Artikel
Late postoperative CT for exact localisation of electrodes in DBS
Exakte Lokalisation von Elektrodenpositionen durch späte postoperative CT bei tiefer Hirnstimulation
Suche in Medline nach
Autoren
Veröffentlicht: | 11. April 2007 |
---|
Gliederung
Text
Objective: We describe a refined CT-based method for postoperative localisation of electrodes. Electrode positions have been delineated by direct visualisation of MR and intraoperative stereotactic X-Ray. However MRI with implanted electrodes carries a risk for patients that is difficult to assess at present. As electrodes create a relatively large artefact on MRI-pictures a submillimeter definition of positions is not possible.
Methods: Between February 2004 and August 2006 63 electrodes in 32 patients have been implanted. Electrode depth was controlled intraoperatively by flouroscopy using a central beam projection. Postoperative CT-images were fused with preoperative images. On the planning workstation (SPS, Elekta) the x- and y-coordinates were determined on the CT and transferred to preoperatively acquired T2 weighted MRI-images. To avoid displacement by brainshift due to air entering the skull late CT scans (5 - 7 days) were obtained in the most recent patients. The exact localisation of electrodes in the STN was defined in an submillimeter scale.
Results: 29 electrodes in 15 patients were available for evaluation (M. Parkinson 14, ET 1). 34 electrode implants had to be excluded due to lack of postoperative CT or information on exact intended positions. The error of localisation due to imperfections of the image fusion could be estimated by visual control to be less than 1 mm. Mean deviation from the planned position was 1.51 mm and 1.93 mm for the first and second operated side respectively. By choosing a near bone window the area of electrode contacts could be differentiated from the proximal part of the electrode. In the half sitting position used in our operations deviations in the anterio-posterior direction were greater than in the lateral direction.
Conclusions: The method described showed to be simple, reliable and accurate. If errors of imagefusion are diligently controlled accuracy in the submillimeterscale can be accomplished. The method therefore appears to be superior to MRI allowing definition of elctrode positions within the STN and in relation of surrounding structures. CT-scans acquired late after resorption of intracranial air are not prone to errors caused by brainshift. In our series large deviations were attributable to intraoperative brainshift or technical difficulties namely in the interface between Leksell-arc and microdrive (FHC). An intraoperative stereotactic X-ray unit is not necessary as flouroscopy in the central beam is not influenced by distortion or deviation by intersecting lines.