BACKGROUND:In spectral CT imaging study, the selection of scanning parameters is considered by most researchers, but the effects of measuring position are often overlooked. Actual measurement found that different measurement location had significant impact on the result of the measurement. Through measurement and mathematical model of a large amount of data, we can correct the measurement results of different location. The results with real data alignment are higher. OBJECTIVE:To explore effects of different measuring positions on results in spectral CT imaging, and to optimize correction coefficients. METHODS:GE standard water phantom was applied to orderly obtain 5 combined scanning parameters with 552 groups of data. Size measurement method was divided into two measuring range:ROI1 and ROI2. We selected 10 points to measure CT value, including Center, North, South, West and East. The measurement data of ROI1 and ROI2 were classified and screened based on the same sequence. Mathematical modeling and probability statistics analysis were used to optimize correction coefficient, get calibration function and draw experimental simulation curve. RESULTS AND CONCLUSION:(1) Measuring methods of ROI1 were superior to the ROI2’s on water phantom in spectrum CT. (2) To different scanning sequences, the measuring results were different. To the same scanning sequences, the measurements for different positions on water phantom in spectrum CT had remarkable influence on the measuring results, which varied from points to points. (3) Through setting up mathematical modeling, using method of statistical analysis, we could get the correction function on different measuring positions. (4) Above results confirmed that compared with the theoretical model and the experimental data of spectral CT scanning parameters, the coefficients of position can be adjusted, which can optimize the measuring results.