Robust endoscope motion estimation via an animated particle filter for electromagnetically navigated endoscopy.

Area of Science:

• Medical Imaging
• Robotics
• Biomedical Engineering

Background:

• Electromagnetically navigated endoscopy is crucial for minimally invasive procedures.
• Challenges include electromagnetic tracking (EMT) inaccuracy due to patient motion and magnetic field distortion.
• Existing EMT methods struggle with dynamic jitter errors.

Purpose of the Study:

• To develop a novel tracking framework to enhance electromagnetic navigation accuracy and robustness.
• To address limitations of current electromagnetic tracking (EMT) methods in endoscopic interventions.
• To improve the precision and smoothness of endoscope navigation.

Main Methods:

• Proposed a new tracking framework integrating an animated particle filter with adaptive particle swarm optimization.
• Validated the method on a dynamic phantom.
• Compared performance against state-of-the-art electromagnetic tracking (EMT) techniques.

Main Results:

• Achieved position and orientation accuracy of 2.48 mm and 4.69°, significantly outperforming previous methods (4.19 mm, 7.75°).
• Improved tracking smoothness from 4.09 mm, 3.37° to 1.84 mm, 2.52°.
• Successfully mitigated particle impoverishment, outperforming standard particle filters.

Conclusions:

• The proposed animated particle filter framework significantly enhances electromagnetic navigation for endoscopic interventions.
• The method demonstrates superior accuracy, robustness, and tracking smoothness compared to existing EMT techniques.
• This advancement offers improved safety and efficacy in image-guided medical procedures.