读者在线:用户名 密码
首页期刊简介投稿须知期刊目录专家风采编委会特约编委特邀顾问广告发行联系我们



刊物信息more
期刊名称:《中华消化病与影像杂志(电子版)》
主管单位:中华人民共和国国家
          卫生健康委员会
主办单位:中华医学会
总编辑:王宝成   孙钢
编辑部主任:刘晓峰
 
网    址:http://www.zhxhyyx.org
地    址:山东济南市
      《中华消化病与影像杂志( 电子版)》编辑部
邮政编码: 250031

电子邮箱:zhxhbyyxzz@126.com
国际标准刊号: ISSN 2095-2015
国内统一刊号: CN 11-9312/R
访问统计
您是第 1955191位浏览者
  您当前的位置:首页 >> 正文

基于CT增强扫描图像的灰度直方图纹理分析评估胰腺癌分化程度 [中文引用][英文引用]

Feasibility study of evaluation of malignant degree of pancreatic cancer by grey histogram texture analysis based on CT contrast enhancement 

作者(英文): 
单位(英文): 
关键词(英文):Pancreatic cancer  Grade  Grey histogram  Texture analysis  CT contrast enhancement 
分类号:
出版年·卷·期(页码):2019·9·第1期(08-13)
DOI: DOI: 10.3877/cma.j.issn.2095-2015.2019.01.003
-----摘要:-------------------------------------------------------------------------------------------

 【摘要】 目的:探究CT增强图像灰度直方图纹理分析技术在术前评估胰腺癌恶性程度的可行性,评价灰度直方图参数和胰腺癌病理分级、分化程度的关系。 方法:回顾性分析扬州大学附属苏北人民医院术后病理证实的胰腺癌患者49例,其中高分化组27例,中-低分化组22例。选取2组肿瘤CT增强扫描静脉期横断位图像最大层面,采用Mazda软件勾画感兴趣区,并进行灰度直方图纹理分析,对高分化、中-低分化两组直方图特征参数进行统计分析。对有统计学意义的参数建立受试者工作特征曲线(receiver operating characterist,ROC),最终对各参数其诊断效能进行比较分析。此外,对差异有统计学意义的参数与高分化、中-低分化组别之间的相关性进行检测。 结果:采用灰度直方图分析提取的9个参数特征中,其中均值(Mean)、第1百分位数(Perc.01%)、第10百分位数(Perc.10%)、第50百分位数(Perc.50%)在高分化、中-低分化2组的差异有统计学意义(P<0.05),变异度(Variance)、偏度(Skewness)、峰度(Kurtosis)、第90百分位数(Perc.90%)、第99百分位数(Perc.99%)在2组中差异无统计学意义(P>0.05)。采用Spearman相关性分析可知均值、Perc.01%、Perc.10%、Perc.50%与分化程度之间呈正相关(r值分别为 0.04、 0.29 、0.32、 0.33,P均<0.05)。评估胰腺癌恶性程度时,均值的诊断效能最高(AUC=0.695);当均值取值147.27时,其对应的敏感度和特异度分别为86.4%、44.4%;当Perc.01%取值117时,其对应的敏感度和特异度分别为95.5%、40.7%;当Perc.10%取值120时,其对应的敏感度和特异度分别为72.7%、63%;当Perc.50%取值146时,其对应的敏感度和特异度分别为86.4%、44.4%。 结论:CT增强图像灰度直方图分析的特征参数在高分化、中-低分化胰腺癌之间存在差异,给术前评估胰腺癌恶性程度提供了新的方法。

 

-----英文摘要:---------------------------------------------------------------------------------------

 【Abstract】  Objective  To explore the feasibility of grey histogram texture analysis based on CT contrast enhancement in preoperative evaluation of the degree of malignancy of pancreatic cancer, and to evaluate the relationship between grey histogram parameters and pathological grade and differentiated degree of pancreatic cancer. Methods  A total of 49 patients with postoperative pathology proven pancreatic cancer including high differentiation (27 cases) and low-moderate differentiation (22 cases) were analyzed retrospectively. Region of interest (ROI) was chosen at axial CT images with maximum enhancement of lesion and grey histogram analysis was performed using Mazda software. The histogram parameters were analyzed statistically in these two groups. Receiver operating characteristic (ROC) curve was established for parameters to compare diagnostic performance. In addition, the correlation was tested between parameters and groups. Results  In the 9 parameters extracted from grey histogram analysis among the two groups, the mean value, Perc.01%, Perc.10%, Perc.50% had statistical differences (P<0.05), but variance, skewness, kurtosis, Perc.90%, Perc.99% had not statistical differences (P>0.05). Spearman correlation analysis showed that the mean value, Perc.01%, Perc.10%, Perc.50% were positively correlated with the differentiated degree (r values were 0.04, 0.29, 0.32 and 0.33 respectively, P<0.05). When evaluating malignant grade of pancreatic cancer, the diagnostic efficiency of the mean value was the highest (AUC=0.695). When the cutoff value was 147.27 for mean value, the sensitivity and specificity were 86.4% and 44.4%. When the cutoff value was 117 for Perc.01%, the sensitivity and specificity were 95.5% and 40.7%. When the cutoff value was 120 for Perc.10%, the sensitivity and specificity were 72.7% and 63%. When the cutoff value was 146 for Perc.50%, the sensitivity and specificity were 86.4% and 44.4%. Conclusion  The characteristic parameters of gray histogram texture analysis based on CT enhanced images are different between high and low-moderate grade pancreatic cancer, providing a new method for preoperative evaluation of the degree of malignancy of pancreatic cancer.

 

-----参考文献:---------------------------------------------------------------------------------------

欢迎阅读《中华消化病与影像杂志(电子版)》!您是该文第 179 位读者!

若需在您的论文中引用此文,请按以下格式著录参考文献:
中文著录格式: 沈力,征锦,叶靖,朱庆强,徐圆.基于CT增强扫描图像的灰度直方图纹理分析评估胰腺癌分化程度.中华消化病与影像杂志(电子版).2019;9(1):08-13.
英文著录格式: .Feasibility study of evaluation of malignant degree of pancreatic cancer by grey histogram texture analysis based on CT contrast enhancement .No Title Settings.2019;9(1):08-13.

与该文相关的文章(仅限于本刊内

已投本刊未发表相似文章

地址:山东济南师范路25号济南军区总医院《中华消化病与影像杂志(电子版)》编辑部;

邮政编码:250031;电话:0531-51665254 投稿邮箱:zhxhbyyxzz@126.com   

本系统由北京博思汇文数字科技有限公司设计开发 技术服务电话:400-921-9838