Accelerated Cardiac CINE MR Imaging on Multi-Channel 3 Tesla: Comparison of Signal Parameters and Volumetric Accuracy to 1.5 Tesla

Purpose: The purpose of this study is to assess the impact of a higher magnetic field strength (3T) on signal parameters of non-accelerated and accelerated CINE SSFP and its volumetric accuracy compared to 1.5T field strength.

Methods: 10 individuals including 9 patients with history of myocardial infarction and one healthy volunteer were enrolled to the study (age 28–63; 49 years). All individuals underwent CINE imaging at 1.5T and 3T, both scanners equipped with 32 independent receiver channels and an equivalent array-coil matrix with a selection of 12 coil-elements.

Data sampling was performed in short axis orientation using standard single-slice CINE and multi-slice TSENSE accelerated CINE (5 slices/breath-hold) with 4-fold acceleration (R=4). All standard sequence parameters regarding spatial resolution were kept constant for 1.5T and 3T with a FOV of 360×293mm2 and a matrix of 192×117. The flip angle was held constant at 60° at 1.5T while at 3T it was maximized with individual adaptation to normal operating mode specific absorption rate (SAR) limits (1.5W/kg). Data was evaluated for left ventricular volumetric parameters (EDV, ESV and EF) as well as for SNR and CNR. Phantom based g-factor evaluation allowed for assessment of noise levels for accelerated data sets. Volumetric results and signal parameters were compared to results of single-slice CINE SSFP at 1.5T as standard of reference (SOR).

Results: At 3T a flip angle of 54–60°was achieved for TSENSE accelerated multi-slice CINE while flip angle was limited by SAR to 39–51° for non-accelerated CINE (P=0.002). SNR of both blood (51.7±8.5 vs. 102.6±34.5; P=0.0004) and myocardium (12.6±2.0 vs. 27.5±10.9; P=0.0009) show a significant increase in non-accelerated data from 1.5T to 3T. Single-slice CINE at 3T showed a ~90% increase in CNR compared to the SOR (P<0.001). At 3T, TSENSE CINE showed a less pronounced loss in CNR (-58±6%) compared to single-slice CINE than at 1.5T (-71±2%). 3T TSENSE CINE showed a 21±18% lower CNR than the non-accelerated 1.5T CINE (P<0.05). Mean differences of EF results were -0.3% (P=0.51) for accelerated 1.5T CINE (R=4), 1.1% (P=0.36) for non-accelerated 3T CINE and 0.6% (P=0.47) for accelerated 3T CINE (R=4). EDV mean differences to SOR were -5.1ml (P=0.009), 11.0ml (P=0.002) and 1.5ml (P=0.69), respectively.

Conclusion: 3T allows compensating for the high CNR loss coming along with 4-fold TSENSE acceleration at 1.5T and shows volumetric accuracy. The use of parallel imaging may help to alleviate SAR limitations at higher field strength.