Comparison of GRE versus SSFP-based cardiac T1-mapping in device patients

Background Cardiac magnetic resonance (CMR) is challenging in patients implanted with metallic devices, such as pacemakers or defibrillators, given metallic susceptibility artifacts. The technique of T1-mapping investigates interstitial fibrosis. The most used method for myocardial T1-mapping is the modified Look-Locker Inversion recovery sequence (MOLLI) using balanced steady-state free precession (SSFP). However, SSFP is susceptible to off-resonance artifacts, leading to errors. Gradient echo (GRE) sequences are less prone to these artifacts. Purpose To investigate whether T1-mapping using GRE was comparable to SSFP in this population. Material and Methods Pre/post-contrast T1-mapping was performed on 16 devices utilizing MOLLI with SSFP and GRE strategies at 1.5 T, as well as 10 non-device controls. The difference in mean T1 time by SSFP versus GRE (both pre- and post-contrast for each slice) for device patients was analyzed. Results Native T1 for the device cohort was 1053 ± 94 ms for SSFP and 969 ± 83 ms for GRE. GRE T1 measurements were shorter than SSFP measurements (difference over all slices for SSFP vs. GRE pre-/post-contrast were 74 ms and 27 ms). Mean λ for GRE (devices) was 0.520 ± 0.194 (p = 0.30). λ for SSFP (devices) was 0.536 ± 0.124 ( P = 0.08). There was no difference in λ between SSFP versus GRE in device patients ( P = 0.91). The percentage of segments identified as artifact on T1 maps for device patients was 24% and 45% for GRE and SSFP, respectively. Conclusion Compared to SSFP, T1 values using GRE were consistently shorter, while λ values remained equivalent. There was less visual artifact on GRE images, suggesting advantageous utility over SSFP in patients with cardiac devices.