Making novel choices – new insights from monkeys into how the brain makes decisions

By Helen Collins

The ability to make informed decisions in unfamiliar environments is a necessary human skill, without which we wouldn’t be able to make appropriate choices when navigating the world. We make these kinds of decisions every day, whether that be what to choose from a menu, or where to book our next holiday. In these situations, we use our previous knowledge and experiences to adapt our decisions to the novel setting, in order to make the best choices and maximise our success.

Our understanding of adaptive decision-making has grown enormously in recent years, as advances in imaging techniques have allowed us to visualise the brain in real-time. In particular, functional magnetic resonance imaging (fMRI) has been instrumental in identifying active brain regions during such tasks. Numerous imaging studies have found that the medial frontal cortex (MFC), a brain region located just behind the eyes, is activated when a person makes a novel decision. Specifically, this region is involved in value-guided decision tasks, meaning tasks where the values of two options are compared in order to choose the one that gives the greatest reward.  

Unfortunately for behavioural psychologists, value-guided tasks are much more challenging to study in animal models. For the sake of good experimental design, the decisions being made have to be simplified so that there are as few variables as possible. For that reason, when working with monkeys, they are often made to choose between abstract shapes and colours. These abstract patterns represent rewards of different value, yet, as a consequence of being so symbolic, it may take many thousands of repetitions for a monkey to learn the association between its choices and the rewards. Thus, by the time the animal gets tested, all the choices are highly familiar to them. These tasks, therefore, poorly reflect everyday scenarios where humans routinely have to make choices between new, spontaneously-occurring options.

New research from Prof. Matthew Rushworth’s group, from the Department of Experimental Psychology in Oxford, took a different approach when studying decision-making to overcome these difficulties. In a recent paper published in Nature, the authors trained four rhesus macaques (Macaca mulatta), a species commonly used in neuroscience research, to recognise different combinations of coloured dots. In this task, the colour of the dot represented the number of drops of juice the monkey could receive, and the number of dots represented the probability of being given the juice. The macaques would then be shown two different combinations of dots and pick which one they wanted. This meant they had to integrate the information and choose the option that would give them the greatest probability of the biggest reward. 

The most innovative part of this study, however, was the stimuli used when the macaques were being tested. Throughout the training, monkeys were only shown a subset of colour-number combinations, so when it came to testing, the experimenters could show them combinations they had not seen before. This experiment more closely resembled novel human decision-making but still meant the macaques relied on their previous training experiences. 

When the macaques performed this task in the scanner, the MFC lit up, just as it does when humans perform similar tasks. This is a significant finding, as previously a different region, the orbitofrontal cortex, was shown to be active in macaques when they chose between familiar options. Because earlier studies hadn’t accounted for the difference between new and familiar choices, researchers had been confused about why humans and monkeys seemed to be using different parts of their brains. For the first time, this study has shown how macaques can use the MFC, just like humans, when they make decisions, and why there was a mismatch between previous animal and human studies.

Experimental design. The colour-number combinations show all the possible options for stimuli. Colour represents the number of drops of juice reward the macaques receive; the number of dots represents the probability of receiving the reward.
Image by Thomas Leslie

The authors then went a step further to demonstrate the causal role the MFC plays in novel adaptive decisions. Transcranial ultrasound stimulation (TUS) is a technique developed in Oxford by Jérôme Sallet that can non-invasively and transiently disrupt neural activity when applied over the brain. When TUS was applied to the MFC in macaques, they were less able to understand the novel dot combinations and thus made less-rewarding decisions. This implies that the MFC is critical for making novel decisions.  

This wasn’t the authors’ only amazing discovery. The researchers think that macaques make these value-guided decisions in a similar way to how they understand physical space. Nobel Prize-winning research by John O’Keefe, May-Britt, and Edvard I Moser previously established that physical space is represented in the brain by grid cells that encode the spatial location of objects. It seems that macaques also understand the landscape of potential dot-combinations similarly, as a sort of abstract space of possible options. Think of this as a simple graph, where the x-axis represents the colour, from least rewarding to most rewarding, and the y-axis represents the number of dots, or the probability of receiving the reward. The top right of such a graph would therefore represent the greatest chance of the best reward. This amazing discovery demonstrates how the brain can apply the same processes to multiple different functions. As Prof. Rushworth, newly-appointed Fellow of the Royal Society, says, ‘It is intriguing to think that mechanisms for finding one’s way in physical space also underlie our ability to navigating an abstract space of choice possibilities’.

This new research represents a major step forward in how psychologists study decision-making. It shows that macaques are much more capable of making novel decisions than previously thought, and that they use similar brain circuitry to humans. These findings open the door to studying more complicated behaviours in monkeys in the future, as well as their underlying neural mechanisms.

Featured image by Fathromi Ramdlon on Pixabay

1.     Bongioanni, A. et al. Activation and disruption of a neural mechanism for novel choice in monkeys. Nature 591, 270–274 (2021).

2.     Bongioanni, A. How monkeys and humans make novel choices. https://www.ox.ac.uk/news/science-blog/how-monkeys-and-humans-make-novel-choices.

Leave a Reply