Is there a fast shortcut from the eyes to the amygdala for faces?
Early brain responses to faces were best explained by a model that includes a direct subcortical route to the amygdala, regardless of the face's emotion.
Source
Subcortical amygdala pathways enable rapid face processing
Study at a glance
- Design
- Human experiment — MEG recorded during a gender-judgement task on neutral, happy and fearful faces; competing connectivity models (with vs without a pulvinar-to-amygdala route) compared with Bayesian model selection over growing time windows.
- N
- N=12 · 12 healthy participants, each viewing 99 faces per emotion condition.
- Population
- Healthy adult volunteers
- Outcome
- Model evidence for dual-route vs cortical-only connectivity models of face-evoked MEG fields; M170 amplitude by emotion
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Twelve participants judged the gender of neutral, happy and fearful faces while their brain's magnetic fields were recorded with MEG. The researchers used dynamic causal modelling to compare two families of network models: one in which visual information reaches the amygdala only through visual cortex, and one that adds a direct connection from the pulvinar, a thalamic nucleus, to the amygdala. They compared how well each model family explained the data as the analysed time window was extended from the first tens of milliseconds up to 300 ms.
What they found
Models with the subcortical pulvinar-amygdala connection won for time windows up to about 180 ms, while for longer windows the cortical-only models did about as well. In the dual-route model, amygdala activity peaked around 140 ms and was roughly ten times larger than in the cortical-only model, where it peaked around 180 ms. Emotion made little difference: the face response was only slightly larger for fearful than happy faces, not different from neutral, and no model with emotion-specific connection changes was favoured.
The limits
What it doesn't show
With only 12 participants and weak emotion effects, the study cannot say whether fear specifically recruits the subcortical route. MEG has limited sensitivity to deep structures such as the pulvinar, and the authors acknowledge that the anatomical labels in their models cannot be validated empirically with MEG. Model comparison only tells us which of the tested models fits best, so a different, untested cortical architecture could also explain the data. Emotion was irrelevant to the gender task, which may have reduced emotional effects.
Key terms
- Amygdala
- A pair of almond-shaped structures deep in the temporal lobes, central to detecting emotionally or socially important stimuli.
- Pulvinar
- A large thalamic nucleus that can relay visual information to the amygdala without going through visual cortex.
- Dynamic causal modelling (DCM)
- A method that fits biologically inspired network models to brain recordings and compares them to infer how regions influence one another.
- Bayesian model selection
- A statistical way of deciding which of several models is most likely to have produced the data, balancing fit and complexity.
- M170
- A face-sensitive MEG response that peaks about 170 ms after a face appears.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Which connection distinguished the dual-route models?
Common questions
Why does the timing of the model comparison matter?
If a subcortical shortcut exists, it should matter most early on. Finding that the dual-route model wins only for early windows supports a fast, early role for that route.
Can MEG really see deep structures like the amygdala?
Only indirectly and with lower sensitivity. DCM helps because it models how deep sources affect signals from detectable regions, but the authors note that their anatomical labels cannot be confirmed with MEG alone.
Does this mean the low road is only for fearful faces?
No. The subcortical route helped explain responses to faces in general; no fear-specific boost was found.
More on Emotion circuits