Research On Intelligent Evaluation System Of Medical Magnetic Resonance Quality Control Detection Based On ACR Phantom

Magnetic resonance imaging(MRI)technology has been widely used in various fields of clinical medicine due to its non-invasive and multislice imaging in arbitrary orientation.For the sake of ensuring the efficient operation of the magnetic resonance system,the engineering personnel introduce the concept of quality control(QC).MR quality control refers to the composite method of checking and maintaining equipment by using scientific and reasonable technical method to meet image quality.The traditional procedure of daily test section is to calculate and analyze the parameters of quality image,as well as compare with the industry standard values to judge the operation condition of the equipment.It needs to debug by manual measurement at the post-processing workstation,which has Iong working time and poor stability.The defect is easy to make the evaluation of equipment performance deviate from the real value so that to hinder the popularization and standardization of MR quality control.For the above limitations,we use the related image processing technology to study the intelligent evaluation system of medical MR quality control detection based on Matlab platform and ACR standard testing phantom.Our goal is to replace the traditional method,so that it can reflect the performance of MRI systems accurately and effectively.First of all,we introduce the research background and describes MR quality control procedure in detail,including the ACR phantom,a brief description of the detected parameters and a discussion of the development present situation at home and abroad.It lead to the purpose and significance of out study through the defect of manual method,and arrange the article structure at the same time.In this paper,we study and explain the principle of image processing technology for each module,so that the design idea of the system in subsequent chapter can be clearly stated.Secondly,in Signal-to-Noise Ratio and Percent Image Uniformity module,we discuss the deviation of measurement value and real value of signal and noise,then we use morphological processing to realize automatic detection of parameters.In geometric accuracy module,we put forward to improve the circle detection with Hough transform for the defect of ACR phantom by double-threshold and piecewise fittinging in order to decrease the invalid accumulation to save time and raise the accuracy.In spatial resolution module,we propose to select region to threshold based on normalized separability measurement,also we synthesize many image processing techniques like the region growing and the line detection to extract region of interest(ROI),and finish the detection of spatial resolution by developing a high-efficiency strategy of template matching after that.Finally,we design the interface and function of intelligent evaluation system,so that it can analyze the quality control images imported and realize the automatic evaluation of parameters.Compared with the traditional manual method,the system has the advantages of short post-processing time,high accuracy and the time performance curve by the detection data is more stable.Besides,it can provide the reference for analysis and procurement of different brand equipment.It is helpful for the development and popularization of magnetic resonance quality control.