2D X-Ray Measurement Correction for 3D Orientation Effects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The dependency of 2D measurement values on the 3D orientation of an organ in X-ray images leads to biased measurements and difficulties in comparing values to existing guidelines, especially when follow-up exams are conducted in different orientations, as shown by studies indicating significant deviations even with minor rotational changes.
Innovation Solution
A method and device for correcting 2D measurement values by predicting a corrected measurement value using a trained model that accounts for the 3D orientation of the examination object, employing landmark detection and probability density functions to estimate and adjust measurements based on a reference orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 2D measurement values are taken directly from X-ray images without correction, then the measurement process is simple and quick, but the measurement precision is reduced due to dependency on 3D organ orientation
Solution Approach 1:
The patent transforms the measurement approach by changing from direct 2D measurement to a corrected measurement that accounts for 3D orientation parameters. The trained model predicts corrected measurement values by considering the relationship between 3D organ orientation and 2D projection measurements, effectively adjusting the measurement parameters based on orientation data.
Solution Approach 2:
The patent introduces a trained prediction model as an intermediary between the raw 2D measurement and the final corrected measurement value. This model acts as a mediator that processes both the 2D measurement and 3D orientation information to produce the corrected measurement, resolving the contradiction between simplicity and precision.
2Adaptability or versatility
If follow-up exams are conducted in different 3D organ orientations, then patient monitoring is flexible and adaptable, but the reliability of comparison between exams is reduced due to measurement value deviations
Solution Approach 1:
Instead of requiring all exams to be conducted in a standard 3D orientation (traditional approach), the patent inverts the problem by predicting what the measurement value would be if the organ were in a reference orientation. This allows flexible acquisition of exams in various orientations while enabling reliable comparison through orientation-independent corrected values.
3Ease of manufacture
If landmark detection is performed in 2D image data without orientation correction, then the detection process is straightforward, but the measurement precision is affected by 3D orientation variations
Solution Approach 1:
The patent performs preliminary detection of landmarks in the 2D image data as usual, then uses these detected landmarks as input to the trained prediction model. The model subsequently corrects the measurement values based on the relationship between landmark positions and 3D organ orientation, maintaining the simplicity of landmark detection while improving measurement precision.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
A method for correcting a 2D measurement value (m) is described. According to the method, 2D image data (I) of an examination object are received. Further, landmarks (LM) in the 2D image data (I) are detected and the 2D positions (a) of the landmarks (LM) are calculated. Furthermore, a corrected measurement value (m^_pdf) of the examination object is predicted, using a trained model (Z(a, I, oref)), which depends on the received 2D-image data (I), the estimated 2D-positions (a) of the landmarks (LM) and a reference parameter (oref) of a reference 3D orientation of the examination object. Moreover, a correction device (50) is described. Furthermore, a medical imaging system (1) is described.