论文部分内容阅读
目的 利用高分辨率CT(HRCT)分析肺结节样病灶的影像学特征 ,结合正向概率判别法探讨对肺结节鉴别诊断有价值的影像学征象。方法 搜集经手术病理证实的 1 2 0例肺结节 ,其中周围型肺癌、炎性结节、结核球及肺内良性肿瘤各 30例 ,所有的患者均于术前行常规平扫和增强扫描 ,并结合HRCT扫描。采用正向概率判别法对结节病灶常见的CT征象进行分析总结 ,并转化为概率值后判定结节的性质 ,为鉴别诊断提供帮助。结果 正向概率判别法对周围型小肺癌、炎性结节和结核球的总体诊断正确率 (78 9%)较常规阅片法 (6 8 9%)为高 ,两者比较差异有统计学意义 (χ2 =4 2 7,P <0 0 5 )。结论 在HRCT扫描的基础上对肺结节样病灶进行概率判别分析法有助于提高诊断正确率。
Objective To analyze the imaging characteristics of pulmonary nodules by high resolution CT (HRCT) and to explore the valuable imaging features of differential diagnosis of pulmonary nodules by combining the positive probability discriminant method. Methods Totally 120 pulmonary nodules confirmed by pathology were collected, including 30 cases of peripheral lung cancer, inflammatory nodules, tuberculoma and benign lung tumors. All patients underwent routine and enhanced scan before operation , Combined with HRCT scan. The positive probability discriminant method is used to analyze and summarize the common CT signs of nodules and to determine the nodules’ properties after transforming into the probability value, which can be helpful for differential diagnosis. Results The positive diagnostic accuracy rate (78 9%) of the peripheral small lung cancer, inflammatory nodules and tuberculoma was higher than that of the conventional radiography (68.9%), the difference was statistically significant Significance (χ2 = 4 2 7, P <0 0 5). Conclusions Probability discriminant analysis of pulmonary nodules on the basis of HRCT scan can help improve the diagnostic accuracy.