MRI Phase Unwrapping via Quality Map Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for phase unwrapping in magnetic resonance imaging (MRI) are inaccurate and lack robustness, leading to artifacts in images due to phase wrapping, which affects the quality of the generated images.
Innovation Solution
A method and system for phase unwrapping that involves obtaining echo signals at different echo times, generating phase wrapped images, creating a phase difference map, unwrapping it to determine a field distribution map, and then unwrapping phase images using this map, with additional steps involving filtering and quality map generation to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase wrapping is performed using inverse trigonometric functions, then the phase can be extracted from echo signals, but the phase becomes wrapped into a range of 2π causing artifacts in images
Solution Approach 1:
The patent introduces an intermediary quality map that evaluates the reliability of phase differences at each pixel location. This quality map serves as a mediator between the wrapped phase data and the unwrapping process, allowing the system to identify reliable regions for unwrapping while avoiding unreliable regions that would generate artifacts. The quality map is computed based on signal strength and phase consistency metrics, enabling selective unwrapping only where confidence is high.
Solution Approach 2:
The patent applies local quality assessment by computing pixel-specific quality values in the quality map that reflect the local reliability of phase measurements. Different regions of the image are treated differently based on their local signal characteristics - high-quality regions undergo unwrapping while low-quality regions are preserved or discarded. This local differentiation prevents artifacts from propagating across the entire image while maintaining accuracy in reliable regions.
2Reliability
If existing phase unwrapping methods are applied, then wrapped phase values can be corrected, but the methods lack accuracy and robustness
Solution Approach 1:
The patent performs preliminary actions by computing the quality map and filtering the phase difference map before executing the unwrapping operation. The quality map is pre-computed to identify reliable pixels, and the phase difference map is filtered to remove noise and inconsistencies. These preliminary steps prepare the data in a state that enables accurate and robust unwrapping, preventing errors from propagating through the subsequent processing stages.
Solution Approach 2:
The patent implements feedback mechanisms where the quality map continuously informs the unwrapping process about which pixels are reliable. The unwrapping algorithm uses quality thresholds to decide whether to unwrap or discard each pixel, creating a feedback loop that adapts to local signal conditions. This feedback ensures that unwrapping is only performed when confidence is sufficient, maintaining both accuracy and robustness.
3Ease of manufacture
If phase unwrapping is performed without quality assessment, then processing is simpler, but accuracy and robustness are reduced
Solution Approach 1:
The patent segments the phase unwrapping process into distinct stages: quality map computation, phase difference map filtering, and conditional unwrapping. Each stage handles a specific aspect of the problem independently. The quality map segmentation separates reliable pixels from unreliable ones, allowing the unwrapping algorithm to focus computational resources only on pixels that will contribute meaningfully to the final image, maintaining accuracy without excessive complexity.
Data Source
AI summary
The present disclosure relates to systems and methods for vehicle management. The systems may perform the methods to obtain a plurality of echo signals representing a subject, the plurality of echo signals being formed at a plurality of different echo times; generate a plurality of phase wrapped images based on the plurality of echo signals, each phase wrapped image comprising one or more first wrapped values; generate a phase difference map based on the plurality of phase wrapped images, the phase difference map comprising one or more second wrapped values; generate an unwrapped phase difference map by unwrapping the one or more second wrapped values in the phase difference map; determine a first field distribution map based on the unwrapped phase difference map; and generate a plurality of unwrapped phase images by unwrapping the one or more first wrapped values in each phase wrapped image based on the first field distribution map.