专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

WO2015164825A1 DUAL SPACE DICTIONARY LEARNING FOR MAGNETIC RESONANCE (MR) IMAGE RECONSTRUCTION 审中-公开
标题翻译：用于磁共振（MR）图像重建的双空间词典学习
公开(公告)号：WO2015164825A1
公开(公告)日：2015-10-29
申请号：PCT/US2015/027648
申请日：2015-04-24
申请人： YUAN, Chun , ZHOU, Zechen , WANG, Jinnan , BALU, Niranjan
发明人： YUAN, Chun , ZHOU, Zechen , WANG, Jinnan , BALU, Niranjan
IPC分类号： G01R33/54
CPC分类号： G01R33/5608 , G01R33/5611
摘要： This disclosure relates to techniques for magnetic resonance (MR) image reconstruction. A herein-described k-space dictionary learning (KDL) technique and a herein-described dual space dictionary learning (DSDL) technique can use dictionaries to approximate an image and/or a data matrix in a signal observation domain for MR imaging by a weighted sum of dictionary entries. A herein-described self-supporting tailored k-space estimation for parallel imaging (STEP) technique can use k-space partitioning and basis selection tailored procedures to promote a spatially variant signal subspace and incorporation into a self-supporting structured low rank model that enforces locality, sparsity, and rank deficiency properties. The model can be formulated into a constrained optimization problem solvable by an iterative algorithm.
摘要翻译：本公开涉及用于磁共振（MR）图像重建的技术。本文所述的k空间字典学习（KDL）技术和本文所述的双空间字典学习（DSDL）技术可以使用字典来通过加权的（k）空间词典学习（DSDL）技术在MR成像的信号观察域中近似图像和/或数据矩阵字典条目的总和。本文描述的用于并行成像（STEP）技术的自支撑定制的k空间估计可以使用k空间划分和基本选择定制程序来促进空间变异的信号子空间并且并入到自我支持的结构化低阶模型中，地方，稀疏性和等级缺陷属性。该模型可以被形成一个由迭代算法解决的约束优化问题。

IPRDB

热门服务

关于我们

友情链接

联系方式