Purpose

Conventional Look-Locker (LL)-based estimation of the optimal inversion time (TI) (“TI-scout”) is a requirement prior to late gadolinium enhancement (LGE) imaging to achieve the appropriate image contrast especially when LGE images are acquired in a magnitude fashion. The TI-scout sequence has multiple limitations, including: the long breath-hold necessary, the collected images are in different cardiac phases, and the overly subjective and user-dependent TI estimation. Therefore, in this study we aimed to develop and test a quantitative T1-mapping-based synthetic inversion recovery (IR_synth) approach^1,2 allowing for the calculation of the most optimal TI for LGE imaging.

Methods

Forty consecutive patients (57±15 years, 23 male) referred for myocardial viability evaluation underwent cardiovascular MRI on a 1.5T system (MAGNETOM Avanto, Siemens, Erlangen, Germany). Twelve minutes after contrast administration (0.1mmol/kg gadobenate-dimeglumine) conventional TI-scout (LL, Field-of-view 300×256mm²; slice thickness 8mm; acquisition matrix 192×104; spatial resolution 2.08×2.08mm²; TR/TE 2.5/1.1ms; Bandwidth 965Hz/pixel; and Flip angle 50°) and T1-mapping (modified LL IR; scheme 4(1)3(1)2; Field-of-view 300×256mm²; slice thickness 8mm; acquisition matrix 196×128; spatial resolution 1.56×1.56mm²; TR/TE 2.6/1.1ms; Bandwidth 1085Hz/pixel; and Flip angle 35°) acquisitions were performed in the 4-chamber view. LL and T1 image sets were acquired in a random order and 2 subgroups were defined accordingly (Figure 1). In the LL group, T1-mapping was performed first followed by the conventional LL acquisition. In the Synthetic group, conventional LL was obtained before T1 mapping. In both groups the second acquisition was used to estimate the optimal TI for LGE imaging in order to reduce the time between the image set used for TI calculation and the LGE acquisition. Based on the T1-maps, magnitude IR_synth images were calculated in a TI range of 200-400 with 5ms increments real-time on the scanner (Figure 2). The optimal TI was determined from both LL and IR_synth images based on TI-Signal intensity curves (Figure 1). Image quality including the quality of nulling, the ability to differentiate LGE from blood, and the ability to differentiate LGE from normal myocardium was subjectively rated by two observers on a 3-point Likert scale. The quality of nulling was also objectively measured based on the myocardial/background signal ratio in the LGE images. The two groups were compared using the Kruskal-Wallis and Mann-Whitney tests.

Results

Out of the 40 patients enrolled in the study, LGE consistent with myocardial infarct was observed in 21 (52.5%) subjects. The optimal TI was measured significantly lower by the conventional LL approach compared to the IR_synth technique (LL group: 259±50ms vs. 290±53ms, P<0.0218; Synthetic group: 248±21ms vs. 287±29ms, P<0.0001). The acquisition order did not influence the value of optimal TI (LL group P=0.2540; Synthetic group P=0.8760). There was significant difference between the TI estimated based on the LL acquisition, and the TI actually applied for the LGE scan (259±50s vs. 285±52ms, P=0.0021) due to the need to account for LL correction (“fudge factor”). However, there was no difference between the TI estimated based on the IR_synth image sets and the TI used for the LGE scans. Using the IR_synth-based TI for LGE acquisition (n=21) provided significantly higher image quality ratings compared to LL-based LGE (n=19) for the quality of nulling (2.4 [2.0-2.7] vs. 1.8 [1.4-2.3], P=0.0044), however there was no difference in the ability to differentiate LGE from blood (2.5 [2.2-2.8] vs. 2.6 [2.1-3.0], P=0.2415) or LGE from normal myocardium (2.7 [2.3-3.0] vs. 2.5 [2.3-2.7], P=0.1194). Myocardial/background signal intensity ratio was lower in the Synthetic group compared to the LL group (1.2±0.4 vs. 2.4±1.1, P=0.0079).

Discussion

Our results indicate that T1-mapping based IR_synth imaging has the potential to accurately measure the optimal TI for LGE imaging. Subjective and objective image quality analysis showed that using the TI accurately determined based on the IR_synth image set yields better quality of nulling. The IR_synth method has multiple further advantages over the conventional LL approach. The IR_synth technique provides higher TI resolution as images can be generated at even 1ms intervals. This technique depends less on the operator as the actual TI used for LGE imaging can be retrospectively selected. Drawing regions of interest is more straightforward as all the images in an IR_synth set are in the same cardiac phase. Further advantage also reducing the subjectivity of the technique is that LL correction is already built in the T1-fitting algorithm³, thus there is no need to TI adjustment with the IR_synth technique, contrary to the conventional LL technique.

Conclusion

T1-based IR_synth imaging provides objective, quantitative, and real-time prescription of the optimal TI for LGE imaging; eliminating the need for LL correction and the substantial operator dependence of the acquisition.

Acknowledgements

No acknowledgement found.