Accuracy Validation of Neuronavigation Comparing Headholder-Based System with Head-Mounted Array—A Cadaveric Study

摘要

BackgroundNeuronavigation is widely used for intracranial neurosurgical procedures and is commonly based on the standard reference array being fixed to the headholder. Some cases require the reference array to be attached directly to the head. The aim of this cadaveric study was to compare operational accuracy of a head-mounted reference array with the standard headholder-based system. MethodsNavigation accuracy was evaluated with 10 cadaveric specimens. Each specimen was prepared with 8 titanium microscrews that served as reference points on the external skull, and computed tomography was performed. Registration of all specimens was done using surface matching with infrared laser on three-dimensional reconstructed high-resolution computed tomography. In all 10 specimens, the head-mounted reference array and headholder-based system were compared by 10 repetitive measurements. The deviation was evaluated for each screw and compared using nonparametric Mann-WhitneyUtest between groups and screws. A Bland-Altman plot was generated for comparison. ResultsA total of 1600 measurements were conducted. Mean deviation was 1.97 mm (95% confidence interval, 1.90–2.03 mm) with the head-mounted reference array and 2.10 mm (95% confidence interval, 2.04–2.18 mm) with the headholder based system. There was no significant difference between methods in 9 of 10 specimens. In 1 specimen, the head-mounted array was superior. The deviation in either method showed a significant correlation, indicating high pertinence for registration (P< 0.001). ConclusionsNavigation with the head-mounted reference array demonstrated comparable accuracy to the headholder-based system and can be used without reduced accuracy. Careful registration is mandatory.