Related Experiment Video
Updated: Jul 18, 2025
Tracking Mouse Bone Marrow Monocytes In Vivo
Published on: February 27, 2015
Monocular endoscope 6-DoF tracking with constrained evolutionary stochastic filtering
Xiongbiao Luo1, Lixin Xie2, Hui-Qing Zeng3
1National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen 361102, China; Department of Computer Science and Technology, Xiamen University, Xiamen 361005, China; Discipline of Intelligent Instrument and Equipment, Xiamen University, Xiamen 361102, China; Fujian Key Laboratory of Sensing and Computing for Smart Cities, Xiamen University, Xiamen 361005, China.
This study introduces a new filtering method for monocular endoscopic camera tracking, significantly improving accuracy for augmented and virtual reality surgery navigation. The enhanced tracking aids in more precise surgical guidance using real-time endoscopic data.
Area of Science:
- Medical Imaging
- Computer Vision
- Robotics
Background:
- Monocular endoscopic 6-DoF camera tracking is crucial for surgical navigation systems.
- Current methods often face challenges like particle degeneracy in stochastic filtering.
- Accurate tracking is essential for multimodal image integration in augmented and virtual reality surgery.
Purpose of the Study:
- To develop an advanced pipeline for monocular endoscopic 6-DoF camera tracking.
- To address limitations in existing stochastic filtering techniques for camera tracking.
- To enhance the accuracy and robustness of endoscope tracking in surgical environments.
Main Methods:
- Proposed a novel pipeline of constrained evolutionary stochastic filtering.
- Introduced spatial constraints and evolutionary stochastic diffusion to overcome particle degeneracy.
- Applied and validated the method on extensive clinical endoscopic data (over 59,000 frames).
Main Results:
- The new pipeline significantly improved tracking accuracy compared to state-of-the-art methods.
- Achieved a notable reduction in tracking error from (4.83 mm, 10.2°) to (2.78 mm, 7.44°).
- Demonstrated effectiveness across diverse surgical procedures using real clinical data.
Conclusions:
- The constrained evolutionary stochastic filtering pipeline offers superior performance for monocular endoscope tracking.
- This advancement holds significant potential for improving the precision of image-guided surgery.
- The method effectively enhances augmented and virtual reality applications in surgical navigation.
