Introduction

SMART1Map (Saturation Method using Adaptive Recovery Times for cardiac T₁ Mapping) is a single-point saturation-recovery balanced SSFP technique [1] that directly measures true T₁. Native T1 changes reflect myocardial pathologies related to excess of deposition of T₁-altering substances, including water, protein, lipid, or iron [2]. The extracellular volume (ECV) can be calculated from T₁ values and is a frequently used parameter to estimate the degree of myocardial alteration [3]. The accuracy of T₁-mapping have been reported for different T1 values [4]. No study has investigated the influence of imaging parameters on these measures. This study aimed to assess the accuracy of SMART1Map by using different acquisition parameters (spatial resolution, heart rate, time delay), and to describe their influence on the measured T₁ values.

Material and Methods

Six phantoms (P0, P1, P2, P3, P4, P5, 500 mL each) were filled with water plus a dose of Gadolinium (0.5 mmol/mL) at respectively 0.0, 0.1, 0.2, 0.3, 0.4, 0.5 mL. All images were acquired on a GE450W 1.5T scanner and 32 channel cardiac coil. The T1 values were measured using standard FSE-IR and 35 different TI values (range 100 ms to 2500ms) by fitting the curve [SI vs TI, T₁= TI/Ln(2)]. The imaging parameters were axial slice (isocenter), thickness 8 mm, TR = 15000ms, 16 ETL (180°), Xres=192, Yres=128, pFOV=75%. In addition, ten SMART1Map measurements were performed (time interval between 2 consecutive exams ranges from 1 day to 1 week). The imaging parameters matched the FSE-IR: axial image, 8mm thickness, flip angle 65°, Xres = 192, Yres=128, BW=100 kHz, pFOV=75%, TE = min full (1.4 ms). Accuracy and reproducibility were tested with 16 different acquisition schemes: Two time delays (systole and diastole), two spatial resolutions (FOV= 30 and 48 cm), and 4 different heart rates (60, 80, 100, and 120 bpm). The saturation-recovery curve-fits for SMART1Map were carried out using Medis software (Leiden, the Netherlands).

Results

The reference T1 values (FSE-IR) of the phantoms ranged between 2550 ms for pure water (P0, no Gd), and 376 ms for P5 = 500 mL water + 0.5 mL Gd (Table 1). Inter-sequence comparison proved that at short T₁ (range 750, ±17 to 376±7; P2-P5) SMART1Map measures accurate and reproducible values at any heart rate, FOV or cardiac-phase. . In contrast for long T₁ (2550 ± 50 ms, and 1350 ± 40 ms for P1), SMART1Map systematically underestimates the T₁ values (respectively 1985 ± 73 ms and 1240 ± 21 ms) regardless of the heart rate (60, 80, 100, 120 bpm), spatial resolution (FOV=30 or 48 cm), or cardiac phases (systole, diastole). Whisker curves analysis (Figure 1) shows that SMART1Map results matches quite well with myocardium native (T₁ = 1350ms) and post contrast (T₁ = 376 ms). Calculation of the standard errors of mean (SEM) demonstrated an important influence of the spatial resolution on the accuracy of the T₁ measurements on all phantoms, regardless of the cardiac phases and the heart rate (Figure 2). For a smaller FOV (30 vs 48 cm, all other parameters equal) increases the T₁ SEM values, on average, from 26ms to 39ms at long T₁ = 1240 ms (pre contrast), and from 7 ms to 15 ms at short T₁ = 376ms (post contrast). In addition, the acquisition phase (diastole vs systole) influences the accuracy of the T₁ measurements. At short T₁ (post-contrast), the SEM is smaller in diastole than in systole, all other parameters equal. (Figure 3). At long T₁ (native values), the SEM is smaller in systole than in diastole. Heart rate has no influence on the accuracy of the SMART1Map measures.

Discussion and Conclusion

The reported normal pre and post contrast T₁ values of the myocardium at 1.5T are around 1100 ms and 300 ms, respectively. We demonstrated that SMART1Map provides accurate measures of post-contrast T₁ of the myocardium, but it systematically underestimates native pre-contrast T₁ values (5 to 10%). These errors can be reduced by prescribing larger FOV and acquiring pre-contrast images during systole and post-contrast images during diastole. Although the effect of these under-estimations is lower on the ECV values (due to the ratio myocardium T₁ / blood T₁), performing SMART1Map using optimized imaging parameters improves the accuracy of T₁-mapping and may lead to improved clinical diagnosis of disease severity.

Acknowledgements

No acknowledgement found.

References

[1] Slavin, Proc. ISMRM, (2012), p.1244.

[2] Riesenkampff et al., Circ Cardiovasc Imaging. (2015).

[3] Moon et al., JCMR, (2013).

[4] Kellman and Hansen JCMR, (2014)