Real Time Cardiac Magnetic Resonance Imaging Based On The Compressed Sensing

Magnetic resonance imaging(MRI)is a non-invasive and non-ionizing imaging technology.It provides various contrast mechanism,making the physiological function of tissues and organs,achieve excellent anatomic structures and pathological information visualization.Due to these advantages,magnetic resonance imaging has been widely used in clinical and scientific research and become one of the most important methods in the field of medical diagnosis.However,the drawback of current MRI techniques is its slow imaging speed,which limits the MRI clinical performance and imaging quality(e.g.,dynamic imaging).Thus,MRI was limited used in cardiac imaging,brain imaging and its detection indication was also greatly reduced.How to improve the MRI speed and image quality is always an important topic in the MRI field.One solution is sparse sampling the k space data to speed up the MRI speed.Based on this method,the main work of this article is to study the reconstruction problem of dynamic cardiac MRI from the sparse sampling k space data.As the dynamic cardiac MR images have a large number of redundant information along the temporal direction,it is possible to use the sparse reconstruction techniques to accelerate the dynamic cardiac MRI.Compressed sensing?dictionary learning and the newly introduced concept of refactoring reference images are the three basic theories of sparse reconstruction algorithms.Based on the theories,two aspects of innovation was made to,improve the accuracy of MRI reconstruction:Frist,make the picture a higher contrast and image quality through the whole refactoring reference image's K space data coming from the MRI;Secondly,we make full use of sparseness and denoising effect of the dictionary learning algorithm.The sparseness can improve the quality of reconstructed image,and the denoising effect of the dictionary learning algorithm ensure that the image contrast and image details.The reconstruction experiments of actual clinical cardiac MRI image show that algorithm we proposed is of highly precision,good stability,and it can effectively save the details of the reconstruct image.