Artikel
Investigation of an image-based tool for planning safe trajectories in deep-brain-stimulation
Suche in Medline nach
Autoren
Veröffentlicht: | 4. Juni 2012 |
---|
Gliederung
Text
Objective: Planning of trajectories for deep brain stimulation (DBS) is nowadays based on MRI-data and adapted to the individual anatomy. In some cases planning is very easy to perform, while in others it can be very complicated and time-consuming. The surgeon has to identify a safe trajectory, never knowing whether there is a better one, which he was not able to identify. While MRI data consists of gray-values, risk-structers are identified by aberrations differing from the gray-matter. Aim of this study is to fasten the procedure of planning by giving the surgeon computer-generated coordinates for potential safe trajectories.
Methods: MRI-datasets of 10 DBS-Implantations were exported from Framelink5®. An algorithm was implemented using C++ on MeVisLab® platform for creating multiple possible trajectories for electrode-placement as vectors in a 3D dataset of a T1-weighted MRI acquired for DBS-procedure. A safety-region of 5 mm was set around each trajectory. For that the algorithm was supplied with the target-point and the entry-point of the clinical DBS-procedure, based on AC-PC coordinates. Multiple trajectories were virtually created and evaluated by the algorithm, by analyzing the grayscale-values along each trajectory. Values excessing or underrunning a specified range were defined as unsafe, a ranking was generated containing all tested coordinates. The five trajectories determined as safest in each dataset were reimported to Framelink5® and a blinded evaluation was performed by a neurosurgeon experienced in DBS, comparing them with the surgical-trajectory.
Results: The automatic calculation and ranking of trajectories in our implementation took less then 3 seconds (measured on an Intel Core i5-750 CPU) Ranking of the trajectories showed that in all evaluated datasets the clinical trajectory was among the first ten software-generated trajectories in ranking. Blinded evaluation of the safest trajectories generated by software led to a different “optimal” trajectory in 9 of 10 cases compared with former surgery. All of the best 5 potential trajectories generated and evaluated by the algorithm were identified as safe alternative by the surgeon.
Conclusions: We present a tool for helping surgeons planning trajectories for DBS. Computer-generated trajectories created by our algorithm seem to be safe and planning time can be impressingly decreased.