EVD Surgical Guidance With Retro-Reflective Tool Tracking and Spatial Reconstruction Using Head-Mounted Augmented...

Area of Science:

• Neurosurgery
• Medical Imaging
• Computer-Aided Surgery

Background:

• Augmented Reality (AR) offers in-situ visual guidance for External Ventricular Drain (EVD) surgery.
• Accurate spatial registration between pre-operative images and patient anatomy is critical for EVD procedures.
• Existing tracking methods face trade-offs between accuracy, efficiency, and non-invasiveness.

Purpose of the Study:

• To develop and validate a convenient and accurate AR guiding pipeline for EVD surgery.
• To leverage Time-of-Flight (ToF) depth sensors for improved spatial relationship estimation.
• To address and correct depth errors inherent in ToF sensors for surgical applications.

Main Methods:

• Utilized Time-of-Flight (ToF) depth sensors for retro-reflective tool tracking and dense surface reconstruction.
• Evaluated and corrected depth errors of the HoloLens 2 depth camera on various materials.
• Reconstructed head surfaces with sub-millimeter accuracy using corrected depth data, validated on phantoms and a sheep head.

Main Results:

• Identified significant depth errors on human skin using the HoloLens 2 depth camera.
• Demonstrated a proposed depth correction method effectively reduced errors on head phantoms.
• Achieved sub-millimeter accuracy in head surface reconstruction.
• Simulated EVD surgery case study showed competitive translational and orientational guidance accuracy.

Conclusions:

• The proposed depth correction method significantly improves ToF sensor accuracy for surgical guidance.
• The developed AR pipeline provides accurate and convenient guidance for EVD surgery.
• The framework shows potential for enhancing surgical navigation and patient outcomes.