低剂量CT图像DICOM数据量化分析

Quantitative Analysis of DICOM Data in Low-dose CT Images

目的对图像质量进行客观准确的评估,指导医学图像处理软件开发及临床低剂量CT检查参数的优化。方法通过CT机扫描获得断层图像,选取兴趣层面使用MATLAB显示灰度值直方图,通过灰度直方图可看出灰度信息的差异,随机选择5个兴趣层面,将不同层面不同mA值图像与150 mA所得图像进行对比,利用MATLAB计算出各图像均方误差和峰值信噪比。讨论由于无法客观的量化评价图像质量,而传统的图像质量客观评价方法主要包括均方误差(MSE,Mean Squared Eror)和峰值信噪比(PSNR,Peak Signal To Noise Rate)。统计数据,分析可得,当mA值在70-150 mA范围内PSNR始终大于30 dB,即在这个剂量范围内,图像质量好,失真程度小。在一定范围内,低剂量CT检查技术具有临床诊断价值。

Objective Objectively and accurately evaluate its image quality,guide the development of medical image processing software and the optimization of clinical low-dose CT examination parameters in the future.Methods CT scan was used to obtain the tomographic image,and MATLAB was used to display the gray level histogram at the level of interest.The difference in gray level information could be seen through the gray level histogram.Five interest levels were randomly selected,and images with different mA values at different levels were compared with images obtained at 150 mA.MATLAB was used to calculate the mean square error and peak signal-to-noise ratio of each image.Discussion Because of that image quality could not be evaluated objectively quantitatively.We can use both MSE and PSNR.According to statistical data and analysis,PSNR is always greater than 30dB when the mA value is within the range of 70-150ma.that is,within this dose range,the image quality is good and the degree of distortion is small.Within limits,low-dose CT examination technical can provide a clinical diagnostic value.