Leuchs et al., 2017. Neural correlates of pupil dilation during human fear conditioning. NeuroImage.
Motivation
SCR sucks, like big time. Pupil diameter might be a better physiological marker than SCR. Let’s find out together.
Introduction
Background
- Pavlovian fear conditioning and extinction are useful tools for studying emotional learning & PTSD
- SCR is usually the main index of sympathetic arousal, and is usually measured as CS+ vs. CS-
- Can also use startle, but this changes the structure of the task slightly
- Disadvantages of SCR
- don’t have to convince me, I know first hand how noisy and unreliable it can be
- Pros of pupillometry
- neutral measuer (compared to EMG startle)
- easily implemented in fMRI (we’ll see)
- tonic changes linked to general alertness and wakefulness
- Most importantly, the pupil dilates in response to aversive and emotionally arousing stimuli
- so we can use it as a read out of fear/threat related processing
- Not only increases to CS+, hints of contraction to CS-, so could act as readout of fear inhibition
- Since there are no direct cortical inputs to muscles controlling pupil response, probably mediated by the Locus coeruleus (LC)
Quantifying responses
- usually baselined to pre-trial responses
- either take mean during CS presentation,
- or take peak change from baseline
- in this method, people have shown that peak is usually right before US presentation
Goal
Describe the neural correlates of pupil dynamics during conditioning and extinction, with the hypothesis that these regions will overlap with normal fear/extinction network. If true, then pupillometry is indeed a good measuer of physiological reactivity during conditioning experiments.
Methods
- N=34
- simple task, CS were colored squares, and were either CS100+, CS60+, or 0% (CS-) reinforced
- Extinction was between groups for different CS+
- Some details about excluding certain trials for analysis based on some pupil data, not worth copying here but refer back to article for details. In sum no more than 5 trials removed per subeject
- GLM relating pupillometry to neural activity modeled both as single PM regressor and as seperate modulators
Results
- neural correlates of pupil diameter match eactly the fear map. QED son.
- really not much else to say - it was a simple paper with a good point. will be incorporating this into my analyses whenever possible.
- No PREE effect, kinda interesting, but we would never use a CS100+