Patient Parameters
Adjust to match the patient profile.
ML-powered clinical decision support for pediatric myopia management. Enter patient parameters on the left to predict annualized axial-length progression.
Model: XGBoost · Validated via LOOCV · Training cohort: n = 171 eyes · For research / demonstration only — not for clinical use
Adjust parameters on the left, then click “Predict Progression”
How to use
- 1Enter the patient's age, eye measurements, and treatment details on the left
- 2Fill the early AXL visits — the early slope is the most predictive feature
- 3Click Predict Progression
- 4Review the predicted mm/yr, risk tier, trajectory and SHAP breakdown
Risk tiers
| Tier | mm/yr | Action |
|---|---|---|
| Slow | < 0.10 | Annual review |
| Moderate | 0.10–0.30 | 3-month follow-up |
| Fast | ≥ 0.30 | Escalate therapy |
Predicted progression
0.000
mm/yr
AXL deviation from age norm:
Risk gauge
mm/yrAnnualized AXL progression band
Slow
<0.10
Moderate
0.10–0.30
Fast
≥0.30
AXL growth trajectory
vs. East Asian norms (He et al., BJO 2023 · n=14,127)
P50 median
P10 / P90 band
This patient
vs. Literature
East Asian children progress faster — compare within same ethnicity for fairness.
Feature contribution (SHAP)
Red bars push prediction higher (faster progression)
Blue bars push prediction lower (slower progression)