Endoscopic AI Recognition Display Gated by Image Blur
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In endoscopic imaging, it is challenging to maintain a consistent distance from the lesion due to non-stationary body tissues, leading to varying image clarity, which affects the accuracy of artificial intelligence recognition processes and may impede diagnosis.
Innovation Solution
A medical imaging apparatus that determines the state of image blur and controls the display of recognition results based on blur using convolutional neural networks and support vector machines, as well as pixel gradient calculations, to ensure accurate display of recognition outcomes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the endoscope tip is kept at a fixed distance from the lesion, then image clarity is improved, but this is difficult to achieve due to non-stationary body tissues
Solution Approach 1:
The system performs preliminary assessment of image quality using blur detection algorithms before the recognition process. By evaluating image clarity in advance, the system can determine whether the current image is suitable for AI analysis, preventing wasted computational resources on blurry images and alerting operators to adjust the endoscope position beforehand.
Solution Approach 2:
The blur detection function serves as an intermediary between image acquisition and AI recognition. It acts as a quality gatekeeper that assesses whether the captured image meets the necessary clarity standards, mediating between the physical act of imaging and the digital processing stage by filtering out substandard images.
2Productivity
If blurry images are processed by the recognition process, then more images can be analyzed, but inaccurate results may be obtained
Solution Approach 1:
The system implements a feedback mechanism where blur detection results directly influence the decision to proceed with or skip the recognition process. When blur is detected above a threshold, the system provides feedback to skip AI analysis for that frame and continues to the next image, ensuring that only clear images undergo computationally intensive recognition processing.
Solution Approach 2:
The system changes the operational parameter of image selection based on blur detection results. By dynamically adjusting which images are submitted for recognition processing based on their clarity parameters, the system optimizes the balance between processing throughput and result accuracy, processing only those images that meet quality thresholds.
Data Source
AI summary
An image acquisition unit sequentially acquires a medical image. A recognition processing unit applies a recognition process to the medical image. A recognition result display determination unit determines to display or not display the result of the recognition process on a display according to blur in the medical image. A display control unit controls the display of the recognition process according to the determination by the recognition result display determination unit.