Artikel
"TREGS" - a new tool for automated registration in neuronavigation: application accuracy
"TREGS" - ein neues Hilfsmittel zur automatischen Registrierung im Rahmen der Neuronavigation: Evaluation der Anwendungsgenauigkeit
Suche in Medline nach
Autoren
Veröffentlicht: | 4. Mai 2005 |
---|
Gliederung
Text
Objective
To make the process of referencing physical space to image space easier and faster, especially when an intraoperative MRI system is used, and to simplify intraoperative updating of navigation data, “TREGS” was developed – a system that provides an automated registration process. We compared the application accuracy of the Brainlab VectorVision® Neuronavigation system when it is used with “TREGS”-registration with its application accuracy when standard fiducial-based registration is performed.
Methods
The automatic referencing tool is based on five markers that are integrated in the MRI-compatible head rest holder we routinely use in our intraoperative MRI setting. The navigation software automatically detects the markers. A phantom was fixed in the head holder, then we acquired multiple optimized gradient echo slices, axial (thickness:220 mm), sagittal (25 mm) and coronary (18 mm), containing the clamp-integrated markers. After that we measured a T1 MPRAGE sequence with a slice thickness of 1.0 mm for navigation. We used a Plexiglas phantom with 32 notched rods of different heights. The deepest points of the surface of the rods were defined as target points in image space.In three measurement cycles we referenced the phantom once with 4, once with 7 fiducials and twice automated with “TREGS”. In one cycle we performed only one automatic registration. The localization error was measured three times per rod and registration. Thus a total number of 1440 error values was obtained.
Results
The median localization error for the standard registration with 7 fiducials was 1.34 mm in the first measurement, 3.05 mm in the second, 1.20 mm in the third and 2.36 mm in the fourth. With four fiducials it was 2.07, 2.21, 1.87, 2.07 mm. For the automatic registration we obtained median localization errors of 0.96, 0.88, 1.63 and 2.13 mm. In 6 of the 8 compared samples, the automated registration showed a highly significant better application accuracy than the fiducial-based standard registration.
Conclusions
The application accuracy found for “TREGS”-referencing is at least not worse than that for standard registration no matter whether 4 or 7 fiducial markers were used. In most series even lower error values were measured than with standard registration. “TREGS” has proved to be reliable and fast as well as easy to handle. In combination with intraoperative MRI it may become a favourable alternative to standard fiducial-based registration, especially when an intraoperative update of the navigation data is necessary.