BRAINnet ~ media releases

HUMAN brain power is being applied in tandem with sophisticated computer modelling and visualisation engines to help scientists understand how the brain works. 

BRAINnet (BDC) at Sydney's Westmead Hospital has been developing an ``integrative neuroscience'' approach designed to explain the workings of the brain as an interactive network. 

Computerised brain imaging technologies that explore the timing of brain function, and where activity occurs, have allowed the multidisciplinary research team to develop new ways of analysing and modelling patterns of brain dysfunction. 

``Essentially we've brought together people from different academic backgrounds -- physics, mathematics, psychology, psychiatry and computer science -- to look at the brain as a system,'' BDC director Dr Evian Gordon says. 

``Physicists have handed us wonderful brain imaging technologies and we've developed software tools to analyse the information in an integrated way.'' 

Most brain research still focuses on the microscopic scale and specialised networks, he says, but there are limits to the extent to which mechanisms operating in a single neuron can be scaled up into useful models of the whole brain. 

But despite understanding of the brain being far from complete, ``sufficient facts already exist for integrative neuroscience to lift us clear of the jungle of detail'', Gordon says. 

He has just published Integrative Neuroscience, the first book to deal with ``the manner in which all the brain's processes are interrelated within and across scale, as well as across disciplines''. It appears some relatively simple rules may underlie overall brain function. 

Excitatory activity is thought to engage widespread networks that are most appropriate to process a stimulus at any point in time, while inhibitory activity serves to damp out networks that are not required. 

A key mechanism thought to constitute learning involves reinforcement of connections between neurons that fire simultaneously -- ``neurons that fire together, wire together''. 

Gordon says the brain seems to operate mainly in two modes -- the neuronal axons either fire or they do not. At the network level there are bursts of firing, while at the whole brain level there is either synchronising or desynchronising activity, and relatively automatic or controlled processing. 

``The ongoing balance between these two modes of function across scale determines the brain's overall stability,'' Gordon says. 

``At marginal stability, an optimal balance occurs between linear synchronised and non-linear desynchronised behaviour, known as the edge of chaos. 

``This edge of chaos may provide the best way to process a lot of information while retaining an optimal balance between flexibility and stability.'' 

Gordon says a more holistic picture is emerging of the flow of information processing in the brain. 

``Professor Peter Robinson and Chris Rennie of Sydney University's physics school and Professor Jim Wright of Auckland have developed a numerical simulation that for the first time bridges the gap between physics and the brain as a biological system,'' he says. 

``This computer model allows us to explore and test specific aspects of overall brain dynamics.'' BDC researchers are collaborating on studies of conditions, including attention deficit hyperactivity disorder (ADHD), post-traumatic stress disorder, social phobias, schizophrenia, borderline personality disorder, Parkinsons disease and traumatic brain injury. 

In the past, each study was done separately, with researchers looking at the stability of the brain's electrical processing, how the brain processes target information, brain arousal levels and so on, he says. 

``We've got a database of patients and controls from these studies, and we're now able to explore and re-examine the information in an integrated way,'' he says. 

``We're starting to get beyond the simplistic averages to see what it means at an individual level, and interrogate data against different profiles and treatment measures.'' 

The challenge is for the researchers to become increasingly focused in the kinds of questions they are asking. 

It's a form of data mining, with the human intelligence looking for the patterns that emerge, and the convergences that give clues to what's happening. 

``We're letting the computers do what they do best, which is detailed processing, and we're finding new ways of doing what the brain does best, which is spatial pattern finding,'' he says. 

``We have some heavyweight computer scientists who are good at visualisation. So we have the computers giving the data back to us in a spatial way. Ironically, you have the human intelligence and the artificial intelligence in a loop -- it has been a lot of fun.'' 

C++ programming has allowed for new proficiencies in rapidly analysing very large amounts of brain function data. Although it's still early days, Gordon says the model will be a very sophisticated engine within a year. 

``We're trying to capture the right tension axis by letting computers do the detailed, serial processing, but sending out the crucial data in a way that enables us to spot new patterns,'' he says. 

Scientists can also use the technique to test competing theories on the same data. 

``That's really important, because it helps flesh out not just which ones really work, but the relative merits of different projects and approaches,'' Gordon says. 

``Every scientist thinks what he or she is doing is the epicentre. 

``When you dispassionately go through all the options in the same data on the same people with respect to a normative database and see which ones are adding more value -- that's an acid test.'' 

Converging on answers 

RESEARCH into ADHD is shedding new light on brain functioning, as well as insight into the disorder itself. 

``In ADHD all the pieces are there, we just need to bring them together,'' BDC director Dr Evian Gordon told a conference, ADHD in the Third Millennium, at Westmead in Sydney last month. 

``The central question in ADHD in our view is, whether these kids have abnormal theta (slow wave) activity -- the most common finding in previous research. 

``The entire point of integrative neuroscience is to add further value to that question, so in the case of simultaneous measures of the body we want to know if arousal affects this. 

``In the case of numerical simulation we want to know if we can get a better explanation of what the mechanism is, and in the case of treatment we want to see how treatment is affecting the measures and our explanations.'' 

Collaboration is the key, bringing in other people who can add ``further explanatory power to questions that are usually only addressed from a very narrow bandwidth point of view''. 

``This convergence is bringing power to central questions that had previously only been answered in a superficial way.'' 

Caption: Integration:Evian Gordon favours a multidisciplinary approach to the brainPicture: Andrew Baker Multi-talented:A painting by Dr Gordon's `Brain Art' web gallery In the picture:A self portrait.

Back to Media Releases