A clinical–radiomics model based on noncontrast computed tomography to predict hemorrhagic transformation after stroke by machine learning: a multicenter study

Abstract

Key points Noncontrast computed tomography (NCCT) is valuable for predicting hemorrhagic transformation (HT) after intravenous thrombolysis (IVT) treatment. Machine learning is vital for predicting HT. NCCT radiomics integrated with clinical factors could facilitate predicting HT.

Keywords

• Acute ischemic stroke
• Hemorrhagic transformation
• Noncontrast computed tomography
• Radiomics
• Machine learning