摘要
目的利用螺旋CT及MPR、VRT、SCTA对中央型肺癌(CLC)行术前分期并与横断面图像对比,探讨其临床应用价值。方法46例CLC行前瞻性的CT检查,取得气管支气管树及纵隔、肺门大血管MPR和VRT图像。盲法阅片,影像诊断结果与手术及病理结果进行了比较。结果横断面图像和后处理图像对于T分类的准确性分别为82.6％(38/46)和95.7％(44/46),χ2=4.039,P=0.044,差异有显著性。N分类中,除4例N3未证实外,其余42例的敏感性为75.0％,特异性为26.9％,准确性为45.2％。横断面分期与病理分期的符合率为58.7％(27/46),后处理图像与病理分期的符合率为67.4％(31/46),χ2=0.746,P=0.338,差异无显著性。结论 图像后处理技术在中心型肺癌T分类中与病理的符合率显著高于横断面,对淋巴结的定位也有一定意义。
I Abstract] Objective To assess image processing techniques including multiplanar reconstruction (MPR), volume rendering technique (VRT) and spiral CT angiography (SCTA) in the staging of central lung cancer (CLC), with transverse image, pathology and findings on the operation table as comparison. Methods Forty-six suspected CLC received spiral CT examination, on basis of prospective randomization, by a Picker 6 000 scanner. Contrast medium was injected at a rate of 4 ml/s with a power injector. The study volume done in a single breath-hold was from the aortic arch to the inferior pulmonary veins, using 3-mm collimation, pitch of 1.0 and reconstruction at 1-mm interval. Six patients were excluded after operation. Transverse CT scans, MPR and VR images were studied in a double-blind way by 3 experienced radiologists, with the findings compared with the surgical and pathological results. Results Post-processing image was more accurate than transverse CT scan in T categories (95.7% vs82.6%), X2 = 4.039, P = 0.044. Four cases of N3 category on transverse CT scans were not proved by surgical or pathological results. The sensitivity, specificity and accuracy rates in N categories were 75.0% (12/16), 26.9% (7/26) and 45.2% (19/42). The accuracy rate of staging was 58.7% (27/46) by transverse CTscansand67 .4% (31/4 6) by post -processing images ( X2 = 0.746, P = 0.338) . Conclusion Post-processing image is more accurate than transverse image in T categorizing.
出处
《中华肿瘤杂志》
CAS
CSCD
北大核心
2003年第1期74-77,共4页
Chinese Journal of Oncology
基金
国家"九五"重点科技(攻关)资助项目(96-920-06-05-4)
关键词
肺癌
分期
螺旋CT检查
图像后处理技术
横断面图像
Lung neoplasms/radiography
Tomography, X-ray computed
I mage processing technique
