The Research And Design For Scintillation Crystal Measurement System

Positron Emission Tomography(PET)is a new nuclear medical instrument with high sensitivity and specificity.It can discover key lesions in the fields of tumor,heart and brain in advance and provide clinical evidence for diagnosis.There are many detection channels in the PET system.The uniformity of the detection efficiency between different channels is the prerequisite for ensuring the imaging quality.As the front-end hardware in PET detector,scintillation crystals’ performance will directly affect the detection efficiency of one channel.However,due to the harsh growth conditions of the scintillation crystals,slight environmental changes during the growth process deteriorates the performance of the grown crystals.Therefore,the performance of the scintillation crystal needs to be measured to ensure the quality and the consistency of performance and obtain better imaging quality.With the aging of the population and the increasing awareness of healthcare,the demand for high-lever medical equipment continues to increase,which expands the market capacity of PET and increase the demand for scintillation crystals.At the same time,in order to obtain higher system sensitivity,the PET system will be developed with longer length,which means that the demand for scintillation crystals will increase further.As a result,the performance measurement of scintillation crystals will face the need for high-volume and efficient testing.However,the existing scintillation crystal measurement device is not mature enough and the system architecture is not perfect.When faced with a large number of crystals,there are still some problems such as low test efficiency,unreliable test results,lack of support for data analysis platforms and the inability to share and mine data.Based on such demand,this thesis studies the architecture of the scintillation crystal measurement system,modularizes and componentizes the system components from the aspects of mechanical control,electronics solutions and software platforms,and proposes a set of scintillation crystal measurement system.The system will solve the problem of the existing measurement device when testing large batch crystals and finally ensure the quality and performance consistency of the large number of scintillation crystals.The mechanical control system proposed in this thesis adopts the design of rotation structure and cross-motion mode to solve the problem of low test efficiency caused by the unstable mechanical structure in the existing system.The design has been demonstrated that the theoretical efficiency can be increased by 48 times when testing crystal arrays.The electronics solution and software platform,as the form of a performance analysis platform are the core components of the measurement system.The platform focus on the data processing,analysis,storage and sharing with high efficient and accurate,to solve the problems of the existing system,such as unreliable test results,lack of data analysis platform support and the inability to share data.Among them,in order to obtain a higher count rate,a digital scintillation detector based on MVT sampling is used to obtain the data stream.In order to provide the platform support for the data analysis and processing,an interactive visualization software system is established based on the Qt framework.The response segmentation and event clustering algorithm are optimized by Do H-based local extremum detection algorithm to improve the speed and accuracy of data analysis.A total of 507 scintillation crystals in 9 test combinations are taken as the measured objects are identified and verified.The accurate recognition rate reaches 100% and the algorithm execution time is less than 1 second.At the same time,for the storage and sharing of performance data,a cloud-based data sharing system is implemented based on the separate front-end and back-end architecture.Besides,a database for scintillation crystal performance parameters is established.Finally,the feasibility of data sharing system is verified through the visualization of test data.