Curved trajectories in stereotactic neurosurgery: is it feasible?
摘要
Stereotactic procedures are planned and performed with straight trajectories. However, brain sulci with vessels and ventricles have to be avoided for bleeding and deviation risks. In distinct pathologies such as insular or pineal lesions, or in patients with brain atrophy, curved trajectories theoratically could provide superior results. So far, research on curved trajectories for stereotactic neurosurgery focusses on aspects like path planning, robot design and control. In a collaborative project of engineers, mathematicians, and neurosurgeons a prototype system for curved cannulas was developed using a concentric tube continuum robot (CTCR) platform.
MethodsTarget precision and follow-the-leader-deviations by movements of the cannulas were assessed. For a set of automatically planned configurations by numerical optimization, the real robot behavior was compared to state-of-the-art models of elastostatic behavior. Target accuracy was tested with CT-imaging of a head model and a deep-seated target, calculation and transformation of stereotactic coordinates to the actuation system, and application of the curved cannulas to the target.
ResultsDetermination of the model’s target point via CT-scan and transformation of the stereotactic coordinates in the path planning software for curved trajectories were possible. The technical operation of the prototype was improved for mounting at the stereotactic system. The target could be calculated with optimal accuracy in the panned configuration by numerical optimization and the test procedure could be performed successfully. However, accuracy with curved cannulas in this prototype system had a target point deviation of 2 mm and more.
ConclusionsTo the authors’ knowledge, this is the first application of curved cannulas for stereotactic neurosurgery in a comparable way to clinical practice. Further research will have to address incorporating iterative learning control of the robot’s tip position to reach target point deviations below 1 mm.