Endoscope Auto Fluorescence Discrimination via Differential Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current endoscope systems face challenges in accurately distinguishing between normal and abnormal regions in biological tissues during auto fluorescence observation, particularly due to variations in illumination conditions and noise components.
Innovation Solution
The system employs a light source apparatus emitting excitation and reference lights, along with a processor that calculates differential values from fluorescence and reference light images, applying threshold processing to discriminate between normal and abnormal regions by generating a determination target image through division or subtraction of these values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If auto fluorescence observation is performed using excitation light and reference light, then fluorescence images and reference light images can be acquired, but accurate discrimination between normal and abnormal regions becomes difficult due to illumination variations and noise
Solution Approach 1:
The system changes the parameter of image processing by calculating differential values between the fluorescence image and reference light image, and further processing these differentials through threshold processing to generate determination target images. This parameter transformation enables clear discrimination between normal and abnormal regions despite illumination variations and noise in the original images.
Solution Approach 2:
The determination target image serves as an intermediary that mediates between the raw fluorescence and reference light images and the final diagnosis. By processing the differential values through threshold processing, the intermediary determination target image simplifies the complex image data into a form that clearly indicates abnormal regions, improving both discrimination accuracy and detection reliability.
2Measurement precision
If threshold processing is applied to the calculation result of differential values, then abnormal regions can be identified, but the system complexity increases due to multiple processing sections
Solution Approach 1:
The image processing system is segmented into functional sections: a differential value calculation section that computes differences between fluorescence and reference light images, and a determination target image generation section that applies threshold processing to identify abnormal regions. This segmentation allows each section to perform its specific function efficiently while maintaining overall system clarity and manageability.
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 enables clear identification of abnormal regions irrespective of observation conditions, enhancing precision by correcting for illumination variations and noise, and displaying the results on a monitor for effective lesion detection.
Implementation Method 1
an object is irradiated with an excitation light and a reference light having specific wavelength bands, and a fluorescence image as an image of auto fluorescence emitted from the object in response to the excitation light
Implementation Method 2
a reference light image as the image of a reflection light which is the reference light reflected in the object
Data Source
AI summary
An image pickup system of the present invention has a light emitting section that emits a first illuminating light and a second illuminating light to an object, an image pickup section that picks up an image of a first return light and an image of a second return light, and outputs the images as image pickup signals respectively, an image generating section that generates a first image and a second image respectively based on the image pickup signals, a differential value calculating section that calculates a first differential value corresponding to the first image and a second differential value corresponding to the second image respectively, a calculation section that performs calculation processing by using the first differential value and the second differential value, and a region discriminating section that discriminates between regions in the object by applying threshold processing to a calculation result of the calculation section.