Myocardial T1 and T2 quantification and water-fat separation using cardiac MR fingerprinting with rosette trajectories at 3T and 1.5T

Purpose: This work aims to develop an approach for simultaneous water-fat separation and myocardial T(1)and T(2)quantification based on the cardiac MR fingerprinting (cMRF) framework with rosette trajectories at 3T and 1.5T. Methods: Two 15-heartbeat cMRF sequences with different rosette trajectories designed for water-fat separation at 3T and 1.5T were implemented. Water T(1)and T(2)maps, water image, and fat image were generated with B(0)inhomogeneity correction using a B-0 map derived from the cMRF data themselves. The proposed water-fat separation rosette cMRF approach was validated in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology MRI system phantom and water/oil phantoms. It was also applied for myocardial tissue mapping of healthy subjects at both 3T and 1.5T. Results: Water T-1 and T-2 values measured using rosette cMRF in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology phantom agreed well with the reference values. In the water/oil phantom, oil was well suppressed in the water images and vice versa. Rosette cMRF yielded comparable T-1 but 2 similar to 3 ms higher T-2 values in the myocardium of healthy subjects than the original spiral cMRF method. Epicardial fat deposition was also clearly shown in the fat images. Conclusion: Rosette cMRF provides fat images along with myocardial T(1)and T-2 maps with significant fat suppression. This technique may improve visualization of the anatomical structure of the heart by separating water and fat and could provide value in diagnosing cardiac diseases associated with fibrofatty infiltration or epicardial fat accumulation. It also paves the way toward comprehensive myocardial tissue characterization in a single scan.

Automated summary

This study developed an approach for simultaneous water-fat separation and myocardial T(1) and T(2) quantification based on cardiac MR fingerprinting. The technique was validated in phantom studies and applied to healthy subjects, showing promising results for visualization of heart anatomy and potential value in diagnosing cardiac diseases.