Comparison of different cardiac MRI sequences at 1.5 T/3.0 T with respect to signal-to-noise and contrast-to-noise ratios - Initial experience

Purpose: To compare image quality, signal-to-noise (SNR) and contrast-to-noise ratios (CNR) of different MRI sequences for cardiac imaging at 1.5 T and 3.0 T in volunteers. Material and Methods: 10 volunteers (5 male, 5 female) with a mean age of 33 years ( 8) without any history of cardiac diseases were examined on a GE Signa 3.0 T and a GE Signa 1.5 T TwinSpeed Excite (GE Medical Systems, Milwaukee, WI, USA) scanner using a 4-element phased array surface coil (same design) on the same day. For tissue characterization ECG gated Fast Spinecho (FSE) T-1- (Double IR), T-1- STIR (Triple IR) and T-2-weighted sequences in transverse orientation were used. For functional analysis a steady state free precession (SSFP - FIESTA) sequence was performed in the 4-chamber, 2-chamber long axis and short axis view. The flip angle used for the SSFP sequence at 3.0 T was reduced from 45degrees to 30degrees to keep short TR times while staying within the pre-defined SAR limitations. All other sequence parameters were kept constant. Results: All acquisitions could successfully be completed for the 10 volunteers. The mean SNR 3.0 T compared to 1.5 T was remarkably increased (p < 0.05) for the T-2 - (160 % SNR increase), the STIR-T-1- (123 %) and the T-1- (91 %) weighted FSE. Similar results were found comparing CNR at 3.0 T and 1.5 T. The mean SNR achieved using the SSFP sequences was more than doubled by 3.0 T (150%), but did not have any significant effect on the CNR. The image quality at 3.0 T did not appear to be improved, and was considered to be significantly worse when using SSFP sequences. Artefacts like shading in the area of the right ventricle (RV) were found to be more present at 3.0 T using FSE sequences. After a localized shim had been performed in 5/10 volunteers at the infero-lateral wall of the left ventricle (LV) with the SSFP sequences at 3.0 T no significant increase in artefacts could be detected. Conclusions: In all cardiac FSE sequences, SNR and CNR at 3.0 T were found to be increased compared to 1.5 T without any major changes of the sequence parameters. The adjusted SSFP sequences fulfilled the expected increase in SNR at 3.0 T but showed no increase in CNR. On the contrary, the overall image quality did not change or was even found to be significantly lower for the SSFP and the FSE sequences at the free wall of the RV. Nevertheless, the results are encouraging for the use of 3.0 T for cardiac tissue characterization and new applications with progressing use of parallel imaging.