The comparative accuracy of the ROSA stereotactic robot across a wide range of clinical applications and registration techniques.

Robot-assisted stereotactic neurosurgery is an emerging technology with a growing range of applications. The ROSA system is a robotic stereotactic system that has been shown to be accurate in laboratory studies and large case series. The goal of this study was to examine the accuracy of the ROSA across different registration methods as well as different clinical applications. Sixteen patients with one hundred and seventeen stereotactic trajectories were examined. Accuracy was compared by measuring the distance between the trajectory target and the actual termination of the device as determined by imaging. Entry error and angular deviation were also measured. Variables included bone fiducials vs. laser facial scanning, the clinical indication for stereotactic surgery, and the effect of lead deflection on accuracy. Bone fiducials did not offer an accuracy benefit over laser facial scanning (mean target error 4.5-3.9 mm, p = 0.34) in these clinical scenarios. Laser interstitial thermal therapy, responsive neurostimulation, and stereo electroencephalography were equally accurate when placed by the ROSA (mean target error 4.4-4.3-4.0 mm, respectively, p = 0.69). Deflection did not affect lead accuracy (mean target error 4.4-3.9 mm, p = 0.11). Similar results are seen for entry error and angular deviation. ROSA is a highly accurate stereotactic system. Laser facial scanning provides the same accuracy as bone fiducials in these stereotactic applications. The ROSA is equally accurate across a wide spectrum of applications. The ROSA is effective at limiting lead deflection, and when it does occur, it does not impact target accuracy in a significant way.