AACE Biomarker Protocol — Experimental Design
This is a collaboration proposal, using facilities at CEBSIT, Chinese academy of Sciences. We hope to develop an experimental protocol for identifying subclinical oculomotor and electrophysiological biomarkers predictive of Acute Acquired Concomitanct Esotropia (AACE), combining high-resolution eye tracking (EyeLink) with 32-channel EEG under dichoptic stimulation.
Rationale
AACE incidence is rising due to excessive near-work and digital device use, which increases medial rectus muscle tension and disrupts fusion. This protocol targets pre-symptomatic biomarkers — measurable before manifest deviation occurs — to enable early detection and preventive intervention.
- Eye tracking is the primary modality, hopefully capturing the most direct downstream motor manifestations of increased medial rectus tension (asymmetric saccade gain, fixation instability, poor binocular coordination).
- EEG (32-64 channel) capturing functional deficits in the visual cortex and dorsal pathway that may change before the oculomotor system completely decompensates.
See the main acute-acquired-concomitant-esotropia page for clinical context, classification, and treatment.
Task Battery
I think we should run these back to back for each subject. I’ve included a 5 min break between tasks 1-3 and task 4.
Pre-task Setup:
- Subjects may need to wash their hair, depending on electrode type.
- EEG Cap fitting and initial impedence checks - 15 to 30 mins.
- Acclimatisation to eyetracker.
- Eyetracker calibration.
- Recheck and fix all EEG channel impedences.
Task 1: Resting State (Baseline Cortical Network) (~10 mins)
- 5 min eyes-closed + 5 min eyes-open (double check time of recording for previous studies!)
- Biomarkers: Alpha-band PSD (cortical idling), IHD-PSD (interhemispheric differences in frontal regions linked to AACE connectivity deficits)
- relevance: Resting functional connectivity in vergence-related frontal networks may predict future decompensation
Task 2: Precision Fixation Stability (~5 mins)
- Central Gaussian spot (0.5°) fixation for 20 s, 8 s inter-trial-interval, repeated 10 times.
- Biomarker: Fixation instability — microsaccades and slow drift. AACE is characterised by a “dominant deviating eye” that may show subtle instability even before manifest deviation
Task 3: Oculomotor Dynamics (~8 mins)
- Saccades: Target jumps stochastically ±20° horizontally; subject makes rapid refixation
- Smooth pursuit: Step-ramp target as a lissajous figure (step-ramp prevents early saccades at pursuit onset)
- Biomarkers:
- Saccade gain asymmetry — differences between adduction/abduction amplitude and velocity. Binocular saccade coordination is significantly poorer in AACE even after clinical “recovery”
- Smooth pursuit gain — reduced gain reflects breakdown in the binocular coordination feedback loop
5 MINS SUBJECT BREAK, RETEST EEG and recalibrate eyetracker if necessary
Task 4: Dichoptic SSVEP & Perceptual Eye Position (PEP) (~15 mins)
- PEP calibration: Participant aligns a dichoptic ”+” symbol within a fixed ”○” circle to record the subjective strabismus angle (perceptual eye position)
- SSVEP design: Circular flickering targets (8° × 8°) presented dichoptically. Left eye: f₁ = 13 Hz; Right eye: f₂ = 15 Hz
- Biomarkers:
- Intermodulation frequency (f₁ ± f₂) — direct measure of binocular integration. Lower intermodulation responses correlate strongly with larger perceptual deviations in AACE
- SSVEP asymmetry index — comparing SNR between eyes to quantify subclinical suppression
Measurement & Analysis Plan
| Category | Measure | Significance |
|---|---|---|
| Eye-Tracking | Binocular Coordination | Difference in gain between dominant and non-dominant eye. Poor coordination is a core AACE signature |
| EEG | α-Band PSD (8–13 Hz) | Lower alpha power in parietal/occipital regions indicates visual stress and compensatory processing |
| EEG | IHD-PSD | Voxel-mirrored homotopic connectivity (VMHC) deficits in frontal lobes — demonstrated in AACE patients |
| EEG | SSVEP IM SNR | Intermodulation SNR reflects cortical “fusion strength”; reductions predict failure of binocularity |
- For smooth pursuit, see this recent analysis technique: [@shishido2026]
- We have previously used SVMs to classify microsaccades in schizophrenia [@liu2023a], we should consider using ML or DL techniques for analysing eye movements to extract biomarkers, see [@asmethajeyarani2023] as an example for other diseases.
Technical Implementation
- Software: MATLAB with Psychtoolbox (PTB) and Opticka toolbox. Analysis will use Nierhorster 2010 toolbox for eye movement analysis.
- Display: IDEALLY Dichoptic presentation via anaglyph (cheap but less good) or shutter glasses (more expensive) to isolate eye-specific responses.
- Hardware: EyeLink binocular eye tracker + 32 (or 64?) channel EEG amplifier
- Synchronisation: Opticka + LabJack T4 provides high-precision temporal syncing between EyeLink and EEG amplifiers, essential for calculating smooth pursuit gain during specific EEG oscillation windows