Our lab pushes the boundaries of the neurobiology of fear and reward expectation, as well as memory formation by integrating theoretical stipulations of learning with neural techniques that prod brain function in real time and examine how alterations in brain function influence information processing.miHAELA IORDANOVA, principal investigator
Our work focuses on studying how the brain learns, how it updates prior knowledge with new information, and how it can make novel inferences about the world on the basis of current knowledge. We do this by studying the contribution of specific neuronal ensembles to learning, the recruitment of specific neuronal circuits as well as the real-time neural processing of environmental input.
Techniques used in the lab
In-vivo neuronal recording
How does the brain predict the future?
Making accurate predictions allows us to behave efficiently and adaptively. Learning to predict the world depends on prediction error or the discrepancy between real and expected events. The greater the error the greater the learning. Our research studies how this teaching signal is implemented in the brain and whether this signal is common between disparate motivational states.
How does the brain update predictions and modify behaviour?
The ability to update predictions and in turn modify behaviour allows us to reduce or even eliminate unwanted or maladaptive actions. To understand how the brain updates past learning with new information, we study two types of extinction learning: extinction driven by outcome omission and extinction driven by overexpectation (Figure 1). Both designs result in reduction in an unwanted behavioural response but they do so differently. Extinction by omission removes the outcome that normally drives responding, whereas extinction by overexpectation maintains outcome delivery but generates an inflation of the expectation of that outcome. In both cases a negative prediction error is generated which drives updating and behavioural change. Using this paradigm-independent approach we are able to examine the generality of brain areas in extinction learning and shed light on the behavioural processes that specific brain areas support.
How does the brain make novel inferences?
This work focuses on understanding how emotional memories can become linked to memories from the past or generalized to newly acquired memories, and thus imbue normally neutral memories with emotional valance. We model this using sensory preconditioning and second-order conditioning. In sensory preconditioning two neutral cues are paired (A-B) prior to subsequent conditioning of one of those cues (A, but not B) with an aversive event (footshock, A-footshock). This emotional learning propagates to cue B imbuing it with emotional valance. In second-order conditioning the emotional learning (A-footshock) takes place prior to pairings of the two cues (A-B). Like in sensory preconditioning, cue B is also imbued with emotional valance. We use these designs to understand the role of specific neuronal ensembles as well as the critical neural circuits that regulate fear inference.