Motion Artifact Detection in Overlapping Medical Imaging Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging procedures, such as CT or MRI of the chest, often suffer from motion artifacts due to breathing, heartbeat, peristalsis, or patient movement, especially when breath-holding is not feasible, which complicates data acquisition and analysis.
Innovation Solution
A method to detect motion artifacts by generating a difference map from overlapping regions in two datasets acquired at different time periods and identifying connected regions with absolute values exceeding a threshold, allowing for the detection and localization of motion artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple shorter acquisitions are made at the same breathing phase and stitched together, then time-resolved data can be acquired and the breath hold requirement is reduced, but motion artifacts occur due to relative motion between subsequent acquisitions
Solution Approach 1:
The patent applies preliminary action by performing motion artifact detection before the stitching process. The method calculates a motion metric from the raw k-space data of multiple acquisitions before they are combined, allowing the system to identify and flag potential motion artifacts in advance. This enables preventive measures to be taken, such as excluding problematic acquisitions from the final stitched image, thereby maintaining image quality while still benefiting from faster multi-phase acquisition.
2Ease of operation
If multiple shorter acquisitions are made at the same breathing phase, then the breath hold requirement is reduced for sick patients or radiotherapy planning, but motion artifacts due to inconsistent breathing motion reduce the usability of the acquired data
Solution Approach 1:
The patent implements feedback by continuously monitoring the motion metric calculated from raw k-space data during the acquisition process. The system provides real-time feedback about the presence and severity of motion artifacts, allowing operators to make informed decisions about whether to repeat acquisitions or adjust scanning parameters. This feedback mechanism ensures that even when patients cannot hold their breath, the system can identify usable data and maintain reliability through iterative optimization.
3Productivity
If motion artifacts are not detected, then the acquisition process is simpler and faster, but undetected motion artifacts compromise the intended purpose of the imaging data
Solution Approach 1:
The patent replaces complex mechanical or manual inspection methods with an automated computational approach for motion artifact detection. By using algorithms that process raw k-space data directly and calculate motion metrics automatically, the system achieves high detection accuracy without adding significant processing time. The automated nature of this substitution maintains productivity while ensuring that even subtle motion artifacts are reliably detected and flagged.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method for detecting a motion artifact (12a-12k) in medical imaging data, which comprises a first acquired dataset (5a) representing a first region and a second acquired dataset (5b) representing a second region, wherein the first and the second region overlap in an overlap region, is provided. A first image (6a) representing the overlap region is generated depending on the first acquired dataset (5a) and a second image (6b) representing the overlap region is generated depending on the second acquired dataset (5b). A difference map (7) is generated by subtracting the first image (6a) and the second image (6b) from each other. The motion artifact (12a-12k) is detected by determining a connected region (8, 8a-8d) in the difference map (7), wherein an absolute value of the difference map (7) is equal to or greater than a predefined threshold value within the connected region (8, 8a-8d).