Recurrent neural network-based simultaneous cardiac T1, T2, and T1ρ mapping.

The purpose of the current study was to explore the feasibility of training a deep neural network to accelerate the process of generating T1, T2, and T1ρ maps for a recently proposed free-breathing cardiac multiparametric mapping technique, where a recurrent neural network (RNN) was utilized to exploit the temporal correlation among the multicontrast images. The RNN-based model was developed for rapid and accurate T1, T2, and T1ρ estimation. Bloch simulation was performed to simulate a dataset of more than 10 million signals and time correspondences with different noise levels for network training. The proposed RNN-based method was compared with a dictionary-matching method and a conventional mapping method to evaluate the model's effectiveness in phantom and in vivo studies at 3 T, respectively. In phantom studies, the RNN-based method and the dictionary-matching method achieved similar accuracy and precision in T1, T2, and T1ρ estimations. In in vivo studies, the estimated T1, T2, and T1ρ values obtained by the two methods achieved similar accuracy and precision for 10 healthy volunteers (T1: 1228.70 ± 53.80 vs. 1228.34 ± 52.91 ms, p > 0.1; T2: 40.70 ± 2.89 vs. 41.19 ± 2.91 ms, p > 0.1; T1ρ: 45.09 ± 4.47 vs. 45.23 ± 4.65 ms, p > 0.1). The RNN-based method can generate cardiac multiparameter quantitative maps simultaneously in just 2 s, achieving 60-fold acceleration compared with the dictionary-matching method. The RNN-accelerated method offers an almost instantaneous approach for reconstructing accurate T1, T2, and T1ρ maps, being much more efficient than the dictionary-matching method for the free-breathing multiparametric cardiac mapping technique, which may pave the way for inline mapping in clinical applications.

Accelerated 3D free-breathing high-resolution myocardial T1ρ mapping at 3 Tesla.

PURPOSE: To develop a fast free-breathing whole-heart high-resolution myocardial T1ρ mapping technique with robust spin-lock preparation that can be performed at 3 Tesla. METHODS: An adiabatically excited continuous-wave spin-lock module, insensitive to field inhomogeneities, was implemented with an electrocardiogram-triggered low-flip angle spoiled gradient echo sequence with variable-density 3D Cartesian undersampling at a 3 Tesla whole-body scanner. A saturation pulse was performed at the beginning of each cardiac cycle to null the magnetization before T1ρ preparation. Multiple T1ρ -weighted images were acquired with T1ρ preparations with different spin-lock times in an interleaved fashion. Respiratory self-gating approach was adopted along with localized autofocus to enable 3D translational motion correction of the data acquired in each heartbeat. After motion correction, multi-contrast locally low-rank reconstruction was performed to reduce undersampling artifacts. The accuracy and feasibility of the 3D T1ρ mapping technique was investigated in phantoms and in vivo in 10 healthy subjects compared with the 2D T1ρ mapping. RESULTS: The 3D T1ρ mapping technique provided similar phantom T1ρ measurements in the range of 25-120 ms to the 2D T1ρ mapping reference over a wide range of simulated heart rates. With the robust adiabatically excited continuous-wave spin-lock preparation, good quality 2D and 3D in vivo T1ρ -weighted images and T1ρ maps were obtained. Myocardial T1ρ values with the 3D T1ρ mapping were slightly longer than 2D breath-hold measurements (septal T1ρ : 52.7 ± 1.4 ms vs. 50.2 ± 1.8 ms, P < 0.01). CONCLUSION: A fast 3D free-breathing whole-heart T1ρ mapping technique was proposed for T1ρ quantification at 3 T with isotropic spatial resolution (2 mm3 ) and short scan time of ∼4.5 min.