Model of the basal amygdala - computational Pavlovian fear conditioning

After the first week of the CAMP course we had to decide which project we intended to do next week. I was always interested in how fear and anxiety evolve in the brain so at the end I decided taking a joint project with my course fellow Cristiano Köhler about modeling the basal amygdala - a part of amygdala which is involved in the formation of fear and extinction memories. The main task was to reproduce a model of the basal amygdala and investigate its behavior. Our helpful supervisor was Arvind Kumar, one of the authors of the paper [1] that describes the model in the first place.

Objectives

At first, we reproduced the model of Vlachos et al., 2011 of the basolateral amygdala. The task of the model is to be able to remember a conditional stimulus (CS, e.g. flute tone) paired to an unconditional stimulus (US, e.g. electric shock) - so basically after applying Pavlovian fear conditioning on the model, it takes the same conditional stimulus (without the shock) as a cue of fear. However, further presentations of CS alone results in a decline of this conditioned response - that process is called fear extinction. Another phenomenon is called fear renewal, which shows the context-dependency of fear conditioning and extinction: after the fear conditioning and extinction period in context A (CTXA) and context B (CTXB) respectively, a repeated CS presentation in context A immediately brings back the fear memories despite the effect of the extinction period.

Model

The model intends to reproduce the functions of the basal amygdala. It's a large-scale spiking neuron network implemented in Python, utilizing the Brian library. The model consists of 4000 leaky integrate-and-fire neurons - 3600 so called excitatory and 400 inhibitory neurons. Three kind of inputs are given to the system as Poisson spike trains: (1) CS-US pairs, (2) context information to a subset of neurons and (3) background noise to all the neurons. Plasticity is established in the connections between the inputs (CS-US, context information) and the excitatory neurons. Further plasticity is introduced in the synapses from inhibitory to excitatory neurons to investigate the effects it may bring about.

Distribution of inputs in the spiking neural network model. CS-US are provided to all neurons while CTX input is fed only to a subpopulations of excitatory neurons - from [1].

Separate subgroups (20-20%) of excitatory neurons received CTXA and CTXB inputs, while both populations received the CS-US inputs. The difference between CS and CS-US presentation can be only found in the context in which the stimuli are given; so CS and CS-US stimuli are the same in the subgroup of neurons they target as well as in the intensity they influence the system. The plasticity rule is implemented so if the pre-spikes of context and CS-US inputs are close in proximity the connections are strengthened (LTP) between the corresponding context and excitatory neurons, otherwise they weakened (LTD) - similar to the BCM rule.

Fear conditioning and extinction

During fear conditioning 5 CS-US presentations are given with the duration of 50ms each separated by 100ms breaks in CTXA. After that, presenting 'only the CS' in CTXB results in the firing of 'fear neurons' - neurons receive input from CTXA. Though 5 CS presentation in CTXB evokes firing of the 'extinction neurons' which by inhibiting the fear neurons reduce their firing rate; so on the behavioral level the fear response to CS is ceased.

Results of our model. All fear conditioning, extinction and renewal are present on the plots. On the bottom-left plot, fear neurons have a number of 2000 - 2800, excitatory neurons 300 - 1100. Fear conditioning is done for the first 750ms, fear extinction from 850ms to 1600ms. Fear renewal can be observed at around 1700ms.

Fear renewal

After extinction training, a single presentation of CS in CTXA results in the massive firing of fear neurons again. This supports the hypothesis that fear extinction is not unlearning of the information stored during fear conditioning.

Plasticity plots: weight values produced by our model. Fear and extinction here means the connection between the inputs, and fear and extinction neurons, respectively.

Extra minigame

I implemented a quite simple game in Pygame that demonstrates the previously described phenomena. By pressing the C and U keys one can present CS or US stimulus, respectively. If pressed together (immediately after each other) a CS-US pair is presented to the mouse. It can show its response to the stimuli in three ways: scared, neutral, happy. Basically happy and neutral means no reaction - the firing rate of fear neurons does not high enough to trigger a fear response. Behind the game our actual model works with just 800 neurons so the simulation does not slow down the game that much.

The model and the game can be freely downloaded from our git repository. Have fun!

[1] I. Vlachos, C. Herry, A. Lüthi, A. Aertsen, and A. Kumar, “Context-Dependent Encoding of Fear and Extinction Memories in a Large-Scale Network Model of the Basal Amygdala,” PLoS Comput Biol, vol. 7, no. 3, p. e1001104, Mar. 2011.