利用CT支气管定量分析技术诊断慢性阻塞性肺疾病的应用研究

The use of bronchial analysis technology based on MSCT in diagnosing chronic obstructive pulmonary disease

目的评价利用基于CT的气管定量分析技术来诊断慢性阻塞性肺疾病(chronic obstructive pulmonary disease,COPD)的价值。方法收集COPD患者35例和健康受试者30例,均在3天内行小宝石CT(GE Optima CT660)64排多层螺旋CT检查,在与所测支气管垂直的平面上测量气管与主要13个支气管的管腔及气道平均直径,气管与支气管管壁面积比值,管壁平均厚度,管壁厚度与气道管径比5个指标;使用T检验和逐步分析法对变量进行筛选,并建立Bayes预测模型,所有COPD患者及志愿者均接受肺功能检查。结果 COPD患者与正常患者中间段支气管的管壁厚度与气道直径比、右肺下叶支气管主干平均管壁厚度、右主支气管管壁面积比、左肺舌叶上段管壁平均厚度、右肺下叶支气管主干管壁厚度与气道直径比分别为:0.54,0.18,0.25,1.68,2.99;0.56,0.14,0.23,1.66,3.5,这5个变量之间差异有统计学意义(P<0.05)。利用该5个变量建立Bayes模型,回代验证预测准确率为82.04%,交叉验证准确率为75.48%。结论基于CT的支气管定量分析技术具有无创、无痛苦的优点,对诊断COPD有一定意义。

Objective To evaluate the value of using bronchial analysis technology in diagnosing the chronic obstructive pulmonary disease. Methods Sixty-five consecutive cases （35 COPD cases and 30 normal people） were all accepted GE Optimatm CT 600 exams. The data of images were then processed in work station of adw 4.5 rt to get the lumen average diameter, airways average diameter, wall area ratio, wallthickness average, wallthickness/ADiam of trachea and 13 main bronchials. T test and stepwise analysis were used to filter the 65 variables to get the ones which had statistic difference between COPD patients and normal people. Then, the filtered variables were used to build Bayes models. All COPD pa- tients accepted Lung function examination. Results The wallthickness/ADiam of middles bronchial, the wallthickness avg of right lower bronchial, the wall area ratio of right main bronchial, the wallthicknessavg of left tongue lobe bronchial and the wallthickness/ADiam of right lower lobe were 0.54, 0.18, 0.25, 1.68, 2.99, 0.56, 0.14, 0.23, 1.66, 3.5 respec- tively. There was statistically significant difference between COPD patients and normal people （ P d0.05）. The predic- tion accuracy of resubstitution of Bayes model built with above 5 variables was 82. 04% and the accuracy of cross validation was 75.48%. Conclusion The bronchial analysis technology is useful for diagnosing the COPD.