Rendering-Based Video-CT Registration with Physical Constraints for Image-Guided Endoscopic Sinus Surgery.

Area of Science:

• Medical Imaging
• Computer-Aided Surgery
• Surgical Navigation

Background:

• Accurate registration of endoscopic data to pre-operative CT scans is crucial for minimally invasive surgery.
• Current methods may lack robustness or require significant manual intervention.

Purpose of the Study:

• To develop and evaluate a robust, semi-automatic system for registering the coordinate frame of an endoscope to CT data.
• To improve the accuracy and efficiency of image-guided interventions.

Main Methods:

• A novel system optimizes the similarity metric between endoscopic images and CT-rendered images.
• Physical constraints, including endoscope-anatomy collisions, are used for initialization and search space limitation.
• A stochastic optimization algorithm leveraging a graphics processing unit (GPU) evaluates similarity metrics in parallel.

Main Results:

• Registration errors of 0.83 mm (cadaver) and 1.97 mm (patient) were achieved.
• The average registration time was 4.4 seconds across 60 trials.
• The system demonstrated robustness against moderate anatomical deformation in a patient study.

Conclusions:

• The proposed system offers accurate and efficient semi-automatic registration of endoscopic views to CT data.
• The integration of physical constraints and GPU acceleration enhances robustness and speed.
• This technology has the potential to significantly aid surgical navigation and planning.