The influence of various registration procedures upon surgical accuracy during navigated controlled petrous bone surgery.

The goal of this study was to investigate the dependence of surgical accuracy with a navigated controlled (NC) drill on selected registration procedures.The target registration error of the instrument and the maximum proximity to a typical high-risk structure (facial nerve) were determined within an artificial petrous bone.The studies took place in two groups: group 1, navigation bow with six integrated markers and attachment at the upper jaw, and group 2, landmark registration with four titanium microscrews. Measurement of the target registration error took place at three targets (3 titanium screws) with 20 repeated registration procedures via evaluation of the deviation between a target and the indicated position in the navigation data.For measurement of the conversion accuracy of the planned cavity, 20 petrous bone models were milled by inexperienced test subjects. The evaluation of 20 cavities was conducted via a microscope by five jurors.Registration accuracy showed a maximum deviation between the actual position achieved and the computed position in the navigation system of 1.73 mm in group 1 and 0.93 mm in group 2. In group 1, the nerve in five of 20 cases was damaged, and a maximum penetration into the nerve of 1.5 mm (0.25 mm SD; milled beyond) was measured. In group 2, the facial nerve was not damaged at all, and a maximum deviation of 0.5 mm (0.63 mm SD; stopped before) was measured.The results for registration and conversion accuracy are significantly better for the landmark-based registration than with the registration of the patient model with registration bow on the upper jaw.