Weighted Histogram Image Processing for Diagnostic Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radiation image processing methods face challenges in obtaining suitable images for diagnosis due to variations in high and low density areas, requiring specific settings for each region of interest, which can lead to defective image quality, especially in small-scale facilities lacking dedicated image processing personnel.
Innovation Solution
An image processing method that detects important areas like bone or soft tissue, creates a weighted histogram, and determines image processing conditions based on a predetermined evaluation function to maximize the shift value, allowing for robust image processing without the need for exclusive region of interest settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If cumulative histogram of image data in a desired area is obtained and data level is set as reference signal value, then image processing can be performed, but depending on difference in rate of high density area to low density area, a visible radiation image suited to diagnosis may not be obtained
Solution Approach 1:
The patent applies local quality by creating a weight image that assigns different weights to different regions based on their importance for diagnosis. The weight image multiplying section multiplies the cumulative histogram by these regional weights, allowing different areas (e.g., lung fields vs. mediastinum) to contribute differently to the overall histogram. This ensures that diagnostically critical regions have greater influence on the reference signal value determination, thereby improving image quality suitability for diagnosis while maintaining automated operation.
2Device complexity
If image processing is performed based on cumulative histogram without considering region of interest settings, then processing can be simplified, but image quality may become defective
Solution Approach 1:
The patent implements self-service by enabling the image processing system to automatically determine appropriate reference signal values without requiring manual region of interest selection or specialized personnel intervention. The weight image multiplying section automatically applies predetermined weights to different regions, and the reference signal value determining section automatically calculates the optimal reference value from the weighted cumulative histogram. This maintains system simplicity while improving reliability through automated adaptive processing that adjusts to the specific characteristics of each radiograph.
3Manufacturing precision
If specific settings are made for each region of interest, then image quality can be improved, but it requires dedicated image processing personnel which is not available in small-scale facilities
Solution Approach 1:
The patent applies universality by creating a weight image that can accommodate multiple diagnostic regions and conditions within a single processing framework. The predetermined weights in the weight image are designed to handle various anatomical regions (lung fields, mediastinum, bones) and different radiograph types (chest, abdominal, skeletal) using the same processing mechanism. This allows the system to provide specialized region-specific processing quality improvement while maintaining ease of operation through automated universal processing that requires no specialized personnel.
Data Source
AI summary
An image processor executes an image processing under an appropriate condition. The image possessor comprises an important area detecting section (120) for detecting a bone portion or soft portion as an important area included in an X-ray image, a weight image creating section (130) for creating a weight image for imparting a predetermined weight to the pixels of the important area, histogram computing section (140) for multiplying the weight of the weight image by the pixel value of the X-ray image and computing a weighted histogram from the results of the multiplication, evaluating section (150) for evaluating the weighted histogram with a predetermined evaluation function and computing a shift value at which the evaluation result takes on a maximum value, image processing condition determining section (160) for so determining an image processing condition so as to obtain a predetermined processing result from the pixel value of the X-ray image corresponding to the maximum value of the evaluation function for obtaining a shift value, and image processing section (170) for executing an image processing under the image processing condition.