Efficient pulse sequence design framework for high-dimensional MR fingerprinting scans using systematic error index.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Magnetic Resonance Fingerprinting (MRF) enables quantitative tissue property mapping.
• Optimizing MRF pulse sequences is critical for accuracy but computationally intensive for high-dimensional models.
• Existing virtual-scan simulations are too slow for complex MRF frameworks.

Purpose of the Study:

• To introduce a novel mathematical model, the Systematic Error Index (SEI), for efficient MRF sequence optimization.
• To address the scalability challenges in high-dimensional MRF sequence design.
• To enable practical optimization of MRF sequences considering complex tissue property interactions.

Main Methods:

• Developed the Systematic Error Index (SEI) model to approximate quantification errors with low computational cost.
• Eliminated the need for computationally expensive dictionary matching.
• Validated the SEI model against virtual-scan simulations.
• Applied the SEI model to optimize high-dimensional MRF sequences (2-4 tissue properties).

Main Results:

• The SEI model closely approximated virtual-scan simulation results.
• Achieved a hundred- to thousand-fold acceleration in computational speed.
• Optimized MRF sequences demonstrated higher measurement accuracy and fewer undersampling artifacts.
• Optimized scans achieved shorter scan times compared to heuristically designed sequences.

Conclusions:

• An efficient method for estimating real-world errors in MRF scans was developed.
• The SEI model provides accurate qualitative and quantitative error approximation.
• Demonstrated the practicality of the SEI model for optimizing high-dimensional MRF sequences.
• Optimized MRF scans show enhanced robustness to undersampling and system imperfections with faster acquisition.