|
The accuracy of PET (Positron Emission Tomography) affects the disease diagnosis。 This research is to focus on PET physics difference analysis and find out the possible influence factor to PET imaging quality in order to increase imaging reliability and establish an evaluation form、research data and method. At first, using calibration source and Phantom to establish experiment group data。 Second, using PET to gain image analysis。 Third, using Xeleris to collect data。 Forth, using statistic analysis software ( Minitab、Matlab and EXCEL) to verify data. The result shown: 1. when used Ge68 standard source at different collimator position for 15 times, we acquired PMT Energy Peak 144.83±0.71、RMS Peak error 0.736﹤2 which matched to system regulation。 2. after calibration procedure was completed, selected high count rate and smooth waveform from Collimator Position 10(P10) then received WCC Factor as well as SUVbw as group analysis by using Regression, and it shown propotional linear Relationship。 3. the correlation analysis of calibration factor influence the imaging quality。 We put 15 group of recombination Flood Phantom out of same activity (30MBq) of WCC Factor group and analyze Flood Phantom signal-to-noise (SNR) by using Matlab software then receive highest SNR 47.873 from Collimator Position 10th imaging。 4. The influence evaluation of WCC factor to clinical human body imaging。 Using statistical method (T-test) to calculate the distribution tolerance of SUVbw and setting up clinical image of SUVbw 0.63523 as expected value. The result shown: Firstly,no obvious and direct influence from single photon to image quality;Secondly, On the procedure of 2D Normalization Correction at PET, we found obviously difference between bed and bed imaging overlap, Non-Normalization group has less 35% SNR. We examined the difference from Flood Phantom imagingm, the SUVbw value differ only 0.02, thus there’s no difference on image identification ; Thirdly, on the procedure of Calibrate the Collimator , we examined on 15 Collimator Position by using Matlab software, we obtain optimum uniformity of Flood Phantom from Collimator Position 10th (P10)。 Fourthly, WCC experiment procedure proved that the WCC will be influenced by collimator position, data from P10 are the most stable one, it won’t affect SUVbw reading. If we compare two group data from same WCC Dose (before and after calibration) by using Regression and analyze the difference between SUVbw and WCC factor, it shown that new calibration factor revised PET system error。 To conclude above calibration processes, we learn that the major influence to the PET imaging quality will be Collimator Position. When collimator position was at optimum position of receiving photon energy, it will be the opposite position of BGO photon receiver (for which we call crystal central) and also has the optimum optical sensitivity. When BGO photon receiver working at its highest value and highest efficiency, the variation of FDG calibration activity to WCC factor will be at smallest influence level, therefore, the correct Collimator Position is able to revise image artifact efficiently and also reduce SNR.
|