Streamlined magnetic resonance fingerprinting: Fast whole-brain coverage with deep-learning based parameter estimation.

Magnetic resonance fingerprinting (MRF) is a quantitative MRI (qMRI) framework that provides simultaneous estimates of multiple relaxation parameters as well as metrics of field inhomogeneity in a single acquisition. However, current challenges exist in the forms of (1) scan time; (2) need for custom image reconstruction; (3) large dictionary sizes; (4) long dictionary-matching time. This study aims to introduce a novel streamlined magnetic-resonance fingerprinting (sMRF) framework based on a single-shot echo-planar imaging (EPI) sequence to simultaneously estimate tissue T1, T2, and T2* with integrated B1+ correction. Encouraged by recent work on EPI-based MRF, we developed a method that combines spin-echo EPI with gradient-echo EPI to achieve T2 in addition to T1 and T2* quantification. To this design, we add simultaneous multi-slice (SMS) acceleration to enable full-brain coverage in a few minutes. Moreover, in the parameter-estimation step, we use deep learning to train a deep neural network (DNN) to accelerate the estimation process by orders of magnitude. Notably, due to the high image quality of the EPI scans, the training process can rely simply on Bloch-simulated data. The DNN also removes the need for storing large dictionaries. Phantom scans along with in-vivo multi-slice scans from seven healthy volunteers were acquired with resolutions of 1.1×1.1×3 mm3 and 1.7×1.7×3 mm3, and the results were validated against ground truth measurements. Excellent correspondence was found between our T1, T2, and T2* estimates and results obtained from standard approaches. In the phantom scan, a strong linear relationship (R = 1-1.04, R2>0.96) was found for all parameter estimates, with a particularly high agreement for T2 estimation (R2>0.99). Similar findings are reported for the in-vivo human data for all of our parameter estimates. Incorporation of DNN results in a reduction of parameter estimation time on the order of 1000 x and a reduction in storage requirements on the order of 2500 x while achieving highly similar results as conventional dictionary matching (%differences of 7.4 ± 0.4%, 3.6 ± 0.3% and 6.0 ± 0.4% error in T1, T2, and T2* estimation). Thus, sMRF has the potential to be the method of choice for future MRF studies by providing ease of implementation, fast whole-brain coverage, and ultra-fast T1/T2/T2* estimation.

A robust SSFP technique for fMRI at ultra-high field strengths.

A non-balanced (nb) SSFP-based fMRI method based on CE-FAST is presented to alleviate some shortcomings of high spatial-specificity techniques commonly used in high static magnetic fields. The proposed sequence does not suffer from the banding artifacts inherent to balanced (b) SSFP, has low geometrical distortions and SAR compared to spin-echo EPI, and in contrast to previous nbSSFP implementations, is applied at a TR, theoretically prescribed for the optimum contrast. Its non-balanced gradient was chosen to just dephase the unwanted signal component (2π dephasing per TR per voxel). 3D data were acquired from nine healthy subjects, who performed a visual-motor task on a 7 Tesla scanner. For comparison, experiments were accompanied by similar bSSFP and spin-echo acquisitions. Consistent activation was achieved in all subjects with theoretically optimal TR, in contrast to previous nbSSFP techniques. The signal stability as well as relative and absolute functional signal changes, were found to be comparable with bSSFP and spin-echo techniques. The results suggest that with suitable modifications, CE-FAST can be regarded as a robust SSFP-based method for high spatial specificity fMRI techniques.

Investigating the spatial specificity of S2-SSFP fMRI: A Monte Carlo simulation approach.

The desirable spatial specificity of spin echo (SE) fMRI cannot be efficiently utilized in high fields due to specific absorption rate (SAR) and B1 inhomogeneity problems. Consequently, S2-SSFP fMRI has been suggested as an alternative to mitigate these problems. Nevertheless, no accurate analysis has been performed thus far to evaluate spatial specificity of this technique. To study spatial specificity, we performed Monte Carlo simulations for evaluating the micro-vasculature contribution in functional contrast along with vessel size sensitivity estimations for a range of relevant imaging parameters. Results showed a spatial specificity at the level of SE fMRI. Simulations further revealed that similar to SE fMRI, an effective echo time (TE) close to the tissue T2 maximizes the micro-vasculature contribution in the obtained contrast. The amount of this contribution, however, showed a slight decrease at ultra-high fields compared to SE fMRI. As for vessel size sensitivity, simulations presented a pattern for S2-SSFP similar to SE fMRI but with a minor shift toward larger vessels. These results are in general agreement with reported experimental studies. Our findings also suggest that the effect of older pathways, rather than primary SE pathway, might be responsible for the observed discrepancies between S2 and SE. Based on this study, provided that optimum experimental parameters are used, S2, with its desirable micro-vasculature contribution and high sensitivity to small vessels, is a promising low SAR approach to replace SE fMRI in high field.