Novel CT Image-Based Intracerebral Bleeding Risk Score for Patients With Acute Ischemic Stroke Undergoing...

Area of Science:

• Neurology
• Radiology
• Medical Informatics

Background:

• Symptomatic intracerebral hemorrhage (sICH) is a significant complication following intravenous thrombolysis for acute ischemic stroke (AIS).
• Predicting sICH risk remains challenging due to its unpredictable nature.
• Current prediction methods may not fully integrate imaging and clinical data.

Purpose of the Study:

• To develop and validate a novel predictive model for sICH.
• The model incorporates clinical variables and noncontrast head CT imaging features.
• To improve individualized risk assessment for patients undergoing thrombolysis for AIS.

Main Methods:

• Multivariable logistic regression analysis was used to derive the model.
• The model was developed using a cohort of 808 AIS patients from Southeast China.
• External validation was performed on a separate cohort of 612 AIS patients from Central China.

Main Results:

• The predictive model identified key factors: cerebral small vessel disease burden, early infarct signs on CT, atrial fibrillation, age, systolic blood pressure, and NIHSS score.
• The model demonstrated strong discrimination (C-statistic 0.80 in derivation, 0.82 in validation).
• Good calibration was observed in both derivation (P=0.816) and validation (P=0.866) cohorts.

Conclusions:

• An internally and externally validated prediction model for sICH was developed.
• This model can aid in the individualized prediction of intracerebral bleeding risk.
• Facilitates informed treatment decisions for AIS patients receiving intravenous thrombolysis.