Research Of Passive Shimming Measurement System And Algorithms For Superconducting Magnets Of MRI

As a novel non-invasive and non-radioactive imaging technique,MRI(magnetic resonance imaging)took an extremely important place in the field of medical imaging diagnosis.Superconducting MRI systems have steadily become the main stream medical imaging devices,with its advantages of high magnetic field strength,stability and homogeneity.The main magnetic field homogeneity within the Diameter Spherical Volume(DSV)of MRI has the determining effects on final image quality.The field homogeneity of DSV is however affected by external factors such as manufacturing production,engineering installation and external environment,it is therefore necessary to adjust and improve the field homogeneity of DSV to certain expected level through a series of shimming procedures before the system can be used in clinic.Passive shimming technique uses shims to produce small magnetic fields,by compensating the field inhomogeneity of DSV to achieve the purpose of shimming.In this paper,according to the design of a new measurement system,actual magnetic field data of the example magnet DSV is measured.Based on the measurement results,the proposed passive shimming algorithm is simulated and the main results are as follows:1.Designed a portable DSV field mapping measurement system,to overcome the drawbacks of traditional DSV field measurement system with MRI systems such as the inaccuracy of probe positioning and movement,tediousness of data recording,and complexity of operating procedure.A labview program(serial communication,data acquisition program and human-computer interaction interface)was developed to control the Gauss meter,and the measurement points on the surface of the DSV were divided,and the position of the probe was precisely moved by adjusting the measurement device.Real-time acquisition of DSV surface measurement point magnetic field data was realized by the system without manual recording and it can save magnetic field data to local by exporting data to TXT file.The system has been actually tested in the project,and the operation is convenient and efficient.The measured DSV uniformity was 416.0506 ppm.2.A passive shimming algorithm based on L1 norm constraint was proposed to solve the shortcomings of the traditional linear programming passive shimming algorithm such as insufficient enhancement of DSV uniformity and difficulty in balancing DSV uniformity and total thickness of shim.The algorithm uses the L2 norm to regularize the magnetic field error,and uses the L1 norm to regularize the thickness of the shim thickness matrix.While ensuring the best homogeneity,minimize the use of shimming plates and weaken eddy currents.Through the constraint conditions,the DSV magnetic field uniformity and the total shimming thickness can be effectively balanced.Compared with the linear programming algorithm,the optimal solution of the algorithm is 0.67% difference in the total thickness of shimming pieces and 48.32% in DSV magnetic field uniformity.The uniformity after the linear programming optimization is 16.0338 ppm and the uniformity after the L1 norm constrained optimization is 8.2892 ppm.