Endoscope Lesion Overlay Imaging Across Tissue Layers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing endoscope systems struggle to accurately and efficiently detect lesions across different tissue layers and overcome obstacles like residues or bleeding, leading to potential oversight of lesions during diagnosis.
Innovation Solution
An endoscope device equipped with multiple illumination lights of varying spectra, combined with AI-driven identification devices, detects lesion candidates in multiple layers and selects a region for display based on reliability scores, aligning and superimposing these regions onto a white light image for enhanced visibility and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple illumination lights with different spectra are used to detect lesion candidates in multiple layers, then lesion detection accuracy is improved, but device complexity increases
Solution Approach 1:
The patent divides the detection process into multiple segments by using different illumination lights (white light, blue light, red light) to detect different tissue layers separately. Each illumination light targets specific depth ranges, allowing the system to segment the complex task of multi-layer lesion detection into manageable parts that can be processed independently and then integrated.
Solution Approach 2:
The endoscope device is designed with multi-functionality by incorporating multiple illumination lights that can serve different purposes: white light for general overview, blue light for deeper tissue penetration, and red light for superficial tissue examination. This universal design allows a single device to perform multiple detection functions simultaneously, improving lesion detection accuracy across different tissue depths without requiring separate devices.
2Reliability
If multiple illumination lights with different spectra are used to detect lesion candidates in multiple layers, then detection of lesions obscured by residues or bleeding is improved, but device complexity increases
Solution Approach 1:
The patent changes the spectral parameters of illumination light to overcome obstacles like residues or bleeding. By switching between different wavelengths (white, blue, red light), the system can penetrate through or around obstructions that may block certain wavelengths, ensuring reliable lesion detection even in challenging conditions where tissues are obscured.
Solution Approach 2:
The system applies partial action by selectively using only the necessary illumination types for each specific detection scenario. When residues or bleeding obscure certain areas, the system can activate specific wavelength illuminations that are most effective for penetrating or highlighting those particular obstructions, rather than continuously using all illumination types, thus improving reliability while managing device complexity.
3Measurement precision
If AI-driven identification devices are used to select lesion regions based on reliability scores, then diagnostic accuracy is improved, but processing time increases
Solution Approach 1:
The AI-driven identification device performs preliminary action by pre-calculating and assigning reliability scores to different detected lesion candidates based on their visual characteristics and detection confidence. This preliminary scoring allows the system to prioritize and quickly identify the most promising lesion regions for further diagnostic review, reducing the time needed for manual assessment while maintaining high diagnostic accuracy.
Solution Approach 2:
The system implements feedback by using the reliability scores generated by the AI to guide subsequent diagnostic processes. The feedback loop allows the system to learn from previous detections and improve its prioritization of lesion candidates, gradually reducing processing time while maintaining or improving diagnostic accuracy through iterative optimization.
Data Source
AI summary
An endoscope apparatus including a processor comprising hardware. The processor being configured to: detect a first region of a lesion candidate from first image information acquired by irradiation with first illumination light; detect a second region of a lesion candidate from second image information acquired by irradiation with second illumination light having a different spectrum from the first illumination light; select a region for display of the lesion candidate out of the first region and the second region, the region for display corresponding to an observation target site of a subject; and generate image information for display, in which the region for display is superimposed on the first image information. Wherein the region for display is larger than any one of the first region or the second region.