Operative Neurosurgery 30:414–421, 2026
This technical note presents a proof-of-concept augmented reality (AR) system that guides external ventricular drain (EVD) placement by integrating cloud-based automatic CT segmentation, PACS compatibility, trajectory planning, point-based image-to-patient registration, and real-time 3D AR guidance via an AR head-mounted display. The low-cost, end-to-end workflow and universal tracking tools aim to reduce attention shifts and improve accessibility compared with conventional neuronavigation.
The system was tested in a phantom study with 29 AR-guided EVD insertions by neurosurgical clinicians, yielding 69% optimal placements (Kakarla grade 1), a mean distance-to-target of 9.49 mm, mean angular deviation of 9.20°, an FRE of 4.00 mm, and mean procedural time of 22:45. Authors identify human–computer interaction, tracking fidelity, registration accuracy, and procedural time as primary areas for refinement before clinical adoption.
Augmented Reality (AR) Guidance: A low-cost, end-to-end AR system was developed for external ventricular drain (EVD) placement, integrating real-time 3D guidance, automated CT segmentation, and compatibility with standard hospital PACS infrastructure, all visualized through AR head-mounted displays (AR-HMDs).
Workflow Components: The system includes cloud-based image storage, automatic segmentation, trajectory planning, point-based image-to-patient registration, and real-time EVD tracking, aiming to reduce attention shifts and improve procedural integration.
Proof-of-Concept Results: In simulated procedures on anatomical phantoms with small ventricles, 69% of placements were optimal (Kakarla 1), with a mean distance to target of 9.49 mm and mean angular deviation of 9.20°, but accuracy is not yet at the level of best clinical standards.
Procedural Time: The mean workflow duration was nearly 23 minutes, which is longer than acceptable for emergency EVD placements, with most of the added time attributed to trajectory planning, marker attachment, and image-to-patient registration.
Usability and Interface Challenges: Users experienced difficulties with human-computer interaction, including issues with holographic controls, visual clutter, and marker tracking, which impacted both speed and accuracy.
Affordability and Accessibility: The AR-HMD system (approx. $4950) is significantly less expensive than traditional neuronavigation systems, potentially increasing access to advanced guidance in resource-limited settings
You must be logged in to post a comment.