Body Motion Detection via Contrast Ratios in Radiographic Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing body motion detection techniques for radiographic images are not effective in accurately detecting motion under varying imaging conditions and subjects, leading to low stability and a narrow range of applicability.
Innovation Solution
A body motion detection device and method that calculate contrast ratios of high and low frequency components in radiographic images at multiple analysis points, determining gradient directions and using statistical processing to identify body motion, independent of imaging conditions or subject type, while excluding inappropriate regions such as artificial objects and gas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If body motion detection is performed using learning-based techniques for specific body parts, then detection accuracy is improved for those specific parts, but the technique cannot be applied to radiographic images of body parts for which learning has not been performed
Solution Approach 1:
The patent applies universality by using gradient directions of edge portions, which are fundamental image features that exist in all radiographic images regardless of body part. Instead of learning body-part-specific features, the invention extracts edges and calculates gradient directions at these edges, making the detection method universally applicable to any body part while maintaining detection accuracy through the universal presence of edges in radiographic images
2Measurement precision
If edge degradation is learned as feature points in two directions, then body motion detection is achieved for specific imaging conditions, but the technique shows low stability and narrow applicability when imaging conditions or subjects differ from learning data
Solution Approach 1:
The patent applies parameter changes by calculating gradient directions in multiple directions (not limited to two fixed directions) and selecting the direction with the maximum gradient magnitude. This adaptive approach allows the method to adjust to different edge orientations and imaging conditions automatically, providing stable detection across varying subjects and imaging parameters without requiring relearning
3Measurement precision
If a radiographic image is enlarged to improve observation, then body motion detection capability is improved, but the workflow efficiency of the technologist decreases
Solution Approach 1:
The patent replaces the mechanical action of manually enlarging and observing images with an automated computational system. The body motion detection device automatically calculates gradient directions, determines edge portions, and detects motion without requiring the technologist to manually enlarge images or perform visual inspection, thereby maintaining detection capability while improving workflow efficiency
Data Source
AI summary
A contrast calculating unit calculates, as each of a contrast of a high frequency component and a contrast of a low frequency component of a transformed radiographic image, a contrast in a gradient direction of an edge portion in an analysis region with each of analysis points set by an analysis point setting unit being the center of the analysis region. A ratio calculating unit calculates, for each gradient direction, a ratio of the contrast of the high frequency component to the contrast of the low frequency component. A determining unit determines the smallest ratio as an index indicating the body motion, and determines whether or not there is a body motion during an imaging operation to take the radiographic image based on a result of statistical processing of the indexes at the analysis points. A display control unit displays a result of the determination on a display unit.