DoseLab软件检测CT图像噪声的程序改进及应用分析

Improvement of DoseLab software program for CT image noise measurement and its application

目的:对DoseLab软件进行程序改进,增加检测CT图像噪声的功能,对改进的程序进行测试分析。方法:首先,通过使用圆的内接多边形顶点位置计算公式,得到圆内接正三十二边形顶点坐标值。然后,在DoseLab软件Catphan 504模体CTP486模块的图像分析程序中,添加一个正三十二边形的感兴趣区(ROI),用于检测CT图像噪声。选取2018年每月由西门子CT模拟机日常质量检测(DQC)程序得到的水模体两个层面(S3和S4)的CT图像,对DoseLab改进程序进行测试。对DoseLab改进程序和DQC程序得到的CT图像噪声数据,进行统计分析和比较研究。结果:根据公式计算得到了半径4 cm圆的内接正三十二边形的32个顶点的坐标值,该多边形ROI的面积为49.94 cm2。计算DoseLab改进程序和DQC程序得到的CT图像噪声的差异(ΔN)。在120 kV情形,S3和S4层的ΔN值分别为(0.06±0.07)HU和(0.03±0.09)HU;在140 kV情形,S3和S4层的ΔN值分别为(0.10±0.09)HU和(0.08±0.09)HU。结论:通过添加正三十二边形ROI得到的DoseLab改进程序,可以自动分析水模体和Catphan模体,得到CT图像噪声数据。

Objective To improve DoseLab software program by adding the function of CT image noise measurement,and to perform a test analysis on the improved DoseLab program.Methods Firstly,the vertex coordinates of the regular 32-sided polygon were obtained by the vertex formula of inscribed regular polygon of a circle.Then,a regular 32-sided polygon region of interest(ROI)used for CT image noise detection was added in the image analysis program of DoseLab software for the CTP486 module of Catphan 504 phantom.The CT images of the water phantom of two slices(S3 and S4)which were obtained by the daily quality check(DQC)program of Siemens CT simulator every month in 2018 were selected to test the improved DoseLab program.The CT image noise data obtained from the improved DoseLab program and DQC program were analyzed and compared.Results According to the vertex coordinates of regular 32-sided polygon in a circle with a radius of 4 cm,the area of the polygon ROI was 49.94 cm2.Difference in image noise(ΔN)between improved DoseLab program and DQC program were calculated,and the results showed that theΔN for S3 and S4 was(0.06±0.07)HU and(0.03±0.09)HU under the voltage of 120 kV,and that the correspondingΔN was(0.10±0.09)HU and(0.08±0.09)HU under the voltage of 140 kV.Conclusion The improved DoseLab program with the addition of a regular 32-sided polygon ROI can be used to analyze water phantom and Catphan phantom automatically,thus obtaining CT image noise data.