Data Processing And Performance Analysis Of Small Animal Panel-PET

In recent years,the role of PET technology in the field of nuclear medicine has become increasingly important,and its application has become more widespread.By injecting drugs with radionuclides into organisms,it can achieve non-invasive accurate imaging of organisms,reflect the metabolic status of birth objects,judge the presence or absence of tumors,examine neurological diseases and cardiovascular diseases,and so on.Due to its unique structure,the panel PET system has important application value and market prospects in the PET field.In the small animal panel PET system,preprocessing of raw data and subsequent image analysis are indispensable,this thesis will conduct relevant research on these tasks.Aiming at the large amount of raw data collected by small animal panel PET systems,an off-line rapid preprocessing method suitable for panel PET system data is proposed.Through this process,the preceding data can be quickly obtained,which facilitates subsequent reconstruction and image analysis.The rapid preprocessing is mainly for the PET data to perform the time coincidence first which removes most of the invalid data,and then perform position correction,energy correction and crystal coding,and finally obtain the LOR information.The results show that the off-line rapid preprocessing method in this thesis can process 13.5 GB in 176 seconds,basically achieving the purpose of rapid processing.Based on the off-line rapid preprocessing,this thesis proposes an online preprocessing method for PET data.The on-line preprocessing method first stores the collected data in a computer memory buffer,and then preprocesses the data through multi-threading,and finally writes the processed conforming data into the hard disk for preservation.The method effectively avoids writing a large amount of original data to the hard disk,and at the same time,it also avoids the problem of slow read and write of the hard disk.At the same time as the data acquisition,the preprocessing can be basically performed synchronously,which fully improves the overall efficiency of the PET system.For the problem that the gaps between the detector modules in the panel PET system affect the imaging quality,this thesis through a series of simulation experiments,artificially adding gaps of different widths in the data preprocessing process,the sensitivity,spatial resolution of the PET system,the performance of the image contrast and coefficient of variation are studied.The results show that gaps of about 4mm width existed between the panel PET systems basically has no effect on the system performance;when the gap width between the crystals reached 16 mm,the sensitivity and image contrast of the PET system have slight influence,the spatial resolution has basically no effect;when the gap width is greater than 20 mm,the performance of the PET system is greatly affected.Finally,after proper modeling,the problem of reconstructed image deformation caused by gaps between detector modules is solved.To test the overall performance of the self-made PET system,the spatial resolution experiment,sensitivity experiment and tumor nude mouse experiment are designed and implemented,and performance comparisons are made with commercial PET systems.The experimental results show that the self-made PET system has a spatial resolution of 1.2 mm,a central sensitivity of 12%,and good tumor nude mouse imaging.In general,the performance of dual panel PET system is not weaker than commercial equipment.