Improving the diagnosis of acute ischemic stroke on non-contrast CT using deep learning: a multicenter study

Abstract

Key points The model simultaneously segments infarcts and estimates ASPECTS by using baseline CT. A mirror-assembly module plus dual-path DCNN model improved the segmentation efficiency. We evaluated the model in a multi-reader and multicenter (MRMC) setting.

Keywords

• Deep learning
• Non-contrast computed tomography
• Early ischemic changes
• Alberta Stroke Program Early Computed Tomography Score
• Multi-reader and multi-center study
• Acute ischemic stroke