Segmented Magnetic Resonance Fingerprinting Dictionary Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic resonance fingerprinting (MRF) techniques face challenges with large dictionaries that require significant memory, storage, and processing time, and existing methods like pre-grouping and singular value decomposition can be resource-intensive and risk truncating signals needed for accurate matching.
Innovation Solution
The use of sub-dictionaries allows for efficient matching of MRF data by dividing the dictionary into smaller, sequentially compared sub-dictionaries, reducing computational resources and time needed to create quantitative parameter maps, even with limited memory or processing power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large dictionary is used to cover the full range of tissue parameters, then matching accuracy is improved, but memory requirements and processing time increase significantly
Solution Approach 1:
The patent divides the large dictionary into multiple smaller sub-dictionaries based on tissue parameter ranges (e.g., T1, T2, proton density intervals). Each sub-dictionary covers a specific parameter range, allowing the system to process only relevant subsets during matching, thereby reducing processing time while maintaining comprehensive coverage for accuracy.
Solution Approach 2:
The patent pre-computes and stores signal templates in organized sub-dictionaries before the actual matching process. By preparing the dictionary structure in advance with proper segmentation and indexing, the system enables faster query-time matching without compromising the ability to accurately match against the full parameter range.
2Measurement precision
If a large dictionary is used to cover the full range of tissue parameters, then matching accuracy is improved, but memory storage requirements increase
Solution Approach 1:
The patent segments the complete dictionary into multiple smaller sub-dictionaries that can be stored in memory-efficient formats. Each sub-dictionary contains templates for a specific tissue parameter range, reducing the memory footprint of any single data structure while collectively covering the entire parameter space needed for accurate matching.
Solution Approach 2:
The patent loads only the necessary sub-dictionaries into memory based on the specific imaging sequence and expected tissue parameter ranges, rather than loading the entire dictionary at once. This allows the system to use partial dictionary coverage appropriate for each application, reducing memory requirements while maintaining sufficient accuracy for the given clinical context.
3Productivity
If pre-grouping and SVD truncation are used to reduce dictionary size, then processing time is reduced, but signal accuracy may be compromised
Solution Approach 1:
The patent avoids SVD truncation by using segmentation into sub-dictionaries based on tissue parameter ranges. This approach maintains the完整性 (integrity) of each signal template while organizing them for efficient processing, thereby preserving signal accuracy for matching while achieving speed improvements through reduced search space.
Solution Approach 2:
The patent performs preliminary organization of the dictionary into sub-dictionaries with proper indexing structures that enable efficient search without modifying or truncating the signal templates themselves. This pre-processing step achieves processing speed improvement through smart data organization rather than through signal approximation or truncation.
Data Source
AI summary
Systems and methods for producing a quantitative parameter map using a magnetic resonance imaging (MRI) system includes providing magnetic resonance fingerprinting (MRF) data acquired with an MRI system from a subject. The MRF data represents a plurality of different signal evolutions acquired using different acquisition parameter settings. The method also includes providing a database comprising a plurality of sub-dictionaries, each sub-dictionary including a plurality of signal templates, sequentially comparing the MRF data to each of the sub-dictionaries to estimate quantitative parameters, and generating a quantitative parameter map of the subject using the estimate quantitative parameters.