Evaluation of augmented-reality based navigation for brain tumor surgery.

To date, several researchers have introduced augmented reality navigation (ARN) into neurological surgery. While its application in brain tumor surgery seems promising, reports on its utility have been limited, thus warranting further evaluation. To clarify the stages and approaches in which ARN is useful and assess the effect of presurgical discussion with surgeons, we assessed usefulness using a hand-held ARN system we had developed, which displays three-dimensional (3D) virtual structures overlaid on a real-time image of the surgical field via a tablet PC monitor. The system was tested in 20 patients undergoing various procedures, with the first 10 consecutive cases being unselected and the following 10 cases being selected, for whom 3D models were prepared per the surgeons' request. Thereafter, the surgeons ranked its usefulness during each stage of surgery. Consequently, case selection and presurgical discussions with surgeons considerably improved the usefulness, with the "useful" gradings improving from 50% to 88% across all surgical stages. Accordingly, usefulness improved from 50% to 90%, 67% to 100%, and 40% to 80% during the skin incision and craniotomy, dura incision, and intradural procedure stages, respectively. ARN was useful for superficial tumor resection, but less so for deep-seated tumor resection, except when using the transcortical and interhemispheric approaches. In conclusion, a tablet-type ARN can be useful during skin incisions, craniotomy and dura incisions, superficial tumor resections, and transcortical and interhemispheric approaches for deep-seated tumors. Case selection and presurgical discussions with surgeons were essential for the efficacy of ARN.