Synopsis

In Dark Blood Late Gadolinium Enhanced (DB-LGE) imaging, simultaneous signal nulling of the healthy myocardium and the blood and thus yields superior contrast-to-noise ratio of the scared tissues. The method employs a T₂-preparation pulse applied after the inversion recovery pulse to rapidly damp the myocardium magnetization relative to the blood. Accurate timing of the inversion pulse, T₂-preparation pulse, and echo acquisition is essential for the success of the technique. In this work, we present a simple method for accurately estimating these parameters through fast low-resolution scouting scans played prior to DB-LGE scans.

Purpose

Methodology

Theory: Given arbitrary set of imaging parameters, D₁, D₂ and D₃, the myocardium signal in DB-LGE sequence is given by:

$$M_{myo}(D_1,D_2,D_3)=M_o(1-E_{3myo}+E_{2myo}E_{3myo}-2E_{1myo}E_{2myo}E_{3myo})$$

, where Mo is the fully recovered magnetization, $$$E_{1myo}=e^{-D_1/T_{1myo}}$$$, $$$E_{2myo}=e^{-D_2/T_{2myo}}$$$, $$$E_{3myo}=e^{-D_3/T_{1myo}}$$$.

Equating the above equation to zero, the time D₃ that nulls the myocardium can be written as a nonlinear function in D₁, and D₂; i.e., $$$D_{3myo}=f_{myo}(D_1,D_2)$$$. For the blood signal, a similar equation can be deduced: $$$D_{3blood}=f_{blood}(D_1,D_2)$$$.

Setting D₂ to a small value in order to maintain the highest possible scar signal, then to simultaneously null the blood and the myocardium, one needs to find D₁ such that: $$$\delta(D_1)=f_{blood}(D_1)-f_{myo}(D_1)=0$$$. A solution of this equation can be achieved using the iterative Newton-Raphson method:

$$D_{1,n+1}=D_{1,n}-\alpha.\delta(D_{1,n})$$

, where n is the iteration number and $$$\alpha$$$ is the inverse of the partial derivative of $$$\delta$$$ with respect to D₁. Instead of calculating $$$\alpha$$$ at each iteration, it is set to a fixed value determined analytically. First, it can be shown that $$$ \alpha\simeq(-\frac{2E_{2blood}}{1+E_{2blood}} +\frac{2E_{2myo}}{1+E_{2myo}})$$$ which is always negative (because T_2blood>>T_2myo). In fact, calculating the exact value of $$$\alpha$$$ for a wide range of (post-contrast) tissue parameter values reveals that the absolute of its optimal value is always above 3 (this was determined by randomly selecting 10000 different tissue parameter combinations from a pool of all possible (post-contrast) tissue parameter values and calculating for each combination). Therefore, we choose to set $$$\alpha$$$=-3 in all scouting scans. This guarantees convergence to the optimal parameter D₁ that yields an image with both the blood and the myocardial signal are null.

Phantom Experiment: A phantom of nine NiCl₂-doped agarose vials with different T₁ and T₂ values was used to test the developed method. Two vials in the phantom have parameters that mimic those of (post-contrast) healthy myocardium and blood. The scout imaging parameters were as follows: FOV=200×200mm², voxel size=2×2×8mm³, TR/TE=2.6/1.3ms, bSSFP acquisition, 21 image with D₃ varied from 130 to 330ms, with flip angle=55°. In the first scout, D_1,0 was intentionally set to a value (= 40ms) that is much smaller than the anticipated correct value (~150). Then, the acquired sequence of scout images is observed to determine the D₃ times of nulling the blood and the myocardium, which in turn are subtracted to obtain $$$\delta$$$(D_1,0). The iterative equation (above) is then applied to calculate D_1,1. The process is then repeated until both the myocardium and the blood are nulled in a single image.

In-vivo Experiment: Scouting for the DB-LGE timing parameters was done on healthy subjects using 1.5T scanner after 35 minutes following Gd-contrast injection. The scouting imaging parameters were: FOV=320×320mm², voxel size=2×2×8mm³, TR/TE=6.2/3.0ms, flip angle=25°, SENSE rate=2, acq. window=80ms.

Results

Phantom: Figure 3.a shows the first scouting scan of the phantom where the signal of the myocardium and blood vials are nulled at D₃=150ms and 190ms, respectively. The second scout is thus performed with D_1,1=160ms, where it can be seen in Fig 3b that the signals of the myocardium- and blood- vials are simultaneously suppressed at D₃=150ms; that is the optimal timing parameters D₁/D₂/D₃ are 160/30/150ms.

In-vivo: Figure 4a shows the few images of the first scout (with D₁/D₂=115/25ms). The signals of the myocardium and blood are null at 80ms and 85ms, respectively. The second scout is thus performed with D₁=115ms, where it can be seen in Fig 4b that the both the myocardium- and blood- are simultaneously suppressed at D₃=85ms; i.e the optimal timing parameters are 115/25/85ms.

Conclusion

Acknowledgements

References