纹理特征分析在肺部磨玻璃结节随访中的应用

发布时间：2018-04-02 05:14

本文选题：结节病　切入点：肺　出处：《中国医学影像学杂志》2017年06期

【摘要】：目的目前临床上主要采用形态学观察及CT值测量的方法对磨玻璃密度结节(GGN)进行评估,缺少有效的图像特征量化评估手段。本研究采用纹理特征分析方法对2年内的GGN进行随访量化分析,以明确GGN合理的随访时间。资料与方法回顾性分析100例GGN随访患者基线及随访3个月(A组)、随访6~12个月(B组)、随访2年(C组)的高分辨CT图像,以每组首次发现的GGN图像为基线,分析其GGN纹理特征(能量、对比度、自相关、逆差矩、熵)。结果随访3个月1个结节缩小;随访6~12个月1个结节增大,1个结节缩小;随访2年4个结节增大,其中2例出现密度差异。剩余结节形态、密度及大小均无明显变化。A组、B组及C组之间纹理特征(能量、对比度、自相关、逆差矩、熵)差异无统计学意义(P0.05)。结论纹理特征分析可以量化评估GGN的属性特征改变,作为GGN随访的量化工具,指导患者选择合理的随访方式。
[Abstract]:Objective to evaluate the GGNs of Grinded Glass density nodules (GGNs) by morphological observation and CT value measurement. In this study, the method of texture feature analysis was used to quantify the GGN in 2 years. Data and methods High resolution CT images of 100 patients with GGN were analyzed retrospectively, including baseline and 3 months follow-up in group A, follow-up of 6 to 12 months in group B and follow-up of 2 years in group C). The GGN texture features (energy, contrast, autocorrelation, deficit moment, entropy) of each group were analyzed based on the first GGN images found in each group. Results 1 node shrank in 3 months, 1 node was enlarged and 1 node shrank in 6 to 12 months of follow-up. There were no significant changes in the morphology, density and size of the remaining nodules. The texture features (energy, contrast, autocorrelation, deficit moments) between group A and group C were observed. Conclusion texture feature analysis can be used to quantitatively evaluate the change of attribute features of GGN, which can be used as a quantitative tool for GGN follow-up and guide patients to choose a reasonable follow-up method.
【作者单位】：解放军总医院海南分院放射科;解放军总医院放射科;
【分类号】：R563;R816.41

【参考文献】