当前位置:主页 > 医学论文 > 肿瘤论文 >

孤立性肺结节CT筛查单一征象及组合征象在疾病诊断过程中的作用及地位分析

发布时间:2018-06-03 18:06

  本文选题:孤立性肺结节(SPN) + 多层螺旋CT ; 参考:《新疆医科大学》2017年硕士论文


【摘要】:目的:探讨多层螺旋CT筛查的影像学单一特征及组合特征在孤立性肺结(SPN)诊断过程中的作用。方法:选取经临床手术切除或穿刺活检等获得病理证实的SPN患者397例作为研究对象,对其临床资料进行回顾性分析。采用Philips 64排螺旋CT机对患者从胸廓入口水平到肋膈角层面,包括所有肺实质进行扫描,统计分析影像学平扫特征,筛选出对孤立性肺结节良恶性鉴别诊断具有统计学意义的单个特征及组合特征,评判各自的诊断效能。结果:397例SPN患者行多层螺旋CT平扫,以经临床手术切除或穿刺活检等获得的病理结果为诊断标准,差异有统计学意义的单个特征有6种,分别为分叶征(P=0.000)、毛刺征(P=0.000)、胸膜凹陷征(P=0.000)、边缘不光滑(P=0.000)、血管集束征(P=0.000)及结节大小(2cm≤D3cm)(P=0.000),诊断效能最高的前三种分别是分叶征、毛刺征和胸膜凹陷征。其中敏感性和阳性预测值都最高的特征为胸膜凹陷征。多种特征结对出现的可能性较单一特征单独出现大幅降低,而在恶性SPN中,同时呈现分叶征、短毛刺征及胸膜凹陷征三种现象的概率最低。结论:多层螺旋CT筛查在对SPN的良恶性鉴别诊断中具有较高的应用价值,它对大多数(准确性58.7%-77.8%)SPN的良恶性能做出准确诊断。单一和组合征象对于SPN的诊断效能无明显统计学意义,但影像学图像具有的典型特征愈多,其确诊SPN恶性病变的可信度和阳性预测值愈高。
[Abstract]:Objective: to investigate the role of single and combined features of multislice spiral CT in the diagnosis of solitary pulmonary nodule (SPN). Methods: the clinical data of 397 patients with SPN confirmed by clinical operation or biopsy were retrospectively analyzed. Philips 64-slice spiral CT was used to scan the patients from the level of thoracic entrance to the angle of costal diaphragm, including all pulmonary parenchyma. The single and combined features of the differential diagnosis of benign and malignant solitary pulmonary nodules were selected and their diagnostic efficacy was evaluated. Results one hundred and ninety-seven patients with SPN underwent multislice spiral CT scan. The diagnostic criteria were based on the pathological results obtained by clinical surgical resection or puncture biopsy. There were 6 single features with statistical significance. The results were as follows: lobulation sign (P = 0.000), burr sign (P = 0.000), pleural depression (n = 0.000), non-smooth margin (P ~ (0.000), vascular bundle sign (n = 0.000) and nodule size (n = 2cm 鈮,

本文编号:1973701

资料下载
论文发表

本文链接:https://www.wllwen.com/yixuelunwen/zlx/1973701.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户1218a***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com