Skip to Main Content

Abstract Details


Abstract Date: 5/1/2016

Raphael Jakubovic
Daipayan Guha, MD
Michael Lu
Shaurya Gupta
David Cadotte, MD, PhD
Chris Heyn, MD, PhD
Peter Howard, MD
Todd Mainprize, MD
Albert Yee, MD
Yang Victor (Toronto, Canada)

Surface anatomy based localization of neurosurgical targets is used in planning appropriate skin incisions, minimizing exposure, and optimizing surgical corridors. Computer-assisted frameless stereotactic navigation is standard for most cranial procedures requiring localization of deeper structures, and is employed in spinal procedures. With pre- and/or intraoperative imaging for navigation, there are workflow limitations including setup time, ease of registration, and ability to account for tissue movement between imaging and navigation.

A novel structured-light-illumination machine vision system was developed for cranial and spinal neuronavigation. 118 patients undergoing craniotomy or spinal procedures requiring stereotactic neuronavigation were enrolled with thin-slice preoperative imaging. Intraoperative patient registrations were performed using the system and compared to existing neuronavigation (Medtronic StealthStation, O-arm, Stryker Navigation). Navigation data was compared to post-operative imaging, where detectable implants' positions were measured on 3D reconstructions and absolute deviations determined. 

The system provided surgical exposure maps for automatic registration using fast (<3sec) iterative closest point algorithm, despite extensive datasets (mean 25,000 points/cranial, 1,500 points/vertebra). Time from optical imaging to navigation was 46+-39sec, statistically faster than existing cranial (111+-81sec) and spinal (254+-192sec) navigation systems’ setup time. Cranial registrations on scalp and face showed absolute error of 2.08+-3.05mm (axial). Lamina spinal registrations showed absolute error of 1.73+-1.34mm (axial) and 1.22+-0.98mm (sagittal). There is no statistically significant difference in absolute positional measurement error between machine vision system and existing neuronavigation techniques.

Optical machine vision is faster and comparably accurate for cranial and spinal neuronavigation. Rapid structured-light-illumination allows efficient initial and repeat registrations with minimal workflow interruptions. 


Article ID: AA-34531

Abstract Center

Support the NREF When You Shop

When you shop at and designate the NREF as your charity, Amazon donates.

Support the NREF

Donate Here