in green. Memory deficits depend on the extent to which damage encroaches upon these connections (the overlap in yellow).

Key team member:

  • Dr Paul Wright

Key collaborators:

Higher-order Cognition

Higher-order cognitive processes include concepts such as attention, working memory, decision-making and cognitive control. These processes are easily disrupted by neurological disease, notably cerebral small vessel disease, which is also a cause of stroke. Our lab is interested in the networks that support these functions in the normal brain and, in turn, damage to these networks in disease.

In previous work, we have shown that executive dysfunction in cerebral small vessel disease is due to damage to certain white matter tracts. More recently, we have shown that variations in subsets of connections in the cingulum bundle account for variations in cognitive control that occur in normal ageing.

A new area of interest – made possible by a collaboration with the Mattingley lab at QBI - is the network architecture of working memory and how it overlaps and differs from networks involved in other higher order functions (see below). A current project will examine frontal lobe connections and structure-function relationships in visual working memory (see task below).

Working memory task

Key team members

Key collaborator

Motivation, Mood and Behaviour

A number of higher-order mechanisms control human behavior and marshal our cognitive resources in the face of complex tasks. Motivation refers to how behaviour is driven by certain desired outcomes and cognitive control describes the process of allocating cognitive resources to tasks. These processes can be damaged by disease. In fact, after stroke and brain injury they are a major cause of morbidity and unmet need.

We are investigating mood, motivation and cognitive control after stroke. In particular, we are interested in understanding the networks for these processes and how they are damaged – and also the role of certain pathological features such as inflammation.

Key team members

Key collaborators  

  • Prof Hugh Markus (Cambridge, UK)
  • Prof Masud Husain (Oxford, UK)

Neuroscience of Resilience and Restoration of Function

There is no simple linear relationship between brain injury and loss of function. Certain features of biological networks can make them resilient to disruption. These observations lead to the idea that there may be features of the brain’s organization that make it resilient to injury. Furthermore, there are likely to be a variety of mechanisms involved in neural adaptation to injury. In a previous study we showed that the cholinergic system may have a role in neural adaptation by supporting the use of undamaged white matter connections. We are currently developing ideas to pursue this area of research. A deeper understanding of natural mechanisms of restoration of function – restorative neuroscience – could lead to major advances in therapy for neurological disorders.

Our previous paper on this topic can be found here, and an editorial commentary in the Journal of Neuroscience here.


STRATEGIC is a longitudinal study of cognitive function (mental skills) after stroke. We recruited just over 200 people who had suffered an acute stroke and assessed them after 1-3 months and again after one year. STRATEGIC looked in most detail at memory function after stroke. A previous large survey of stroke survivors had shown that memory problems are a leading cause of unmet need – memory problems are common and we have little to offer currently to treat them. We recruited individuals who had suffered their first stroke, in a variety of different locations in the brain. Understanding the mechanisms of memory problems is the first step in the design of new treatments.

Apathy after Stroke (ApathyAustralia)

Apathy is a major determinant of quality of life after stroke, for which there are currently no effective treatments. We host the Australian arm of an international clinical study of apathy after stroke. We will recruit patients with stroke from the Royal Brisbane and Women’s Hospital and the Sunshine Coast University Hospital. The aim is to get a better understanding of the frequency and persistence of apathy in different subtypes of stroke, to investigate the patterns of brain injury that underpin apathy, and to use this knowledge to design clinical trials.