Student Tsinghua University, China (People's Republic)
Background: Cardiac T1 mapping has been routinely used for assessment on myocardium heterogeneity.1 The conventional SASHA2 is a breath-holding sequence, which challenges for patients with poor compliance and may induce physiological instability. Given the nature of breath hold, conventional SASHA has limited spatial coverage. The slice tracking technique3, using diaphragm navigator, has been proposed for real-time slice localization with free breathing. Therefore, this study aims to investigate the feasibility of free-breathing SASHA sequence (fbSASHA) with slice tracking for accurate whole-heart T1 measurement.
Methods: The conventional SASHA was extended to free-breathing protocol with two key modifications, as depicted in Figure 1. First, respiratory synchronization was achieved through the application of slice tracking. Second, acquisition of T1-weighted images for different slices was performed in an interleaved order to minimize the disturbance from irregular breathing pattern. We validated fbSASHA on a 3T Philips Ingenia CX scanner by in-vivo studies. A cohort of twenty healthy volunteers (Age: 34 ± 12 years, 9 males) was enrolled. Within the similar scan time, the fbSASHA sequence covered the left ventricular (LV) with nine short-axis (SAX) slices. As the comparison, the conventional SASHA finished three SAX slices. Except for the free-breathing modification, all the other imaging parameters were kept the same. Same post-processing was used for both sequences. In brief, nonrigid motion correction was performed before the pixel-wised T1 map fitting using a two-parameter saturation-recovery model. The myocardium was manually segmented for each slice. Papillary muscle and confounding blood were carefully excluded. AHA 16-segment model was used to assess the homogeneity distribution through the LV. The mean, standard deviation (SD) and coefficient of variation (CV) of T1 values were calculated to evaluate the accuracy and precision.
Results: Homogeneous T1 maps were obtained from fbSASHA (Figure 2A), with evenly distributed T1 measurements across the entire left ventricular coverage (Figure 2B, C). The fbSASHA yielded comparable image quality to SASHA (Figure 3A), with consistent trends mean of T1 estimation among all segments (Figure 3B). Myocardial T1 measured by fbSASHA was slightly lower than SASHA (1420 ± 55 ms vs 1537 ± 57 ms), with comparable precision (11.1 ± 3.2% vs 9.9 ± 2.1%). The underestimation might come from the different physiology stability between the breathing strategies, though it is still under investigation.
Conclusion: The free-breathing SASHA sequence enabled by slice tracking technique achieves comparable performance to the conventional breath-hold SASHA. The free-breathing SASHA is promising to benefit the patient compliance for whole-heart T1 mapping.
