Cortical surface shift estimation using stereovision and optical flow motion tracking via projection image...

Area of Science:

• Neurosurgery
• Medical Imaging
• Computer Vision

Background:

• Intraoperative brain shift necessitates accurate monitoring during cranial surgery.
• Stereovision offers non-invasive imaging of the exposed cortical surface.
• Real-time estimation of cortical surface shift is crucial for surgical accuracy.

Purpose of the Study:

• To develop and validate an automatic and robust registration technique for compensating cortical surface displacement during open cranial surgery.
• To utilize optical flow (OF) motion tracking for efficient and accurate estimation of brain shift.
• To enable real-time intraoperative adjustments based on dynamic changes in brain anatomy.

Main Methods:

• Reconstructed 3D cortical surfaces from stereo images acquired at multiple time points.
• Established a local coordinate system for 2D projection images.
• Applied dense optical flow (OF) motion tracking to determine 2D displacement fields.
• Inverted spatial mapping to calculate full 3D cortical surface displacement.

Main Results:

• Digital phantom validation showed high accuracy (0.05 pixels).
• Surgical case comparisons with a tracked stylus demonstrated average agreement of 1.7-2.1mm for feature coordinates.
• Feature displacement tracking agreement was comparable to probe data (<1mm difference).
• Average cortical surface displacement measured at 7.9 ± 5.7 mm, with gravity-dependent shift of 5.2 ± 6.0 mm.

Conclusions:

• The developed optical flow technique provides accurate and efficient (∼15 s) compensation for intraoperative brain shift.
• This method is suitable for real-time applications in the operating room.
• The technique enhances surgical navigation by accounting for brain deformation.