腮腺良恶性肿瘤CT影像色度学定量研究

COLORIMETRIC QUANTITATIVE STUDY ON CT IMAGES OF PAROTID GLAND WITH BENIGN AND MALIGNANT TUMOR

目的:探讨通过腮腺CT影像色度学定量分析,为腮腺CT图像和其它浓淡图像和彩色图像的定量诊断找到了途径。方法:各选取30例经过病理确诊的罹患单侧良性及恶性腮腺肿瘤的CT数字影像片,利用软件Photoshop Extended对患侧和健侧进行伪彩色处理并用直方图方法来测定其浓淡图像的灰度值,这些灰度值不同于CT值,分为0到255共256个灰度级,属于计算机图像灰度值,同时通过SPSS 17.0统计包对腮腺健侧、患侧灰度值进行统计比较。结果:对患侧和健侧进行伪彩色处理后,比较精确地分别测量得出患良、恶性肿瘤腮腺的灰度值,结果显示,罹患有良、恶性肿瘤的腮腺图像经过伪彩色处理后,进行数字化直方图测量灰度值存在明显差异,即正常组、良性组和恶性组每两组间均数差异均具有统计学意义,供鉴别诊断使用。结论:伪彩色可以对黑白图像进行分层,从而使得人眼能够从经过伪彩色处理的图像中提取更多的图像信息,达到降低人为测量误差、提高判读精度的目的。通过Photoshop伪彩色处理并进行直方图测量后,可以得到正常腮腺及罹患良、恶性肿瘤腮腺的灰度值均值,临床上对腮腺良、恶性肿瘤的鉴别有一定指导意义,为腮腺CT图像和其它浓淡图像和彩色图像的定量诊断找到了方法。

Objective:By making quantitative chromatics analyses of the parotid gland CT image,to explore a method of quantitative diagnosis of parotid gland CT images( not only black and white images but also colorful ones ) . Methods:60 CT images of pathologically diagnosed benign and malignant tumor of unilateral parotid gland were selected,with 30 cases each group. Pseudo-color processing was made to both diseased and healthy sides by means of Photoshop Extended,and the grey scales value of black and white images were measured with histogram. Different from the CT values, the grey scales value belongs to the grey values of the computerized image ranging from 0 to 255,with 256 grey levels as total. At the same time,the grey values of both the diseased and healthy sides of parotid gland were compared by means of SPSS 17 . 0 . Result:After pseudo-color processing to both diseased and healthy.