Bronchoscopic Navigation and Visual Tracking
Image-guided bronchoscopy using visual SLAM, anatomical analysis, and CT-based navigation
Overview
This project focuses on image-guided bronchoscopic navigation using computer vision, visual SLAM, deep learning, and pre-operative CT information.
The goal is to improve bronchoscope localization, tracking robustness, anatomical understanding, and navigation accuracy during bronchoscopic procedures.
Research Topics
- Visual SLAM-based bronchoscope tracking
- ORB-SLAM-based localization and reconstruction
- Robust feature matching and outlier elimination
- Bronchial orifice segmentation and tracking
- Bronchial branching-level estimation
- Bronchial anatomical structure analysis
- Bronchoscope-to-CT registration
- Bronchial nomenclature estimation
- Navigation path generation
- Depth estimation from bronchoscopic images
Research Period
2016–Present
Collaborations
- Nagoya University
- Olympus Corporation
- Clinical collaborators in bronchoscopy and respiratory medicine
Representative Publications
- Visual SLAM for Bronchoscope Tracking and Bronchus Reconstruction in Bronchoscopic Navigation, SPIE Medical Imaging, 2019.
- Improved Visual SLAM for Bronchoscope Tracking and Registration with Pre-operative CT Images, SPIE Medical Imaging, 2020.
- Bronchial Orifice Segmentation on Bronchoscopic Video Frames Based on Generative Adversarial Depth Estimation, SPIE Medical Imaging, 2021.
- Bronchial Orifice Tracking-based Branch Level Estimation for Bronchoscopic Navigation, SPIE Medical Imaging, 2022.
- Anatomy Aware-based 2.5D Bronchoscope Tracking for Image-guided Bronchoscopic Navigation, Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2023.
- Real Bronchoscopic Images-based Bronchial Nomenclature: A Preliminary Study, SPIE Medical Imaging, 2023.
Keywords
Bronchoscopic Navigation, Visual SLAM, Computer-Assisted Intervention, Bronchoscope Tracking, Medical Image Analysis, Computer Vision, Deep Learning, CT Registration