【摘要】 目的　探讨胃和结肠CT仿真内窥镜 (CTvirtualendoscopy ,CTVE)的成像方法及技巧。方法　使用HispeedFX/i扫描机 ,扫描层厚 3～ 5mm ,螺距 1～ 2 ,重叠重建率 5 0 %～ 67%。利用NavigatorSmooth软件对 48例 (包括 4例正常志愿者 )容积扫描数据进行三维重建 ,获得仿真胃、结肠内窥镜的影像 ,并与纤维内窥镜和 /或手术所见对照分析 ,全部病例经病理证实。结果　通过选择不同的观察方位、角度和阈值 ,CTVE能清楚显示靶器官内正常解剖结构的立体影像和病变的部位、大小、形态及其侵犯范围。CTVE对胃、结肠病变形态的显示与纤维内窥镜 (fiberopticendoscopy ,FE)或 /或标本显示具有良好的对应性 ,CTVE显示病变的准确性为 95 .6%。结论　高质量的胃、结肠CTVE图像依赖于扫描前准备、扫描及图像重建参数的合理选择和匹配 ,选择合适的观察方位和阈值 ,以及操作者的技巧和熟练程度更多还原

【Abstract】 Objective To investigate the methods and techniques of CT virtual endoscopy(CTVE) of stomach and colon.Methods 44 patients proved pathologically and 4 normal volunteers were detected by volume scanning using helical CT(Hispeed Fx/i,GE).The CT scan parameters were:collimating width 3～5 mm,pitch 1～2,reconstructed in overlapping rate 50%～67%.Volumetric data were analyzed by the navigator soft ware and along the longitudinal axis of the tissues using the software.Fly through.CTVE images were observed comparatively with the findings of fiberoptic endoscopy and /or surgery.Results By choosing different viewpoint location,orientation and threshold values,CTVE could clearly demonstrate the three-dimensional image of normal structures and the location,size and extension of lesions.The appearances of the inner lesions and normal wall stomach and colon on CTVE were similar to those on fiberoptic colonoscopy.The accuracy in detecting lesion by CTVE was 95.6% in 44 cases.Conclusion A high quality of CTVE image of stomach and colon is depended on selecting and matching of the scanning and reconstruction parameters properly,choosing an appropriate viewpoint direction and threshold values as well as the operator’s technique and skill.更多还原