Dynamic Airway Deformation Tracking for Navigational Bronchoscopy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigational bronchoscopy systems struggle with accurately tracking interventional instruments within the dynamically moving and complex airways of the lung, particularly due to the flexible and dynamic nature of the lungs, which leads to inherent positioning errors and inaccuracies.
Innovation Solution
A method and system for tracking the positioning of interventional instruments within the lung airways using a skeletonized model of the airways, incorporating anatomically-based constraints to limit shape changes and using machine learning to score the plausibility of airway segment arrangements, while adjusting relative angles and positions based on respiratory phases and body motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a static skeletonized model of airways is used for navigation, then the system complexity is reduced, but positioning accuracy deteriorates due to lung movement and deformation
Solution Approach 1:
The patent applies dynamics by transforming the static airway model into a dynamic one that adapts to real-time lung movement. The system continuously updates the skeletonized model based on measured positions of interventional instruments and respiratory phase data, allowing the model to reflect actual lung deformation during procedures. This resolves the contradiction by maintaining simplicity while incorporating temporal variability through dynamic parameter adjustment.
Solution Approach 2:
The system changes parameters of the airway model based on measured positions and respiratory phase. By adjusting model parameters (such as airway segment positions and orientations) according to real-time measurements and anatomical constraints, the system maintains positioning accuracy without increasing overall system complexity. The parameter updates are performed incrementally based on observed deviations from the static model.
2Measurement precision
If anatomically-based constraints are applied to limit shape changes, then positioning accuracy is improved, but device complexity increases due to additional computational requirements
Solution Approach 1:
The patent applies preliminary action by pre-defining anatomical constraints based on known lung geometry and airway structure. These constraints (such as maximum bending angles, segment length variations, and hierarchical relationships) are established before the procedure and stored in the system. During navigation, the system only needs to check and enforce these pre-computed constraints, reducing real-time computational complexity while maintaining positioning accuracy.
Solution Approach 2:
The system uses feedback by continuously comparing measured instrument positions with predictions from the constrained dynamic model. When deviations are detected, the system adjusts model parameters within the bounds of anatomical constraints to reduce the error. This closed-loop feedback mechanism improves positioning accuracy while the pre-defined constraints limit the computational search space.
3Measurement precision
If the model dynamically adjusts to respiratory movements, then positioning accuracy is improved, but the reliability of the model deteriorates due to excessive flexibility and potential over-fitting
Solution Approach 1:
The patent applies local quality by allowing dynamic adjustment only in local regions where measurements indicate actual deformation, while maintaining the global structure and anatomical constraints of the overall lung model. Each airway segment can be adjusted independently within its anatomical bounds, preventing over-fitting while improving local tracking accuracy. The hierarchical nature of the airway tree ensures that local adjustments propagate appropriately without compromising global model integrity.
Data Source
AI summary
Systems and methods for tracking movement of a catheter within airways of a lung. A deformable model of the lung represents airways of the lung as airway segments joined at bifurcations. Deformation of the lung model uses modification of the angles and/or positions of the airway segments with respect to each other. An initial model may be generated, for example, based on segmentation of a CT image. A baseline deformable registration of the initial model to a lung shape of the patient at the beginning of the procedure may be established from position measurements of the catheter along plurality of different pathways. Optionally, the baseline registration is dynamic according to respiratory phase. Relative to the baseline registration, real-time changes in lung shape may be modeled by using further measurements obtained using the catheter as it navigates to a target, preferably using position sensors distributed along the catheter body.