Endoscope Image Processing Device for Site-Specific Observation Support
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current endoscopy systems lack the capability to provide tailored observation support information for varying anatomical structures during procedures, leading to inadequate guidance for operators, especially when dealing with multiple sites with distinct internal structures, and fail to notify operators about unsuitable medical images for observation.
Innovation Solution
An image processing device that identifies the observation target within a medical image using a trained algorithm, selects appropriate support algorithms based on the target, and provides real-time notification through guide images and voice instructions for optimal endoscope positioning and operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single type of trained model is used to output operation support information, then the device complexity is reduced, but the adaptability to different observation sites is insufficient
Solution Approach 1:
The patent divides the observation support information generation into multiple independent trained models, each specialized for a specific observation site (e.g., nasal cavity, pharynx, larynx). This segmentation allows each model to be optimized for its specific anatomical region while maintaining overall system manageability through modular architecture.
Solution Approach 2:
The system dynamically selects and switches between different trained models based on the current observation site detected by the observation target identification algorithm. This dynamic model selection enables the system to adapt to varying anatomical structures in real-time without requiring all models to be simultaneously active, balancing adaptability with computational efficiency.
2Reliability
If observation support information is not provided for unsuitable medical images, then the processing time is reduced, but the reliability of observation support is compromised
Solution Approach 1:
The observation target identification algorithm performs preliminary analysis of the medical image to determine whether the image is suitable for observation before generating support information. This preliminary action filters out unsuitable images early in the processing pipeline, preventing wasted computational resources on images that cannot provide useful observation support.
Solution Approach 2:
The system incorporates feedback mechanisms where the observation target identification algorithm continuously monitors the quality and suitability of captured medical images. When unsuitable images are detected, the system provides feedback to the operator through notifications, enabling real-time adjustment of the endoscope position or angle to capture suitable images for effective observation support.
Data Source
AI summary
A processor included in the image processing device acquires a medical image, outputs observation target identification information indicating an observation support target part or indicating that the medical image is out of an observation support target by inputting the medical image to an observation target identification algorithm, selects, from a plurality of observation support algorithms, one specific observation support algorithm based on the observation target identification information, outputs observation support information by inputting the medical image to the specific observation support algorithm, and performs a control of notifying of the observation support information.