期刊文献+

三期CT图像纹理分析对肺良恶性结节的鉴别诊断价值 被引量:6

The value of three-phase CT images texture analysis in differential diagnosis of benign and malignant pulmonary nodules
下载PDF
导出
摘要 目的探讨基于胸部三期CT图像的纹理分析对鉴别肺内良恶性结节的价值。方法选取本院159例肺结节(恶性83例,良性76例,均有病理证实)的CT平扫、动脉期及静脉期(三期)图像进行回顾性纹理分析。手动勾画三期均为同一层面的病灶为感兴趣区并提取33个纹理特征。通过秩和检验及独立样本t检验找出有统计学差异的纹理参数。进一步分析三期图像有统计学差异纹理参数中的能量、熵、相关性、惯性并分别进行ROC曲线分析,比较三期图像的诊断效能差异。结果对于肺内良恶性结节的鉴别诊断,三期图像均得出较多具有统计学意义的纹理参数。其中相关性、惯性在动脉期图像的诊断效能较高,两者的P值均为0.000。以相关性≥0.00004、惯性≤12272作为诊断恶性结节标准时,ROC曲线下AUC值分别为0.731、0.702,敏感度分别为77.1%,85.5%。能量、熵两个参数在平扫、动脉期、静脉期三期的诊断效能无显著差异。结论胸部三期CT图像的纹理分析能够为肺内良恶性结节的鉴别诊断提供帮助,其中以动脉期图像的诊断效能较高。 Objective To explore the value of texture analysis based on three-phase CT images of the chest in distinguishing benign and malignant nodules in the lung.Methods From June 2015 to June 2019,159 cases of pulmonary nodules(83 cases of malignant and 76 cases of benign,all confirmed by pathology)were retrospectively analyzed by CT plain scan,arterial phase and venous phase(three-phase)images.We manually outlined the lesions in the three-phase images at the same level as the region of interest(ROI)and extracted 33 texture features.The rank sum test and independent sample t test were used to find out the statistically different texture parameters.We further analyzed the energy,entropy,correlation,and inertia in the texture parameters of the three-phase images with statistical differences,and performed ROC curve analysis to compare the differences in diagnostic efficiency of the three-phase images.Results For the differential diagnosis of benign and malignant nodules in the lung,the three-phase images obtained more statistically significant texture parameters.Among them,correlation and inertia had higher diagnostic efficiency in arterial phase images,and the P value of both was 0.000.When correlation≥0.00004 and inertia≤12272 were used as the criteria for diagnosing malignant nodules,the AUC values under the ROC curve were 0.731 and 0.702,and the sensitivity was 77.1%and 85.5%,respectively.The two parameters of energy and entropy had no significant difference in the diagnostic efficacy of plain scan,arterial phase,and venous phase.Conclusion The texture analysis of three-phase CT images of the chest can provide help for the clinical diagnosis of benign and malignant nodules in the lung,and the arterial phase images have higher diagnostic efficiency.
作者 陈梓盼 黄远明 罗文暄 罗树存 陈济琛 黄东玲 罗泽斌 陈晓东 CHEN Zipan;HUANG Yuanming;LUO Wenxuan;LUO Shucun;CHEN Jichen;HUANG Dongling;LUO Zebin;CHEN Xiaodong(Department of Ultrasound, Shiyan People's Hospital of Shenzhen, Baoan District, Shenzhen 518108, P.R.China;Department of Radiology, The Affiliated Hospital of Guangdong Medical University, Zhanjiang 524001, P.R.China)
出处 《医学影像学杂志》 2021年第10期1677-1681,共5页 Journal of Medical Imaging
基金 广东省湛江市科技发展计划项目(编号:2019A01026) 广东医科大学附属医院博士基金项目(编号:BJ201521)。
关键词 纹理分析 肺内良恶性病灶 体层摄影术 X线计算机 Texture analysis Benign and malignant lesions in the lung Tomography,X-ray computed
  • 相关文献

参考文献7

二级参考文献69

  • 1姚宏宇,李弼程.基于广义图像灰度共生矩阵的图像检索方法[J].计算机工程与应用,2004,40(34):98-100. 被引量:19
  • 2汪家旺,于立燕,王德杭,俞同福,舒华忠,张廉良.肺癌分形维数特征的研究[J].中国医学物理学杂志,2005,22(1):393-395. 被引量:7
  • 3朱福珍,吴斌.基于灰度共生矩阵的脂肪肝B超图像特征提取[J].中国医学影像技术,2006,22(2):287-289. 被引量:23
  • 4王惠明,史萍.图像纹理特征的提取方法[J].中国传媒大学学报(自然科学版),2006,13(1):49-52. 被引量:77
  • 5Lin Daw-yung, Yan Chung-ren, Chen Wen-tai. Autonomous detection of pulmonary nodules on CT images with a neural network-based fuzzy system [ J ]. Computerized Medical Imageing and Graphics,2005,29:447-458.
  • 6Xu X W, Doi K, Kobayashi T, et al. Development of an improved CAD scheme for automated detection of lung nodules in digital chest images [ J ]. Medical Physics, 1997,24(9) :1395-1403.
  • 7Kanazawa K, Kawata Y, Niki N, et al. Computer-aided diagnosis for pulmonary nodules based on helical CT images[J]. Computerized Medical Imaging and Graphics, 1998, 22 : 157-167.
  • 8Satoh H, Ohmatsu H, Kakinuma R,et al. Lung cancer detection based on helical CT images using curves surface morphology analysis [ J ]. Proceeding of SPIE, 1999, 3661 : 1307-1314.
  • 9Vapnik V. The nature of statistical learning theory [ M ]. New York : Springer-Verlag, 1995:91-188.
  • 10Cristianini N,Shawe-Taylor J. An introduction to support vector machines and other kernel-based learning methods [ M ]. Cambridge : Cambridge University Press, 2000 : 47-98.

共引文献96

同被引文献76

引证文献6

二级引证文献16

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部