论文部分内容阅读
目的:分析良恶性孤立性肺结节的不同CT征象,并探讨各种不同征象及联合多种征象在肺部孤立性结节定性诊断中的价值,以期提高对其的诊断符合率。方法:回顾性分析105例经手术病理和临床随访证实的孤立性肺结节的相关临床及CT资料(包括后处理重建资料)。结果:分叶征、短毛刺征、血管集束征、胸膜凹陷征、空泡或空气支气管征在恶性结节组中的出现率依次为72.4%、65.5%、51.7%、53.4%、36.2%,明显高于良性结节组的25.5%、8.5%、8.5%、23.4%、10.6%,差异具有统计学意义(P<0.05),而上述征象在细支气管肺泡癌组与非细支气管肺癌泡组中的差异无统计学意义(P>0.05),联合上述2种或2种以上征象诊断恶性结节的敏感性(81.0%)、特异性(80.9%)及准确性(81.0%)较高,而假阳性率(19.2%)和假阴性率(19.0%)相对较低。良性结节多边缘光滑(74.5%),长毛刺的出现率(29.8%)较高,结节内钙化的发生率(21.3%)等与恶性结节组相比较,差异均具有统计学意义(P<0.05)。结论:联合分叶征、短毛刺征、血管集束征、胸膜凹陷征、空泡或空气支气管征中2种或2种以上的征象并结合强化特征等对恶性结节的诊断准确性较高。
Objective: To analyze the different CT features of benign and malignant solitary pulmonary nodules, and to explore the value of different signs and multiple signs in the qualitative diagnosis of solitary pulmonary nodules in order to improve their diagnostic accuracy. Methods: Retrospective analysis of 105 cases of solitary pulmonary nodules confirmed by surgical pathology and clinical follow-up clinical and CT data (including post-processing reconstruction data). Results: The incidence of lobulation, short spiculation, vascular bundles, pleural indentation, vacuolization or air bronchial signs in malignant nodules were 72.4%, 65.5%, 51.7%, 53.4%, 36.2% Which was significantly higher than that of benign nodules (25.5%, 8.5%, 8.5%, 23.4%, 10.6%, P <0.05), but the above signs were significantly different between the bronchioloalveolar carcinoma group and the non-bronchioloalveolar carcinoma group (81.0%), specificity (80.9%) and accuracy (81.0%) of malignant nodules combined with the above two or more signs were higher than those of the control group (P0.05) The false positive rate (19.2%) and false negative rate (19.0%) were relatively low. The benign nodules had more smooth edges (74.5%), longer burr (29.8%) and nodular calcification (21.3%) than malignant nodules P <0.05). Conclusion: The diagnosis of malignant nodules is more accurate with the combination of lobulation, short spiculation, vascular bundling, pleural indentation, vacuolization or air bronchial signs in combination with enhanced features.