[Virtual endoscopy with post-processing helical CT data sets].

PURPOSE

The purpose of this work was to test a newly developed, post-processing software for virtual CT endoscopic methods. Virtual endoscopic images were generated from helical CT data sets in the region of the shoulder joint (n = 2), the tracheobronchial system (n = 3), the nasal sinuses (n = 2), the colon (n = 2), and the common carotid artery n = 1). Software developed specifically for virtual endoscopy ("Navigator") was used which, after a previous threshold value selection, makes the reconstruction of internal body surfaces possible by an automatic segmentation process. We have evaluated the usage of the software, the reconstruction time for individual images and sequences of images as well as the quality of the reconstruction. All pathological findings of the virtual endoscopy were confirmed by surgery.

RESULTS

The post-processing program is easy to use and provides virtual endoscopic images within 50 seconds. Depending of the extent of the data set, virtual tracheobronchoscopy as a cine loop sequence required about 15 minutes. Through use of the threshold value-dependent surface reconstruction the demands on the computer configuration are limited; however, this also created quality problems in image calculation as a consequence of the accompanying loss of data.

CONCLUSIONS

The Navigator software enables the calculation of virtual endoscopic models with only moderate demands on the hardware.