CT影像纹理分析对肺磨玻璃样小结节良恶性的鉴别诊断-现代医学

CT影像纹理分析对肺磨玻璃样小结节良恶性的鉴别诊断

单位：淮安市肿瘤医院 CT室, 江苏淮安 223200

分类号：R730.44;R734.2

出版年·卷·期（页码）：2021·49·第九期（1075-1079）

摘要：

目的：探讨CT影像纹理分析鉴别诊断肺磨玻璃样小结节良恶性的价值。方法：选取113例患者共115个肺磨玻璃样小结节，将CT影像导入MaZda 4.6软件中，手动勾画病变的感兴趣区域（ROI），分别通过交互信息（MI）、分类错误概率联合平均相关系数（POE+ACC）、Fisher系数及3种方法联合（Fisher+POE+ACC+MI）选择最具鉴别价值的纹理特征参数，并构建人工神经网络（ANN）模型。以病理结果作为金标准，对比分析CT影像纹理分析和影像医师评估的误判率。结果：对于肺磨玻璃样小结节，MaZda软件提取的30组纹理参数中共10组差异具有统计学意义，其中MI 2组，POE+ACC和Fisher系数均为4组。采用MI方法进行纹理分析的总误判率为45.22%（52/115），POE+ACC为34.78%（40/115），Fisher系数为37.39%（43/115），3种方法联合为14.78%（17/115），影像医师的误判率为45.22%（52/115），其中3种方法联合的误判率最低。同时，影像医师对恶性微小病变的误判率明显高于良性微小病变（P<0.05）。结论：CT影像纹理分析可用于鉴别诊断肺磨玻璃样小结节的良恶性。

Objective: To investigate the value of CT images texture analysis in differential benign and malignant small pulmonary ground-glass nodules. Methods: A total of 113 patients with 115 pulmonary ground-glass nodules were selecteded. The CT images were imported into Mazda 4.6 software and regions of interest(ROI) were manually drawed. The optimum texture parameters were selected through Mutual information (MI), Fisher coefficient, probability of classification error and average correction coefficient (POE+ACC) and the combination of three methods (Fisher+POE+ACC+MI) respectively.The model of artificial neural network was builded. The assessment results of texture analysis and radiologists were compared with the pathological results. And misdiagnostic rates were analyzed between texture analysis and radiologists. Results: For small pulmonary ground-glass nodules, there were 8 groups with statistical differences among the 30 groups of texture parameters extracted by Mazda software, MI were 2 groups, Poe+ACC and Fisher coefficient were 4 groups each. The misdiagnostic rates of MI, POE+ACC, Fisher, Fisher+POE+ACC+MI and radiologists were 45.22%(52/115),34.78%(40/115),37.39%(43/115),14.78%(17/115),45.22%(52/115) respectively. Misdiagnostic rate of Fisher+POE+ACC+MI was lowest. Meantime, the misdiagnosis rates of benign nodules was significantly higher than that of malignant nodules by radiologists (P < 0.05). Conclusion: CT images texture analysis can be used to differentiate benign and malignant small pulmonary ground-glass nodules.

参考文献：

[1] OKUYAMA A.Lung cancer incidence rates in the world from the cancer incidence in five continents Ⅺ[J]. Jpn J Clin Oncol, 2018, 48(3):300-301.
[2] KADOYA N, TANAKA S, KAJIKAWA T, et al. Homology-based radiomic features for prediction of the prognosis of lung cancer based on CT-based radiomics[J]. Med Phys, 2020, 47(5):2197-2205.
[3] WEIR-MCCALL J R, JOYCE S, CLEGG A, et al. Dynamic contrast-enhanced computed tomography for the diagnosis of solitary pulmonary nodules:a systematic review and meta-analysis[J]. Eur Radiol, 2020, 30(1):3310-3323.
[4] 赵飞, 徐剑霞, 蒋婷, 等.对混合性毛玻璃样结节和单纯性毛玻璃样结节的临床Ⅰ期非小细胞肺癌患者进行淋巴结转移率比较的Meta分析[J]. 南京医科大学学报(自然科学版), 2018, 38(2):277-282.
[5] 陈真伟, 滕晓东.2015版WHO肺肿瘤组织学分类解读[J]. 中华肿瘤防治杂志, 2016, 23(1):60-64.
[6] ESPINASSE M, CHARMETTANT B, BIDAULT F, et al. CT texture analysis challenges:influence of acquisition and reconstruction parameters:a comprehensive review[J]. Diag, 2020, 10(5):258-269.
[7] 黄栎有, 王延花, 高先聪.基于CT平扫图像纹理分析鉴别浸润性肺腺癌与非钙化结核球[J]. 中国医学影像技术, 2020, 36(4):545-549.
[8] 高先聪, 黄栎有.基于乳腺X线图像不同区域的纹理分析鉴别乳腺肿块良恶性[J]. 放射学实践, 2020, 35(8):1037-1041.
[9] 张衡, 舒政.纹理分析在甲状腺结节影像学中的研究进展[J]. 中国医学计算机成像杂志, 2018, 24(5):70-76.
[10] 王楠, 刘爱连, 李烨, 等.基于单源双能CT平扫图像的纹理分析对肝脓肿和肝转移瘤的鉴别价值[J]. 放射学实践, 2019, 34(11):76-80.
[11] 王波涛, 刘刚, 何蕾, 等.纹理特征分析在肺部磨玻璃结节随访中的应用[J]. 中国医学影像学杂志, 2017, 25(6):441-446, 451.
[12] 鄂林宁, 张娜, 王荣华, 等.计算机体层摄影术纹理分析对孤立性肺结节良恶性鉴别诊断的价值[J]. 中华肿瘤杂志, 2018, 40(11):847-850.
[13] CHEN C, OU X, LI H, et al. Contrast-enhanced CT texture analysis:a new set of predictive factors for small cell lung cancer[J]. Mol Imaging Biol, 2020, 22(3):745-751.
[14] 李传旺, 邓开盛, 吴景强.螺旋CT靶扫描及重建对肺部小结节良恶性判断的价值与限度[J]. 现代医用影像学, 2018, 27(7):13-15.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】

提示：您还未登录，请登录！点此登录