nnU-Net7TMRI

Corinne Donnay1, Henry Dieckhaus2, Charidimos Tsagkas1

  • 1Translational Neuroradiology Section, National Institute of Neurological Disorders and Stroke (NINDS), National Institutes of Health, Bethesda, MD, United States.

Frontiers in neuroimaging
|December 18, 2023
PubMed
概括

深度学习方法,特别是伪标签辅助nnU-Net (PLAn),在多发性硬化症 (MS) 患者的7T核磁共振时显示出改善的自动大脑和病变细分. 与其他技术相比,PLAn显著提高了病变检测的准确性.

相关概念视频