Improvement of registration accuracy for navigatedcontrol drill in mastoidectomy (Autopilot)

Background: The goal of this study is the improvement of the surgical accuracy of a navigate-controlled drill for mastoidectomy in a lab test. Methods: For lab tests an artificial model of the temporal bone with color-coded injury identification of the facial nerve (solution of 0.5 mm) was used. Two different registration methods were examined: (group 1) navigation bow with 4 integrated markers at the upper jaw; (group 2) landmark registration with 4 titanium micro screws. An optical navigation system was used. The targets were illustrated by 3 titanium screws within the range of the planum mastoideum. The accuracy of the navigate-controlled drill in drilling the planned cavity were evaluated at 20 temporal bone models. The measurement of the registration accuracy was evaluated by deviation between the target screw and the calculated position in the navigation system. The evaluation of the resulted cavities was done by 5 senior surgeons with the help of the microscope. Results: The registration accuracy shows a maximum deviation between the real position and the calculated position of 1, 73 mm in group of 1 and 0.93 mm in group 2. In group 1 the nerve was hurt in 5/20 cases and a maximum deviation of - 1.5 mm (Std 0.25 mm) (drilled beyond the nerve) was measured. In group 2 the nerve was not hurt, a maximum deviation of 0.5 mm (too early stopped before the nerve) was measured. Conclusions: Significantly better results of the registration and drilling accuracy show up in group group 2. Thus the preconditions for clinical use are fulfilled.