---
id: wiki-2026-0508-amygdala-hyperactivity
title: Amygdala Hyperactivity
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [Amygdala Hyperreactivity, Limbic Hyperactivation]
duplicate_of: none
source_trust_level: A
confidence_score: 0.9
verification_status: applied
tags: [neuroscience, anxiety, ptsd, amygdala]
raw_sources: []
last_reinforced: 2026-05-10
github_commit: pending
tech_stack:
  language: python
  framework: nilearn/mne
---

# Amygdala Hyperactivity

## 매 한 줄
> **"매 amygdala 의 exaggerated response 의 threat / emotional stimuli — 매 anxiety, PTSD, depression 의 transdiagnostic biomarker."** 매 fMRI BOLD response 의 elevation (특히 face/threat tasks) — 매 prefrontal regulation 의 hypoactivity 와 pair — 매 2026 에 SSRI, CBT, TMS, psychedelic-assisted therapy 의 normalization target.

## 매 핵심

### 매 circuit
- **Amygdala**: 매 basolateral (BLA, threat learning) + central (CeA, autonomic output).
- **vmPFC / dlPFC**: 매 top-down inhibition — 매 hyperactivity 와 anti-correlation.
- **HPA axis**: 매 amygdala → CRH → cortisol — 매 chronic activation 의 maladaptive.

### 매 conditions
- **Anxiety disorders**: GAD, social anxiety, panic.
- **PTSD**: 매 trauma-associated cue 의 sensitization.
- **MDD**: 매 sad-face bias.
- **BPD**: 매 emotional reactivity.
- **Autism**: 매 mixed — face-processing 의 atypical activation.

### 매 응용
1. Diagnostic biomarker (research stage).
2. Treatment response prediction (SSRI, exposure therapy).
3. Neurofeedback / TMS target localization.

## 💻 패턴

### fMRI BOLD extraction
```python
from nilearn import image, masking, datasets

# Load Harvard-Oxford amygdala mask
atlas = datasets.fetch_atlas_harvard_oxford('sub-maxprob-thr25-2mm')
amyg_mask = image.math_img("img == 10", img=atlas.maps)  # left amyg label

# Extract task BOLD
bold = image.load_img("sub-01_task-faces_bold.nii.gz")
amyg_ts = masking.apply_mask(bold, amyg_mask).mean(axis=1)
```

### Threat > neutral contrast
```python
from nilearn.glm.first_level import FirstLevelModel

events = pd.DataFrame({
    "onset": [0, 20, 40, 60],
    "duration": [10]*4,
    "trial_type": ["threat", "neutral", "threat", "neutral"],
})

flm = FirstLevelModel(t_r=2.0, hrf_model="spm")
flm.fit(bold, events=events)
contrast = flm.compute_contrast("threat - neutral", output_type="z_score")
```

### Functional connectivity (amyg-vmPFC)
```python
from nilearn.connectome import ConnectivityMeasure

# Time series from amyg + vmPFC ROIs
ts = np.column_stack([amyg_ts, vmpfc_ts])
conn = ConnectivityMeasure(kind="correlation")
fc = conn.fit_transform([ts])[0]  # 2x2 corr matrix
amyg_vmpfc_fc = fc[0, 1]  # negative in healthy, weaker in anxiety
```

### HRV proxy (peripheral readout)
```python
import neurokit2 as nk

ecg = nk.ecg_clean(ecg_signal, sampling_rate=500)
peaks = nk.ecg_peaks(ecg, sampling_rate=500)[0]
hrv = nk.hrv_time(peaks, sampling_rate=500)
# Low RMSSD ↔ high sympathetic ↔ amyg overdrive
```

### Real-time fMRI neurofeedback target
```python
def neurofeedback_signal(current_volume, amyg_mask, baseline_mean, baseline_std):
    activation = masking.apply_mask(current_volume, amyg_mask).mean()
    z = (activation - baseline_mean) / baseline_std
    # Display inverted bar — patient learns to downregulate
    return -z
```

## 매 결정 기준
| 상황 | Intervention |
|---|---|
| Acute anxiety | Benzodiazepine (short-term), breathing |
| Chronic anxiety | SSRI/SNRI + CBT |
| PTSD | Trauma-focused CBT, EMDR, prazosin (nightmares) |
| Treatment-resistant | TMS (dlPFC), ketamine, psilocybin trials |
| Research / monitoring | fMRI + HRV biomarkers |

**기본값**: 매 CBT + SSRI — 매 6-12 weeks 의 expected normalization.

## 🔗 Graph
- Adjacent: [[Autism-Spectrum-Disorder]]

## 🤖 LLM 활용
**언제**: 매 patient psychoeducation, 매 literature summarization.
**언제 X**: 매 diagnosis, 매 treatment prescription — 매 clinician 의 only.

## ❌ 안티패턴
- **Single-region focus**: 매 amygdala alone — 매 circuit (vmPFC, hippocampus) 의 consideration.
- **State vs trait conflation**: 매 task-induced state ≠ stable trait.
- **Reverse inference**: 매 amyg activation = "fear" — 매 many functions.
- **fMRI as diagnostic**: 매 group-level 의 individual 의 X.

## 🧪 검증 / 중복
- Verified (Etkin & Wager 2007 meta-analysis, Shin & Liberzon 2010, Stein et al. 2007).
- 신뢰도 A.

## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — circuit + biomarker patterns |