Study: Mixed reality system shows promise as alternative to optical neurosurgical navigation

Editor's Note

A mixed reality navigation (MRN) system for neurosurgery achieved performance comparable to traditional optical navigation (TON) in certain clinical scenarios, while significantly reducing equipment costs, according to an August 1 study published in BMC Surgery.

The MRN system, built around a Microsoft HoloLens headset and infrared tracking tools, was tested in both laboratory and clinical environments. Laboratory validation involved 10 head models and five neurosurgeons performing spatial registration and localization tasks. Clinical evaluation was conducted in 46 patients undergoing neurosurgical procedures such as hematoma evacuation, tumor resection, or ventricular drainage.

In the laboratory, the fiducial registration error (FRE, which reflects the alignment accuracy between corresponding anatomical fiducial points on the physical and virtual models), ranged from 1.70 to 2.20 mm. The target registration error (TRE, reflecting the final localization accuracy) ranged from 1.30 to 1.70 mm. No significant differences in accuracy or task completion time among surgeons, authors report, and the learning curve stabilized after five localization trials.

In clinical settings, navigation time with MRN was 6.36 ± 1.27 min compared to 6.23 ± 1.30 min with TON. FRE ranged from 2.57 to 1.07 mm. Using TON as a reference, TRE was 2.14 ± 1.23 mm preoperative abd 3.65 ± 1.49 mm interoperative (after head-position change), with a mean increase in error of 1.51 mm.

MRN integration did not disrupt procedures, authors report. In fact, informal feedback from surgical teams noted ease of use and compatibility with standard operating room communication. Challenges reportedly include errors lined to patient movement and lighting changes; battery life and hardware fatigue; and patient discomfort, among others.

Nonetheless, authors conclude MRN is suitable for neurosurgical procedures requiring sub-centimeter accuracy and shows potential for broader clinical use, particularly in resource-constrained settings. Potential directions for future research include improved registration algorithms, ergonomic refinements, and expanded validation through randomized and multi-center trials.