Application Of Multi Parameter Landau Distribution Fitting Model In MVT Sampling Method

Multi-Voltage Threshold method that can solve the problem of accurate digitization of high-speed pulses is highly expected in the research of all-digital Positron Emission Tomography technology.The MVT method can accurately extract the energy information and time information from the reconstructed scintillation pulse fitting by a small number of samples and a priori model of scintillation pulse.Therefore,it is very important to choose the fitting model of scintillation pulse.However,Due to the traditional linear exponential fitting model and double exponential fitting model are sensitive to the MVT threshold,the accuracy of energy information which is calculated by reconstructed pulse fitting by those two models still can be improved.Aiming at this problem,a Multi Parameter Landau Distribution fitting model(MPLD)is proposed based on the Landau Distribution fitting model and the shape characteristics of scintillation pulse,the advantages and disadvantages of this model compared with traditional models are explored.The specific work is as follows:Firstly,the performance of MPLD model applied to traditional four threshold MVT which has been well verified and widely used in the development of digital PET detectors is studied.For the four threshold MVT simulation data,relying on the energy resolution,coincidence time resolution,true coincidence count and other indicators,the performance of the MPLD model is evaluated,and compared with that of traditional linear exponential model and double exponential model.A comparative analysis was carried out to explore the advantages and disadvantages of the MPLD model.Secondly,the performance of MPLD model applied to novel nine points MVT and dynamic MVT is studied.The nine points MVT provides additional scintillation pulse characteristics for the four threshold MVT,it’s expected to improve the accuracy of pulse fitting.The dynamic MVT can adaptively adjust the number of thresholds according to amplitude characteristic of the pulse,it is suitable for the calculation of pulse energy in a large dynamic range.The adaptability and potential of MPLD model are explored by comparing and evaluating the performance of MPLD model and traditional models based on simulation data.To sum up,the results show that MPLD model is relatively insensitive to sampling methods,and it’s energy performance and counting performance are relatively good and stable.In the four threshold MVT,the MPLD model achieves good energy performance and counting performance,and the optimal performance of the MPLD model is easier to achieve in applications,and the model has relatively low threshold dependence under certain conditions.In the nine points MVT and the dynamic MVT,the MPLD model obtains good energy performance,and the counting performance is the best.The temporal performance of the MPLD model is relatively poor,but it is expected to be optimized by the study of digital timing methods,and there is room for further exploration.It is hoped that the above research can provide new methods and new ideas for the development of all-digital PET technology.