Artikel
Virtual Endoscopy of the Paranasal Sinuses with Haptic Force-Feedback for Operation Planning
Suche in Medline nach
Autoren
Veröffentlicht: | 8. August 2007 |
---|
Gliederung
Text
Introduction: A high rate of complications in endoscopic surgical interventions of the paranasal sinuses because of missing landmarks or malignant tumors requires an exact operation planning. Beside navigation a patient´ individual preparation with the help of an intuitive virtual endoscopy with force-feedback devices may reduce the rate of complication.
Material and Methods: We developed a prototype to make a visualization of the endoscopic view, orthogonal slices and a three-dimensional overview on the base of direct volume rendering possible. The basic principles are modules for image processing and visualization for an exact illustration and generation of the haptic feedback with low computation effort during runtime. The haptic devices guide the user through the data. The PHANToM with and without force-feedback and the spacemouse have been compared and were evaluated by 6 subjects
Results: The visualization of the surface, that depends on the data resolution and is restricted because of missing information of texture and colour, is sufficient fort he planning of surgical interventions It was confirmed by the positive evaluation of the users. A support of the camerawork by haptic allows an efficient exploration of the data by avoidance of tissue collisions and is favoured in the items satisfaction, learnability, usefulness and 3-D-orientation in comparison to the spacemouse or missing force-feedback. The unusual application of the spacemouse is not intuitive and its acceptance low.
Conclusion: Previous operation training simulators are characterized by high technical and time effort and not qualified for patient´ individual preoperative planning. The method described above is because of its accuracy and speed in clinical routine preoperatively applicable. The intuitive usability requires neither previous knowledge nor experience in 3-D-environments. A time-consuming learning training is avoided. Extensive visualization options for optimizing (e.g. transparency) depend on a time-consuming post-processing and are therefore sensible only in particular cases.
Figure 1 [Fig. 1]
Figure 2 [Fig. 2]