基于T2WI图像的纹理分析对鼻咽癌与鼻咽部炎性增生性疾病的鉴别诊断价值

The value of texture analysis of T2WI in differential diagnosis between nasopharyngeal carcinoma and nasopharyngeal hyperplasia

目的:探讨磁共振T2WI图像纹理分析鉴别鼻咽癌(nasopharyngeal carcinoma,NPC)与鼻咽部炎性增生(nasopharyngeal hyperplasia,NPH)的价值。方法:收集经手术及病理学检查证实为NPC和NPH的患者共48例,均经过磁共振成像(magnetic resonance imaging,MRI)检查,其中NPC 30例,NPH 18例。应用MaZda软件对两组病灶进行纹理分析,在T2WI图像上选取病灶范围最大层面勾画感兴趣区(region of interest,ROI),测得294组参数,进行统计学分析。最后,利用多参数联合鉴别NPC和NPH。结果:两组间S(0,5)熵、S(5,5)熵、S(5,-5)熵、45°方向游程长不均匀度、135°方向游程长不均匀度、梯度偏度及小波低高频转换系数s-3等7组参数差异有统计学意义(P<0.05),其中135°方向游程长不均匀度、梯度偏度及小波低高频转换系数s-3这3组参数曲线下面积(area under curve,AUC)分别为0.759、0.803及0.731,诊断准确率中等。另外,分析发现,135°方向游程长不均匀度联合梯度偏度、梯度偏度联合小波低高频转换系数s-3鉴别两组病变的诊断效能分别为0.824和0.833,诊断准确率较单一参数高。结论:基于磁共振T2WI序列的纹理分析可用于鉴别NPC和NPH性疾病。

Objective: To evaluate the value of texture analysis derived from T2 WI in differential diagnosis between nasopharyngeal carcinoma(NPC) and nasopharyngeal hyperplasia(NPH). Methods: Forty-eight cases of NPC and NPH confirmed by operation and pathology were analyzed retrospectively. All of them were examined by magnetic resonance imaging(MRI). Thirty cases were NPC and 18 cases were NPH. Two groups of lesions were analyzed by MaZda software. A total of groups of parameters were obtained by selecting the largest layer of the lesion on the T2 WI image and measuring the parameters of the region of interest(ROI). Finally, multi-parameter combination were used to identify NPC and NPH. Results: The seven parameters of S(0,5) Entropy, S(5,5) Entropy, S(5,-5) Entropy, 45 dgr_RLNonUni, 135 dr_RLNonUni, GrSkewness and WavEnLH_s-3 were of statistical significance between the two groups(P<0.05), among which, the area under curve(AUC) of 135 dr_RLNonUni, GrSkewness and WavEnLH_s-3 were 0.759, 0.803 and 0.731, respectively, and the diagnostic accuracy was at a moderate level. In addition, the twoparameter combined analysis of this study found that the diagnostic efficacy of 135 dr_RLNonUni + GrSkewness and GrSkewness + WavEnLH_s-3 in the identification of two groups of lesions were 0.824 and 0.833, respectively, and contrast to the single parameter, the accuracy of combined analysis of two parameters was better. Conclusion: Texture analysis of T2 WI can be used to identify NPC and NPH.