邮件订阅 RSS 中文 | English
高级检索

3.0T磁共振T1WI对比增强图像纹理分析鉴别眼眶淋巴瘤与炎性假瘤

Differentiation of orbital lymphoma and inflammatory pseudotumor using texture analysis based on contrast-enhanced T1-weighted images of 3.0T MRI

    摘要
  • 摘要:
    目的 探讨纹理分析在鉴别眼眶淋巴瘤与炎性假瘤中的价值。
    方法 纳入经手术病理证实的眼眶淋巴瘤25例和炎性假瘤24例，使用MaZda软件提取T1WI对比增强（contrast-enhanced T1-weighted images，CE-T1W）肿瘤纹理特征。比较两组游程长矩阵（run-length matrix，RLM）的20个纹理特征参数，对差异有统计意义的参数进行主成分分析（principal component analysis，PCA）。采用多变量logistic回归模型分别对在两组间差异有统计学意义的特征参数及主成分进行建模，绘制受试者工作特征曲线（receiver operating characteristic curve，ROC）以评价模型效能。
    结果 淋巴瘤组灰度不均匀度（grey level non-uniformity，GLNU）、长游程优势（long run emphasis，LRE）的4个参数高于炎性假瘤（P < 0.001），短游程因子（short run emphasis，SRE）、游程图像分数（fraction of image in runs，FIR）的4个参数低于炎性假瘤组（P < 0.001）；游程长不均匀度（run length non-uniformity，RLNU）的4个参数在两组间差异无统计学意义（P>0.05）。在两组间差异有统计学意义的参数中，以GLNU的参数建立的多变量logistic回归模型的诊断效能最高，灵敏度、特异度及曲线下面积（area under curve，AUC）分别为95.9%、72.0%、0.905；PCA提取出的2个主成分建立的多变量logistic回归模型的灵敏度、特异度及AUC分别为87.5%、88.0%、0.913。
    结论 基于CE-T1W的RLM特征可有效鉴别眼眶淋巴瘤与炎性假瘤，其中以GLNU的诊断效能最高。

     

    Abstract:
    Objective To investigate the diagnostic value of texture analysis in differentiating orbital lymphoma from inflammatory pseudotumor.
    Methods There were 25 patients with orbital lymphoma and 24 patients with orbital inflammatory pseudotumor confirmed by postoperative pathological test. MaZda software was used to extract tumor texture features of contrast-enhanced T1-weighted images (CE-T1W). Twenty texture feature parameters of run-length matrix (RLM) between the two groups were compared. Principal component analysis (PCA) was performed for the parameters with statistical significance between the two groups. Multivariate logistic regression was used to model the features that showed statistical significance between the two groups and principal components respectively, then ROC curve was drawn to evaluate the efficiency of the models.
    Results Four parameters of grey level non-uniformity (GLNU) and long run emphasis (LRE) were higher in lymphoma than inflammatory pseudotumor (P < 0.001). Four parameters of short run emphasis (SRE) and fraction of image in runs (FIR) were lower in the lymphoma group than those in the inflammatory pseudotumor group (P < 0.001). There were no significant differences of four parameters of run length non-uniformity (RLNU) between the two groups. Among the features that showed statistically significant differences between the two groups, the model with GLNU parameters showed the highest effect with sensitivity, specificity, and area under curve (AUC) of 95.9%, 72.0%, and 0.905. The sensitivity, specificity, and AUC of the multivariate logistic regression model established with two principal components extracted by PCA, were 87.5%, 88.0%, and 0.913, respectively.
    Conclusions Texture feature parameters based on CE-T1W could effectively distinguish orbital lymphoma from orbital inflammatory pseudotumor, among which, GLNU had the highest diagnostic efficiency.

     

    • HTML全文
    • 参考文献(13)
    • 相关文章
    • 施引文献
  • 资源附件(0)

/

返回文章
返回