能谱曲线方程数学参数对碘溶液鉴别能力初探

A Preliminary Study on the Ability of Mathematical Parameters of the Spectral Curve Equation to Identify the Iodine Solution

目的:探讨能谱曲线方程数学参数对不同浓度碘溶液的鉴别能力。材料与方法:将碘海醇注射液与纯净水按比例(1:20,2:20,3:20及4:20)配制成4支不同浓度的溶液,应用GE Discovery CT 750 HD能谱扫描仪对4支溶液行能谱成像扫描(gemstone spectral imaging,GSI)。扫描数据传入GE AW4.6后处理工作站处理,测量各溶液单能量下(40-140Ke V,间隔5Ke V)的CT值。应用WPS Office(2016)表格软件将CT值分别以指数、线性、对数、多项式及幂函数作为趋势线进行曲线拟合,得到各函数的拟合方程。比较各方程决定系数(R2值),以愈接近1为拟合程度最佳。将拟合度最佳的曲线作为溶液能谱曲线方程,运用Matlab R2014a(8.3.0)软件计算其定积分(∫14040ydx)及极限(limx→140Ke V)。采用SPSS 19.0统计学软件中单因素方差分析比较4支溶液各函数能谱曲线方程R2值、拟合度最佳曲线方程的定积分及极限。结果:曲线拟合以幂函数拟合度为最佳,R2mean=0.9979(F=824.186,p=0.000);各溶液能谱曲线方程(以幂函数)的定积分、极限差异均有统计学意义(F定积分=7628.401,p=0.000;F极限=2306.063,p=0.000)。结论:能谱曲线方程数学参数可为物质鉴别提供更多信息。

Purpose： To investigate the distinguishing ability of the mathematical parameters of the spectral curve equation for iodine solutions with different concentrations. Materials and Methods： 4 different concentrations of Iohexol Injection and purified water mixed in different proportions（ 1： 20,2： 20,3： 20 and 4： 20） were prepared,All solutions underwent gemstone spectral imaging（ GSI） mode by GE Discovery CT 750 HD. The obtained data was sent to the GE AW4. 6postprocessing workstation,the CT attenuation value of 4 solutions were measured in monochromatic images（ 40- 140 ke V,interval 5Ke V）. The CT values are fitted by the exponential,linear,logarithmic,polynomial and power function as the trend line,and the fitting equation of the function is obtained by WPS Office table（ 2016） software. The coefficient of determination（ R2value） of these equation is compared,the more close to 1,the more optimum. The best fit curve was used as the spectral curve equation of these solutions,and Matlab R2014a（ 8. 3. 0） calculated the definite integration and the limit of these equations. The single factor variance analysis was used to compare the R2 value of the equation of energy spectrum for each function of 4 Solutions、The definite integral and limit of fitting the best curve equation in SPSS 19. 0 software. Results： On curve fitting,The fitting degree of power function was the most optimum,R2 mean = 0. 9979（ F = 824. 186,p = 0. 000）; the difference of definite integration and limit of the spectral curve equation was statistically different（ F = 7628. 401,p = 0. 000; F =2306. 063,p = 0. 000,respectively）. Conclusion： The mathematical parameters of the spectral curve equation can provide more information for material identification.