Advances in artificial intelligence (AI) and machine learning have led to successful applications of automated medical image analysis. Deep learning algorithms, a subset of machine learning, are particularly suited to learn complex patterns in data to make automated predictions for new images.1 Several studies have demonstrated that deep learning algorithms applied to medical images may have broad future clinical applications.2,3 If successfully embedded within future healthcare systems, AI may reduce the costs and time associated with reading and interpreting brain scans. However, using these tools in clinical practice has been elusive for various reasons.4,5
