肺内炎性肿块与周围型肺癌的CT鉴别诊断

目的 分析肺内炎性肿块和周围型肺癌的影像特征，探讨确定肺内肿块性质的方法，方法 搜集28例肺内炎性肿块和33例周围型肺癌的CT影像资料，全部病例均由病理诊断证实。根据形态将病灶大致分为2类：典型肿块和不典型肿块，同时观察病变征象，以征象观察的结果对形态分类结论进行修正，采用操作者特性曲线（ROC）曲线验证本方法。结果 以形态分类为标准，诊断炎性肿块的敏感度、特异度、准确度分别为64.3%、72.7%、68.9%。结合征象修正后的敏感度、特异度、准确度分别为75.0%、81.8%、78.7%。经ROC曲线分析，2位医师各自2次阅片的曲线下面积（Az）明显不同，统计学上差异有非常显著性意义（P＜0.01）。结论 采用ROC验证表明，应用以形态为基础、以征象为补充的方法判定肺内病灶的性质，可明显提高诊断的准确性。

The differential diagnosis of pulmonary inflammatory mass and peripheral lung cancer by computed tomography

Objective To analyze the CT characteristic of pulmonary inflammatory mass and peripheral lung cancer, and to look for the method of diagnosis solitary pulmonary nodule. Methods 28 patients of pulmonary inflammatory masses and 33 patients of peripheral lung cancers were included in the study,all patients conformed by pathology. The lesions were divided into typical and atypical masses according to morphology. Observing the signs of lesions and using the results ameliorate the morphological classification. Tested the method by ROC. Results According to the standard of morphology, the diagnosis sensitivity, specificity and accuracy of inflammatory mass are 64.3%, 72.7% and 68.9% respectively. The ameliorated sensitivity, specificity and accuracy are 75%, 81.8% and 78 7% respectively. Conclusion The method of making morphology as basement and sign as supplement can improve the sensitivity, specificity and accuracy of diagnosis.