Position-Adaptive Lesion Detection in Endoscopic Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image processing systems face challenges in accurately detecting lesion regions from endoscopic images due to varying structures and characteristics of mucous membranes at different in-vivo positions, making it difficult to identify optimal lesion regions.
Innovation Solution
A medical image processing device and system that acquires images at multiple in-vivo positions, utilizes positional information to select appropriate region-of-interest detection units, and employs learned models specific to each position for accurate lesion detection, enabling real-time detection and display of optimal lesion regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single lesion detection model is used for all positions, then the device complexity is reduced, but the detection precision deteriorates due to varying mucous membrane structures at different in-vivo positions
Solution Approach 1:
The patent divides the detection system into multiple region-of-interest detection units, each corresponding to a specific in-vivo position (e.g., different sections of the colon). Each detection unit is specialized for detecting lesions at its corresponding position, allowing the system to adapt to position-specific mucous membrane structures and improve detection precision without requiring a single complex universal model
Solution Approach 2:
The patent implements local quality by creating detection units with position-specific characteristics. Each region-of-interest detection unit is configured with parameters and features optimized for its specific anatomical location, enabling the system to account for variations in mucous membrane structure, lighting conditions, and lesion appearance at different in-vivo positions
2Measurement precision
If multiple region-of-interest detection units are used for different positions, then the detection precision is improved, but the device complexity increases
Solution Approach 1:
The patent introduces a dynamic selection mechanism that automatically switches between different region-of-interest detection units based on the current in-vivo position detected from the medical image. This dynamic adaptation allows the system to use the most appropriate detection unit for the current position, maintaining high detection precision while managing complexity through automated position-based selection rather than requiring all units to be active simultaneously
Solution Approach 2:
The patent employs an in-vivo position detection unit as an intermediary that determines the current anatomical position and selects the corresponding region-of-interest detection unit. This intermediary component coordinates between the multiple detection units and the image processing system, enabling seamless switching and integration without requiring direct complex interactions between all detection units
3Measurement precision
If position-specific learned models are used, then the detection accuracy is improved, but the loss of time increases due to selecting and switching between multiple detection units
Solution Approach 1:
The patent performs preliminary action by pre-configuring multiple region-of-interest detection units with position-specific learned models before actual lesion detection begins. Each detection unit is pre-trained and optimized for its specific anatomical position, so when the system needs to detect lesions at a particular position, the corresponding pre-prepared detection unit can be immediately activated without requiring time-consuming model selection or training during the detection process
4Adaptability or versatility
If multiple detection units are implemented, then the adaptability to different positions is improved, but the ease of operation deteriorates due to managing multiple detection units
Solution Approach 1:
The patent implements self-service by enabling the system to automatically detect the current in-vivo position and select the appropriate region-of-interest detection unit without requiring manual intervention from the operator. The system autonomously manages the complexity of multiple detection units by using position information to automatically configure and switch between the appropriate detection models, maintaining high adaptability while preserving ease of operation
Data Source
AI summary
There are provided a medical image processing device, an endoscope system, a medical image processing method, and a program which detect an optimal lesion region according to an in-vivo position of a captured image. Images at a plurality of in-vivo positions of a subject are acquired from medical equipment that sequentially captures and displays in real time the images; positional information indicating the in-vivo position of the acquired image is acquired; from among a plurality of region-of-interest detection units that detect a region of interest from an input image and correspond to the plurality of in-vivo positions, respectively, a region-of-interest detection unit corresponding to the position indicated by the positional information is selected; and the selected region-of-interest detection unit detects a region of interest from the acquired image.