Correction Of Rotational Motion Artifacts In MRI

Magnetic resonance imaging (MRI) is an imaging technology which using the nuclear magnetic resonance (NMR) property of organism. It has many features such as high resolution, multi imaging parameters, non-invasive and arbitrary slice imaging, etc. Now MRI is one of the most important tools in medical diagnostics and widely used for in-vivo studies.During the MRI scanning process, due to patients is often difficult to avoid the Motion, resulting in motion artifact, which have significant impacts on serious disruption of the normal medical diagnosis. Motion is the most prominent one of MRI technical difficulties. If objects do not consider deformation of the organization, motion is usually assumed to be rigid motion, rigid motion decomposed into translation motion and rotational motion. For translation artifact correction many scholars at home and abroad carried out in-depth research, but rotational artifact correction is more complex than translation artifact correction, research results is less. Therefore, this paper focuses on correcting motion artifact caused by rotation, improving the quality of magnetic resonance imaging, so that magnetic resonance technology in biological and medical aspects has more practical significance.In this paper, the principle of MRI imaging, rotational artifact causes and correction method was carried out in-depth research, we carried out the research work about rotational artifact correction from pre-treatment and post-processing. When MRI system is collecting data and not reconstructing the MRI image, you can repair of damaged data through PROPELLER (Periodically Rotated Overlapping Parallel Lines Enhanced Reconstruction) rotational correction algorithm, to achieve rotational artifact correction. However, running time of the traditional algorithm is very long, and rotation parameter estimation accuracy is not high. This study proposes a circular gridding-based algorithm for rotation correction, running speed of the method is much faster than that of conventional algorithm, and image quality can be improved. For MR images which have been created and contain rotational motion artifacts, this study proposes a new post-processing algorithm- Rotational Artifacts Correction Based On Fast Scanning Sequences, the proposed algorithm can remove the rotational artifacts efficiently, with high speed and accuracy, have a great improvement than Similar algorithm . Our meaningful and detailed research work is organized as follows:(1) A circular gridding-based algorithm for rotation correction is proposed. the data in central overlapped area is gridded onto a series of round, which can avoid many of the grid when data strips are rotation corrected. Besides, we propose more efficient similarity measure formula, by calculating the value of measure, we can estimate the rotation parameters and correct data strips. Experimental results demonstrate that running speed of the method is much faster than that of conventional algorithm, and image quality can be improved.(2) Rotational artifacts correction based on fast scanning sequences is proposed. First, it divides the whole K-space into several data strips, and then calculates the rotational angle from the similarity of the overlap data between the rotational strips. Second, it compensates the density of the sampled data,and then gets the corrected image by the promoted gridding method. The test results show the proposed algorithm can remove the rotational artifacts efficiently, with high speed and accuracy.