Comparison of Multislice CT and Its Postprocessing Techniques in Examination of Trachea and Main Bronchus Neoplasms
罗明月,单鸿,徐川,胡冰,李露芳,张建生
DOI: https://doi.org/10.3969/j.issn.1001-9324.2005.06.009
2005-01-01
Abstract:Objective To comparatively study the clinical application of multislice CT (MSCT) and its postprocessing techniques in examination of trachea and main bronchus neoplasms.Materials and Methods Thin-slice and HQ mode scanning was performed with Light Speed QX/i MSCT scanner in 31 patients with suspected trachea and main bronchus neoplasms, reconstruction data were formed after retro-reconstruction of raw data. Multiplanar volume reformation (MPVR), volume rendering (VR) and virtual bronchoscopy (VB) images were obtained retrospectively by postprocessing of reconstruction data with MPVR, VR and VB software in AW workstation. The findings of MSCT initial images, MPVR, VR and VB images were compared and analyzed with surgical and pathological results, longitudinal extensions and luminal stenosed degree of trachea and main bronchus neoplasms. Results With MSCT initial images, MPVR, VR and VB images, the diagnosis correction rates of neoplasm morphologies were 70.9%, 100.0%, 74.2% and 65.5% respectively, MPVR was the best, there were significant statistical difference between MPVR and MSCT initial images, VR, VB images; the diagnosis correction rates of longitudinal extensions were 80.6%, 100.0% 100.0% and 74.2% respectively, MPVR and VR were the best, there were significant statistical difference between MPVR, VR and MSCT initial images, VB images; the diagnosis correction rates of luminal stenosed degrees were 83.9%, 100.0%, 100.0% and 80.6% respectively, MPVR and VR were the best, there were significant statistical difference between MPVR, VR and MSCT initial images, VB images.Conclusion MSCT initial images is basic; MPVR images can accurate display neoplasm morphologies, longitudinal extensions and luminal stenosed degree without three dimensional effect; VR images may three- dimensional accurate reveal longitudinal extensions and luminal stenosed degrees, while its disclosing of neoplasm morphologies isfairly poor; clinical application value of VB images is limited.