News & InsightsEvidence & Regulatory

Hematoma Expansion Prediction - 20 years, no clinical standard

Share

🧠 Hematoma Expansion Prediction — 20 Years In, Still No Clinical Standard

After two decades of research, predicting hematoma expansion (HE) in ICH remains one of neurocritical care’s most critical unsolved problems.

The evidence shows:

  • 📊 NCCT markers (blend sign, black hole sign, hypodensities) → High specificity (~~92–95%) but low sensitivity (~~28–32%) → ~70% of expanders still missed
  • • 📈 Composite & CTA-based tools → BAT score: C-statistic ~0.77 → CTA spot sign: pooled OR 8.49 (29 studies) → Yet no routine clinical adoption
  • • ⚠️ Definition inconsistency → “Expander” rates vary from 13% to 44% depending on criteria → IVH-inclusive definitions improve outcome prediction, but challenge biomarker consistency
  • • 🔬 Latest validation (Zhu et al., Neuroradiology, Mar 2026, n=685) → Combined models maintain AUC 0.82–0.87 across definitions
  • • 🤖 AI / Radiomics → Single-center: AUC up to 0.90+ → Multicenter reality: 0.76–0.81 → Persistent overfitting gap

The field needs to go to:

  • 🧩 Multimodal integration — imaging + CTA + clinical data in one architecture
  • • 📉 Beyond AUC — calibration + decision-curve validation
  • • 🏥 Tiered deployment — from NCCT-only settings to comprehensive stroke centers
  • At PurpleAI, our HE prediction research is built on this exact framework:
  • ➡️ Automated ➡️ Multimodal ➡️ Multicenter-validated

The gap is no longer in model development — it’s in bridging published performance to real clinical utility.

image.png