A novel registration method for image-guided neurosurgery system based on stereo vision

Area of Science:

• Neurosurgery
• Computer Vision
• Medical Imaging

Background:

• Accurate spatial registration is crucial for image-guided neurosurgery systems (IGNS).
• Existing methods can be costly and complex.
• A need exists for accessible and effective registration techniques.

Purpose of the Study:

• To develop and validate a novel, low-cost spatial registration method for IGNS using stereo-vision.
• To enable precise alignment of patient data in real-time during neurosurgical procedures.

Main Methods:

• Utilized stereo-vision to reconstruct sparse facial data points from patient head images captured by a calibrated video camera.
• Employed an optical tracking system to track the camera.
• Applied a surface matching technique to register the reconstructed sparse points with preoperative facial surface data.

Main Results:

• Simulation experiments confirmed the feasibility and accuracy of the proposed stereo-vision-based registration method.
• The method demonstrated effective spatial alignment of patient-specific data.

Conclusions:

• The developed method offers a new, cost-effective, and user-friendly approach to spatial registration for IGNS.
• This technique holds significant potential for widespread clinical application in neurosurgery.