Design And Implementation Of Magnetic Resonance Imaging Algorithm Based On PROPELLER Technology

Magnetic resonance imaging (MRI) can provide images with high soft tissue contrast, spatial resolution and without harmful radiation. Therefore, MRI has been widely applied in clinics. However, long data acquisition time makes patient susceptible to independent or non-independent motion. As a result, patient motion may reduce image artifact and limit detection of pathological findings. Since magnetic resonance imaging equipment emerge, how to effectively overcome the impact of move has been hotspot and difficulty of magnetic resonance imaging study.This paper has studied how to eliminate motion artifact and magnetic sensitivity artifact. Design and realize the algorithm of magnetic resonance imaging based on the PROPELLER sampling technique, and achieve good results. Firstly, this paper adopts PROPELLER sampling technology to eliminate K-Space offset through phase correction, eliminates high weight of K-Space center through rotation correction and eliminates linear translation of data strip through translation correction. Finally, this paper reconstructs image with the Jackson gridding algorithm and eliminates image artifact.In addition, because the disadvantage that the existing PROPELLER technology adopted programmable graphics hardware GPU acceleration causes conflict of write-write to result in inaccurate faults. This paper proposes the reverse gridding algorithm to solve the problem,whose reconstruction result is consistent with the PROPELLER technology, ensures the consistent results after reconstruction algorithm based on GPU PROPELLER.Through a lot of experiments, this paper has confirmed the PROPELLER and the reverse gridding algorithm is effective, eliminated artifact to a certain extent and improved the quality of magnetic resonance imaging. This algorithm is applied to Neusoft Philips Superstar 0.35T MRI system.