New computational neuroscience methodology proves long-held theory about dopamine modulation during decision-making processes

"The methodology takes techniques from functional analysis, machine learning and dynamical systems and applies them to areas such as neuroscience," explains Dr Balaguer-Ballester. "By mixing up approaches from different subject areas, the theory has allowed researchers in the US and Canada to prove a theoretical hypothesis proposed by neuro-computational models over fifteen years ago."

"The methodology allows for the reliable identification of the neural dynamics from recordings gathered from monitoring animal behavior," said Dr Balaguer-Ballester, "For example, a neuroscience study looking at decision making in rats might be able to gather data on simultaneous brain activity in about 100-1000 neurons.

"This provides only a small part of the picture, relative to the number of neurons involved in the decision making process, but our methodology enables the identification of an expanded state space where patterns of neural activity associated with "metastable" states of the network, are robustly represented.

"The states identified are cross-validated through multiple trials of a particular behavioral task, which ensures their reliability. This is a particularly useful novel approach, because the enhancement method makes it easier to spot wider trends and dynamic patterns in relation to behaviors which are not accessible by classical approaches."

The method was originally published in PLoS Computational Biology in 2011, but until this year, had not been used in practice. A joint project by universities in the UK, America, Canada and Germany used this method of data analysis to demonstrate the effects of amphetamines in rats. The study was published in the Journal of Neuroscience in July.

Thanks to the enhancement of the recorded data provided by this method, researchers were able to demonstrate that after receiving a low dose of amphetamine; stable states associated with cognitive decisions became more stable, while the rats consistently improved their performance in making correct decisions in a maze task. At higher doses, the higher dopamine concentration in frontal areas caused the disruption of such stable states rendering the rats unable to make the correct decisions.

Speaking about the next steps of his research, Dr Balaguer-Ballester explained, "I intend to use what we have learned through this study to continue to improve the original methodology.

In time, I hope that it will be able to prove how metastable neural states naturally emerge and evolve, which will help to establish a causal link between behaviors and the decision making process."