Medical Image Standardization via Anatomical Feature Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical images captured by MRI apparatuses face challenges in accurate anatomical region extraction due to non-standardized pixel values, leading to variations in image quality and contrast, which complicates automatic or semi-automatic extraction processes.
Innovation Solution
A medical image processing device and method that includes a feature point extraction unit and a standardization conditions acquisition unit to extract anatomical feature points and acquire standardization conditions based on statistical quantities of pixel values around these points, allowing for conversion processing and region extraction with reduced pixel value and contrast variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If pixel values are not standardized in MR images, then image quality and contrast vary greatly with magnetic field strength or pulse sequence, but this makes automatic or semi-automatic extraction difficult
Solution Approach 1:
The patent applies parameter changes by standardizing pixel values through statistical transformation. It calculates statistical quantities (mean, standard deviation) from pixel values around anatomical feature points and uses these to normalize pixel values across different imaging conditions. This transforms the parameter space of pixel values to enable consistent extraction accuracy across varying magnetic field strengths and pulse sequences.
Solution Approach 2:
The patent replaces manual window width and level adjustment with an automated statistical processing system. Instead of mechanical/manual parameter tuning, it substitutes a computational system that automatically calculates standardization conditions based on anatomical feature points and their surrounding pixel statistics, eliminating the need for operator intervention.
2Measurement precision
If manual adjustment of window width and window level is performed for each imaging, then image observation in desired state is possible, but time for adjustment is required
Solution Approach 1:
The patent applies preliminary action by pre-calculating standardization conditions based on anatomical feature points before actual image analysis. It performs statistical analysis of pixel values around feature points in advance, storing standardization parameters that can be quickly applied during extraction processes, eliminating the need for time-consuming manual adjustments at the time of use.
Solution Approach 2:
The system performs self-service by automatically determining appropriate window width and level settings through statistical analysis of the image data itself. It extracts anatomical feature points, calculates surrounding pixel statistics, and generates standardization conditions autonomously without requiring operator input or manual parameter tuning.
3Measurement precision
If threshold value processing is used for CT images, then anatomical regions can be extracted with certain accuracy, but this method cannot be applied to MR images with non-standardized pixel values
Solution Approach 1:
The patent achieves universality by creating a standardized processing framework that works across different imaging modalities. By transforming MR image pixel values through statistical normalization based on anatomical feature points, it brings MR images into a comparable parameter space with CT images, enabling the same threshold-based extraction methods to be applied universally to both modalities.
Data Source
AI summary
There are provided a feature point extraction unit that extracts an anatomical feature point included in a medical image based on the medical image and a standardization conditions acquisition unit that acquires standardization conditions of pixel values of the medical image based on some pixel values around the anatomical feature point among the pixel values of the medical image.