For this study, scientists determined the navigational accuracy of an advanced augmented reality (AR)–based guiding system for minimally invasive neurological surgery, biopsy, and other operations. The researchers drilled Five burr holes into a human skull model’s frontal, temporal, and parietal bones. 5 optical fiducials (AprilTags) printed with CT-visible ink were inserted on the frontal, material, and parietal bones. They fixed three 0.5-mm-diameter targets on nylon stakes around the level of the tentorium cerebelli and the pituitary fossa in the inside of the skull. They filled the head with ballistic gelatin to represent brain tissue. A CT scan was acquired to combine three targets and five access holes, and virtual needle tracts were labeled on the preoperative 3D workstation (15 target tracts). By viewing an encrypted, printed QR code issued to the research by the preoperative workstation, the annotated study was uploaded to and activated by VisAR software running on the HoloLens 2 holographic visor. By aligning the holographic fiducials with the AprilTags mounted to the skull, the DICOM pictures were transformed into 3D holograms and registered to the head. Five participants familiar with the VisAR utilized the software and visor to direct an 18-gauge needle/trocar through a succession of burr holes to the target, resulting in 70 data points (15 for 4 users and 10 for 1 user). After each try, the needle stayed in the skull, supported by the ballistic gelatin, and took a high-resolution CT.
About 2.30° ± 1.28° was the combined angle of error. Users had a mean radial inaccuracy of 3.62 ± 1.71 mm. The average depth of the target was 85.41 mm. The mean radial error and angle of error and the accompanying variance measures showed that VisAR navigation could be useful for guiding a small needle to brain lesions or targets with a 3.62 mm accuracy. Many neurological operations, such as ventriculostomy, can be navigated using these parameters because they were accurate.
