Endoscope Treatment Tool Tracking via Virtual Model Intersection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical systems require specific magnetic sensors or color markers on treatment tools to track their position, which is cumbersome and inefficient, especially when tools are replaced or multiple tools of the same type are used, as they need to be distinguished and tracked accurately within endoscopic images.
Innovation Solution
A medical system that uses an endoscope and a processor to extract treatment tools from images, determine their longitudinal axes, calculate the intersection coordinates of these axes, and judge whether they serve as follow targets based on these calculations, allowing the endoscope's field of view to be adjusted to follow the treatment tools without the need for additional markers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If magnetic sensors or color markers are attached to treatment tools for position detection, then the position tracking accuracy is improved, but the device complexity and operational burden increase
Solution Approach 1:
The patent creates a virtual model (copy) of the treatment tool that mirrors its physical appearance and characteristics. This virtual model is used for tracking and identification purposes, eliminating the need for physical markers or sensors on the actual tool. The system captures images, extracts features to create the virtual model, and uses this model to follow the tool's position and orientation throughout the procedure.
Solution Approach 2:
The patent replaces the mechanical/optical marker-based tracking system with an image processing-based system. Instead of using magnetic sensors or color markers that require physical attachment and detection hardware, the system uses standard endoscopic imaging combined with computer vision algorithms to detect, extract features, and track the treatment tool.
2Measurement precision
If markers are attached to treatment tools for identification, then the tool identification accuracy is improved, but the ease of operation deteriorates due to frequent re-marking
Solution Approach 1:
The system creates a virtual model that captures the unique visual characteristics of each treatment tool. This model serves as an identifier that persists throughout the procedure, eliminating the need to physically re-mark tools. The virtual model is generated once from image data and used for continuous identification and tracking.
Solution Approach 2:
The treatment tools themselves provide their own identification features through their inherent visual characteristics (shape, color, pattern, reflective properties). The system extracts these features automatically from standard endoscopic images, allowing the tools to identify themselves without requiring external markers or labels that would need to be applied and maintained.
3Adaptability or versatility
If multiple treatment tools of the same type are used, then the treatment capability is improved, but the difficulty of detecting and measuring increases due to differentiation requirements
Solution Approach 1:
The system segments and extracts multiple feature parameters from each treatment tool's appearance, including shape characteristics, color properties, reflective patterns, and geometric features. By analyzing multiple features simultaneously, the system can differentiate between multiple tools of the same type based on subtle variations in their visual characteristics.
Solution Approach 2:
The system analyzes local features and characteristics of each treatment tool individually, such as specific patterns, reflections, or geometric details that are unique to each tool instance. This local feature analysis enables differentiation even when tools share the same overall type or category.
Data Source
AI summary
A medical system includes: an endoscope and at least one treatment tool; a treatment-tool coordinate calculating unit that extracts the treatment tool by processing two or more images acquired at different times by the endoscope, that determines directions of longitudinal axes of the extracted treatment tool, and that calculates a coordinate of an intersection of the determined two or more longitudinal axes; and a judgment unit that judges whether the treatment tool serves as a follow target, on the basis of the coordinate of the intersection calculated by the treatment-tool coordinate calculating unit.