Lecture 2: Neural Correlates Flashcards
Intro to neuro correlates
The neural correlates of face recognition too can tell us a lot about how we recognise faces and more particularly provide insight into which models are best
Experiments that support Bruce and Young
Schweinberger et al., 2002- priming effect
Neural correlates priming effect study what did they do?
Schweinberger et al., 2002- priming effect
• The target in the experiment was the same pic (Paul McCartney) difference is that the prime is different. Change in neural processing of stimulus depending on what was shown before.
• Showed one of three conditions:
o Unprimed (priming stimulus is someone completely different)
o Primed same (exactly same image both times)
o Primed different (same person but different picture of the same person)
o So what does this do? Can measure the ERP to the same stimulus post priming
Neural correlates of priming effect study what were results?
Schweinberger et al., 2002- priming effect
• Results: priming effect in the N250 priming effect occurs because it is easier to recognise the
• target when they have been presented before. It is easier for the brain to recognise the face so less negative N250 when seeing primed faces.
o See even priming effect for same face but different photo because what happens theoretically is when you see first face a representation of that face is activated if you then see the same person that representation is already activated so it is easier to recognise.
• Because FRU’s recognise people from any image it should also work here ie. paul mccartney representation should be same irrespective of photo.
• Shows an image independent face recognition effect kind of represents what we had in the Bruce and Young model where we say that FRU’s do not depend on image of person.
Image independent neurons study
Quiroga et al., 2005- Jennifer Aniston neuron study
Epilepsy patient has electrodes in brains- person presented with hundreds of pics and found by chance that electrode in neuron in hippocampus always increased firing to any pic of Jennifer Aniston irrespective of what image (image independent). This neuron shows image independent recognition and this is what we want because is a neural correlate of predictions.
Knowledge that there exist in hippocampus that there are image invariant facial neurons.
What is the issue with the Quigora et al., 2005 study and how can one overcome this
The problem with this is that can not use it hugely as is a patient and cant record this type of thing often as is a patient. thus can look using fMRI
what is FRM-adaptation
FRM-adaptation The population of neurons that code a specific attribute when you present that attribute over and over the brain region involved will decrease its firing. If you have a brain region that is able to recognise a face from any image that brain region will find it interesting to begin with but after multiple presentations of different photos of the same person the brain region will adapt.
What would Haxby model predict?
- Haxby et al. (2000): Differences between familiar and unfamiliar faces should be observed in the OFA-FFA pathway (processing of invariant aspects).
- Release from adaptation in FFA (and OFA) and STS (from haxby model we predict that area that should code invariant info and identity should be FFA) as large for different images of the same familiar face as for different faces!
Evidence that FFA is sensitive to configural properties of faces.
- Brain lesions resulting in acquired prosopagnosia (i.e., the inability to identify familiar faces) affect fusiform gyrus (including FFA).
Issue with idea that FFA is sensitive to configural properties of faces.
BUT: fMRI studies produced variable results concerning role of FFA for face familiarity (Gobbini & Haxby, 2007).
Whose study went against haxby model?
Davies-Thompson et al. 2009
What is Davies-Thompson experiment an example of?
Example of an adaptation experiment with fMRI (e.g. negative contrast to black and white and see normal colours). The overall idea is that if a brain region codes info and if you present same info over again the brain will decrease activity to this.
Method of Davies-Thompson experiment
• Use of FMR-adaptation.
Experiment with three conditions:
- Celebrity faces and always the same picture- so a brain region that would be interested in processing the familiarity in these should first go up but then adapt and go down.
- Different pics of different people- a brain region interested in identity processing should go up and stay up as it is changing.
- Same person but a different picture.- so the question is do we get an adaptation effect to these pics because is the same person and boring (would expect this) as if the area is interested in identity and not the specific picture should reduce its activity even with different pics of identity.
According to Haxby model what would we expect from the Davies Thompson experiment?
• If there is a brain region that is interested in identity and not different images then brain region should adapt to the second condition. Adaptation to picture differences and person differences should be different.
Results of Davies-Thompson experimetn
- activity of FF for condition 3 it looks like condition 2. So FFA doesn’t code for identity it responds the same to all different images whether these are the same or different people. So the FFA doesn’t care about person identity
- FFA’s response dominated by differences between images, not between identities. As in the second condition as the FFA does not adapt there is no difference to different images of same person and different images of different people
- Different images of celebrities clearly recognised as the same person, but not by FFA.
Thus, this is not in line with the Haxby model. The only condition where FFA adapts is first condition which is simply the same image. Makes use of humans ability to recognise different pics of same person but FFA doesn’t do this.
