The curious brain: behavioural and neural mechanisms of spatial updating and reward surprise
<p>In an ever-changing environment, organisms learn about non-random events and use that information to form predictive models to facilitate the processing of expected stimuli or planned actions and their outcomes. Predictions about stimuli enable for an efficient and appropriate response and...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Published: |
2016
|
| Summary: | <p>In an ever-changing environment, organisms learn about non-random events and use that information to form predictive models to facilitate the processing of expected stimuli or planned actions and their outcomes. Predictions about stimuli enable for an efficient and appropriate response and predictions about action outcomes determine the value and therefore likelihood of repetition of that action. In primate brains, these models have been described in the framework of spatial attention and reward expectation. Unexpected observations in space trigger an attentional reorienting process and surprising action outcomes are defined as reward prediction errors. These error signals can be used by the brain to update and refine its internal model that contains probabilistic information about events in the environment (Chapter 1). In what follows, I will report effects of spatial surprise and unexpected rewards on behaviour and brain activity in humans and macaques. I will show how humans react to visually surprising events and how they use that information to form future predictions (Chapter 2). Brains reacting to surprise have been primarily described in the attention literature but there are dissociable networks for an attentional shift and for an expectation updating process following unexpected but informative observations (Chapter 3). Macaques also form spatial predictions and violations of these predictions result in a behavioural cost. Furthermore, both expectations about the size of a reward and prediction errors influence behaviour (Chapter 4). FMRI in macaques performing a task reveals a network of reward sensitive areas. Nodes of this network however carry different information about the reward history and therefore play different roles in forming a reward expectation. Spatially surprising events activate networks similar to those described in human fMRI studies (Chapter 5). In summary, unexpected observations and action outcomes influence behaviour and activate brain networks that process the surprise as well as form new predictions.</p> |
|---|