Motion Prediction in Digital Subtraction Angiography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for reducing motion artifacts in digital subtraction angiography images fail to accurately account for current frame motion, leading to persistent artifacts due to the use of displacement vectors calculated from past frames, which lag behind actual motion.
Innovation Solution
Calculating a virtual pixel-shift vector based on previously determined pixel-shift vectors to predict the motion in dynamic image data and adjust the reference image data accordingly, ensuring better alignment and reduced artifacts in the subtracted image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a pixel-shift vector is calculated based on a past frame of the x-ray image, then motion artifacts are reduced to an extent, but the displacement vector lags behind the actual motion in the current frame, resulting in persistent artifacts
Solution Approach 1:
The patent calculates a predicted pixel-shift vector for the current frame based on pixel-shift vectors from previous frames, rather than waiting to calculate it after motion occurs. This preliminary prediction allows the reference image data to be adjusted in advance to match the expected position of the dynamic image data, eliminating the time lag inherent in conventional methods that rely solely on past frame analysis.
Solution Approach 2:
The patent uses previously calculated pixel-shift vectors as feedback to predict the current frame's displacement. By analyzing the motion pattern from past frames and feeding this information back into the prediction algorithm, the system continuously refines its estimate of current motion, improving alignment accuracy without sacrificing real-time performance.
2Device complexity
If conventional subtraction methods are used with past-frame displacement vectors, then processing is simpler, but motion artifacts persist due to incorrect displacement vectors when current frame motion differs from past frame motion
Solution Approach 1:
The patent transitions from a static displacement vector (based on a single past frame) to a dynamic predicted pixel-shift vector that adapts to changing motion conditions. The prediction algorithm continuously updates the displacement estimate based on the most recent motion patterns, allowing the system to respond to varying motion conditions while maintaining computational efficiency.
Data Source
AI summary
A system and method for calculating a pixel-shift vector that predicts the motion in dynamic image data and can be used to shift reference image data in order to generate a subtracted image with reduced motion artifacts. The pixel-shift vector is calculated based on displacement vectors that have been calculated based on past motion in the dynamic image data. One embodiment determines a virtual pixel-shift vector as a function of stored pixel-shift vectors that were previously determined as a function of a previously captured live frame. The virtual pixel-shift vector is a prediction of the location of the current live frame before the current live frame is captured. The mask frame is adjusted as a function of this virtual pixel-shift vector. The current live frame is then subtracted from the mask frame adjusted as a function of the virtual pixel-shift vector. The adjusted mask frame and the live frame will then overlap and the subtracted image will have reduced artifacts.