Fluorescent Endoscopic Image Calculation for S/N Ratio Improvement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fluorescent endoscopic devices face challenges in effectively discriminating between normal and lesion tissues due to low signal-to-noise ratios (S/N) in fluorescent images, making it difficult to accurately identify lesions.
Innovation Solution
The device incorporates an irradiation portion for illuminating and exciting tissues, a light receiving portion for capturing reflection and fluorescent images, and a calculation portion that processes these images through addition and division operations to enhance image quality and contrast, creating a fluorescent observation image that improves S/N ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fluorescent images are obtained by irradiating excitation light to living tissue, then fluorescent images can be obtained to display lesion tissue, but the signal-to-noise ratio becomes extremely low making discrimination between normal and lesion tissue difficult
Solution Approach 1:
The patent combines multiple fluorescent images obtained by irradiating excitation light at different wavelengths into a single composite fluorescent image. By merging information from multiple wavelength regions, the system improves the signal-to-noise ratio and enables accurate discrimination between normal and lesion tissues that would be difficult to achieve with single-wavelength imaging alone.
Solution Approach 2:
The patent changes the wavelength parameter of excitation light to obtain fluorescent images at multiple different wavelength regions. By varying the excitation wavelength and combining the resulting images, the system enhances measurement precision and overcomes the low signal-to-noise ratio problem inherent in single-wavelength fluorescent imaging.
2Measurement precision
If standardized calculation is carried out using fluorescent images with low S/N, then a calculated image can be created, but the S/N of the calculated image also becomes extremely low
Solution Approach 1:
The patent performs preliminary action by obtaining and combining multiple fluorescent images at different wavelengths before conducting standardized calculation. By pre-combining the images to create a higher quality composite fluorescent image, the system ensures that subsequent calculations are performed on data with improved signal-to-noise ratio, preventing the propagation of low S/N through the calculation process.
3Loss of information
If excitation light is irradiated to obtain fluorescent images, then lesion tissue can be displayed, but the fluorescent intensity from living tissue is extremely weak
Solution Approach 1:
The patent merges fluorescent images obtained at multiple wavelength regions to accumulate signal information. By combining the weak fluorescent signals from different wavelengths, the system recovers sufficient information about lesion tissue that would be undetectable in any single wavelength image, effectively compensating for the extremely weak fluorescent intensity from living tissue.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the brightness and contrast of fluorescent images, alleviates the drop in S/N ratios, and creates image information reflecting the spectral shape, facilitating better discrimination between normal and lesion tissues.
Implementation Method 1
irradiating illumination light and excitation light to a subject, a light receiving portion for receiving a reflection light image generated from the subject on the basis of the illumination light and a first fluorescent image and a second fluorescent image generated from the subject on the basis of the excitation light
Data Source
AI summary
An inter-image calculation portion includes a switch circuit made up of three switch portions for switching each of three image data (R, G, B) from a 3-board processing portion, a first divider, a second divider, a first adder, a second adder, a first LUT, a second LUT, a first clip portion, and a second clip portion. The inter-image calculation portion executes addition processing of two different fluorescent images with two different wavelength bands and division processing the two different fluorescent images with two different wavelength bands and then it executes addition processing which adds the addition results and the division results of the two different fluorescent images with two different wavelength bands or it executes subtraction processing which subtracts the addition results from the division results of the two different fluorescent images with two different wavelength bands.