Predicting Intracerebral Hemorrhage Expansion With Noncontrast Computed Tomography: The BAT Score.

Area of Science:

• Neurology
• Radiology
• Biomarkers

Background:

• Hematoma expansion (HE) is a critical determinant of outcome in spontaneous acute intracerebral hemorrhage.
• Computed tomographic angiography (CTA) spot sign is a known HE biomarker but not routinely used in emergency settings.
• Noncontrast computed tomography (NCCT) is the standard imaging modality in acute stroke evaluation.

Purpose of the Study:

• To develop and validate a novel prediction score for HE based on readily available NCCT findings.
• To provide a practical tool for identifying patients with intracerebral hemorrhage at high risk for HE.

Main Methods:

• A 5-point prediction score (BAT score) was developed using logistic regression in a single-center cohort (n=344).
• NCCT markers analyzed included hypodensities, blend sign, hematoma shape/density, and fluid level.
• The score was validated in two independent cohorts: a large clinical trial population (n=954) and a multicenter cohort (n=241).

Main Results:

• Key predictors of HE were blend sign, intrahematoma hypodensity, and time from symptom onset to NCCT <2.5 hours.
• The BAT score demonstrated good discriminative ability across validation cohorts (c-statistics ranging from 0.65 to 0.77).
• A dichotomized BAT score (≥3) achieved 89% specificity for predicting HE.

Conclusions:

• The 5-point BAT score is an accurate and specific tool for predicting HE using baseline NCCT.
• This easily applicable score can assist in selecting intracerebral hemorrhage patients for antiexpansion therapies and clinical trials.