深度学习在头颈动脉病变医学影像诊断中的研究进展与挑战OA

Head and neck carotid artery diseases are an important cause of disability and death worldwide.Its accurate diagnosis is crucial for clinical decision-making.Traditional imaging diagnosis methods face bottlenecks such as low efficiency and strong subjectivity.Deep learning technology has provided an innovative solution to overcome these limitations.This article systematically reviews the latest progress of deep learning in the diagnosis of cerebrovascular arterial disease.The study also reveals breakthrough applications of deep learning in the diagnosis of arteriovenous malformations and Moyamoya disease.Despite limitations such as insufficient model generalizability(mostly based on single-center data),false-positive rates(e.g.,YOLOv5),and the"black-box"issue of interpretability,deep learning has significantly improved diagnostic efficiency and has promoted personalized risk assessment and clinical decision support.In the future,it will be necessary to overcome existing bottlenecks through multicenter validation,development of interpretable algorithms,and federated learning.The fusion of multimodal imaging and multidisciplinary collaboration will become an important direction for intelligent diagnosis and treatment of cerebrovascular diseases.