Flip Angle Imaging Research Articles

Optimal radiofrequency and gradient spoiling for improved accuracy of T1 and B1 measurements using fast steady‐state techniques

Variable flip angle T(1) mapping and actual flip-angle imaging B(1) mapping are widely used quantitative MRI methods employing radiofrequency spoiled gradient-echo pulse sequences. Incomplete elimination of the transverse magnetization in these sequences has been found to be a critical source of T(1) and B(1) measurement errors. In this study, comprehensive theoretical analysis of spoiling-related errors in variable flip angle and actual flip-angle imaging methods was performed using the combined isochromat summation and diffusion propagator model and validated by phantom experiments. The key theoretical conclusion is that correct interpretation of spoiling phenomena in fast gradient-echo sequences requires accurate consideration of the diffusion effect. A general strategy for improvement of T(1) and B(1) measurement accuracy was proposed based on the strong spoiling regimen, where diffusion-modulated spatial averaging of isochromats becomes a dominant factor determining magnetization evolution. Practical implementation of strongly spoiled variable flip angle and actual flip-angle imaging techniques requires sufficiently large spoiling gradient areas (A(G)) in combination with optimal radiofrequency phase increments (phi(0)). Optimal regimens providing <2% relative T(1) and B(1) measurement errors in a variety of tissues were theoretically derived for prospective in vivo variable flip angle (pulse repetition time = 15-20 ms, A(G) = 280-450 mT.ms/m, phi(0) = 169 degrees) and actual flip-angle imaging (pulse repetition time(1)/pulse repetition time(2) = 20/100 ms, A(G1)/A(G2) = 450/2250 mT.ms/m, phi(0) = 39 degrees) applications based on 25 mT/m maximal available gradient strength.

Cine cardiac imaging using black‐blood steady‐state free precession (BB‐SSFP) at 3T

To propose a new black-blood (BB) pulse sequence that provides BB cine cardiac images with high blood-myocardium contrast. The proposed technique is based on the conventional steady-state free precession (SSFP) sequence. Numerical simulations of the Bloch equation were conducted to compare the resulting signal-to-noise ratio (SNR) to that of conventional BB imaging, including the effects of changing the imaging flip angle and heart rates. Simulation results were verified using a gel phantom experiment and five normal volunteers were scanned using the proposed technique. The new sequence showed higher SNR and contrast-to-noise ratio (CNR) (approximately 100%) compared to the conventional BB imaging. Also, the borders of the left ventricle (LV) and right ventricle (RV) appear more distinguishable than the conventional SSFP. We were also able to cover about 80% of the cardiac cycle with short breath-hold time (approximately 10 cardiac cycles) and with reasonable SNR and CNR. Based on an SSFP conventional sequence, the new sequence provides BB cines that cover most of the cardiac cycle and with higher SNR and CNR than the conventional BB sequences.

Joint brain parametric T1-map segmentation and RF inhomogeneity calibration.

We propose a constrained version of Mumford and Shah's (1989) segmentation model with an information-theoretic point of view in order to devise a systematic procedure to segment brain magnetic resonance imaging (MRI) data for parametric T1-Map and T1-weighted images, in both 2-D and 3D settings. Incorporation of a tuning weight in particular adds a probabilistic flavor to our segmentation method, and makes the 3-tissue segmentation possible. Moreover, we proposed a novel method to jointly segment the T1-Map and calibrate RF Inhomogeneity (JSRIC). This method assumes the average T1 value of white matter is the same across transverse slices in the central brain region, and JSRIC is able to rectify the flip angles to generate calibrated T1-Maps. In order to generate an accurate T1-Map, the determination of optimal flip-angles and the registration of flip-angle images are examined. Our JSRIC method is validated on two human subjects in the 2D T1-Map modality and our segmentation method is validated by two public databases, BrainWeb and IBSR, of T1-weighted modality in the 3D setting.

On the steady‐state properties of actual flip angle imaging (AFI)

AFI (actual flip angle imaging) represents an interesting approach to map the B(1) transmit fields by measuring the spatial variations of the effective flip angle. However, the accuracy of the technique relies on the adequate spoiling of transverse magnetization. In the present work configuration theory was employed to develop a proper RF and gradient spoiling scheme for the AFI technique, making the sequence robust against off-resonance without the need of large spoiling gradients. Furthermore, numerical simulations were performed to predict the steady-state signals and, hence, the accuracy of the AFI technique as a function of the sequence and tissue parameters. It is shown that the spoiling properties of the sequence are mainly defined by the phase shift increment phi of the RF pulses and the diffusion sensitivity resulting from the unbalanced gradients of the sequence. Adequate spoiling may be achieved for a reasonable range of tissue parameters and flip angles for moderate spoiling gradients if a favorable value for phi is chosen. Phantom and in vivo head imaging experiments show an excellent agreement with the theoretical predictions, indicating that the proper operating range of the approach may be reliably predicted by the theory.

Rapid 3D‐T1 mapping of cartilage with variable flip angle and parallel imaging at 3.0T

To rapidly acquire T(1)-weighted images using a three-dimensional fast low angle shot (3D FLASH) sequence in combination with generalized autocalibrating partially parallel acquisitions (GRAPPA) and variable flip angle (VFA) method at 3.0T. 3D T(1) maps of model systems (gadolinium [Gd] and agarose phantoms), bovine cartilage, and human subjects were constructed on a 3.0T clinical whole-body MR scanner. The T(1) values of model systems measured using the 2D inversion-recovery fast-spin-echo (IR-FSE) sequence were considered as a reference method to validate the rapid 3D method for comparison. The root mean square coefficient of variation percentage (RMS-CV%) of the median T(1) of agarose phantom across different acquisition methods was approximately 6.2%. The RMS-CV% of the median T(1) of bovine cartilage across different acquisition methods was approximately 4.1%. The RMS-CV% of median T(1) of the cartilages among the subjects was between approximately 7.3% to 11.1%. In our study, rapid 3D-T(1) mapping with VFA and parallel imaging with different acceleration factors (AFs) (AF = 1, 2, 3, and 4) seems to have no obvious influence on the T(1) mapping (before and after contrast agent administration). The preliminary results demonstrate that it is possible to quantify 3D-T(1) mapping of the whole knee joint (with 0.7 mm(3) isotropic resolution) under approximately five minutes with excellent in vivo reproducibility at 3.0T.

Actual flip‐angle imaging in the pulsed steady state: A method for rapid three‐dimensional mapping of the transmitted radiofrequency field

A new method has been developed for fast image-based measurements of the transmitted radiofrequency (RF) field. The method employs an actual flip-angle imaging (AFI) pulse sequence that consists of two identical RF pulses followed by two delays of different duration (TR(1) < TR(2)). After each pulse, a gradient-echo (GRE) signal is acquired. It has been shown theoretically and experimentally that if delays TR(1) and TR(2) are sufficiently short and the transverse magnetization is completely spoiled, the ratio r = S(2)/S(1) of signal intensities S(1) and S(2), acquired at the beginning of the time intervals TR(1) and TR(2), depends on the flip angle (FA) of applied pulses as r = (1 + n * cos(FA))/(n + cos(FA)), where n = TR(2)/TR(1). The method allows fast 3D implementation and provides accurate B(1) measurements that are highly insensitive to T(1). The unique feature of the AFI method is that it uses a pulsed steady-state signal acquisition. This overcomes the limitation of previous methods that required long relaxation delays between sequence repetitions. The method has been shown to be useful for time-efficient whole-body B(1) mapping and correction of T(1) maps obtained using a variable FA technique in the presence of nonuniform RF excitation.

Simultaneous spin‐echo refocusing

A new approach to spin-echo imaging is presented in which the 180 degrees RF pulse refocuses two or more spin-echoes at different positions in the readout period. When simultaneous echo refocusing (SER) is implemented using multiple 180 degrees pulses, an undesirable mixing of stimulated echoes and primary echoes from different slices can occur. A novel periodic gradient spoiler scheme eliminates this potential source of artifacts without spoiling the correctly timed stimulated echoes, which, similar to RARE (TSE) sequences, add coherently to the primary echoes. Comparisons show equivalent artifact elimination using phase cycling, periodic spoiling, and a previously developed spoiling scheme for non-Carr-Purcell-Meiboom-Gill sequences. A comparison of head images at 1.5 T acquired with SER-TSE and conventional TSE T1-weighted sequences show no degradation in image quality nor SNR. T2-weighted imaging is not achievable with the current implementation, but possible solutions are proposed. The proposed technique might prove especially beneficial at higher field strengths, where the reduced number of refocusing pulses for multislice SER-TSE decreases RF power deposition. SER spin-echo imaging offers an approach that is very different from low flip angle imaging to mitigate RF heating limitations in high-field clinical imaging.

High flip angle imaging of metallic stents: implications for MR angiography and intraluminal signal interpretation.

Intraluminal stent signal characterization by MRI is generally hampered by signal loss from the metallic stent material. This signal loss is related to magnetic susceptibility and RF shielding. Even when stent materials with low magnetic susceptibility are used, RF shielding can still be problematic. In this article we have shown that high flip angle imaging enables morphology assessment and tissue characterization in stents made of stainless steel 316L, NiTinol, and ABI-alloy.

Myocardial tagging with SSFP.

This work presents the first implementation of myocardial tagging with refocused steady-state free precession (SSFP) and magnetization preparation. The combination of myocardial tagging (a noninvasive method for quantitative measurement of regional and global cardiac function) with the high tissue signal-to-noise ratio (SNR) obtained with SSFP is shown to yield improvements in terms of the myocardium-tag contrast-to-noise ratio (CNR) and tag persistence when compared to the current standard fast gradient-echo (FGRE) tagging protocol. Myocardium-tag CNR and tag persistence were studied using numerical simulations as well as phantom and human experiments. Both quantities were found to decrease with increasing imaging flip angle (alpha) due to an increased tag decay rate and a decrease in myocardial steady-state signal. However, higher alpha yielded better blood-myocardium contrast, indicating that optimal alpha is dependent on the application: higher alpha for better blood-myocardium boundary visualization, and lower alpha for better tag persistence. SSFP tagging provided the same myocardium-tag CNR as FGRE tagging when acquired at four times the bandwidth and better tag- and blood-myocardium CNRs than FGRE tagging when acquired at equal or twice the receiver bandwidth (RBW). The increased acquisition efficiency of SSFP allowed decreases in breath-hold duration, or increases in temporal resolution, as compared to FGRE.

Functional, Dynamic, and Anatomic MR Urography: Feasibility and Preliminary Findings

Rationale and Objectives The authors assessed the feasibility of using magnetic resonance (MR) urography to acquire functional, dynamic, and anatomic information in human subjects with normal and hydronephrotic kidneys. Materials and Methods In subjects known to have or suspected of having hydronephrosis, split renal filtration fractions were measured with a customized magnetization-prepared, inversion-prepared gradient-recalled echo sequence to determine the T1 of flowing blood in the inferior vena cava and aorta before and after contrast medium administration and in the renal veins and arteries after contrast medium administration. Multiple timed sets of coronal fast spoiled gradient-echo 70° flip-angle images were acquired before and after contrast medium administration to derive MR renograms from changes in the signal intensity of the cortex and medulla. Precontrast T2-weighted images were obtained with a three-dimensional fast spoiled gradient-echo maximum intensity projection pulse sequence, and postcontrast T1 maximum intensity projection images were also obtained to depict the renal anatomy. Results Split filtration fraction differentiated normal from hydronephrotic kidneys. MR renograms depicted vascular, tubular, and ductal phases and differentiated between normal and hydronephrotic kidneys ( P < .05, n = 20). Contrast medium dose correlated with the peak of the cortical signal intensity curves on the renogram ( r = 0.7, P < .0005; n = 20). The sensitivities for the visual determination of hydronephrosis and unilateral delayed excretion of contrast material were both 100%, and the specificities were 64% and 85%, respectively. Conclusion The preliminary findings show promise for the use of MR urography in the comprehensive assessment of renal function, dynamics, and anatomy.

A low flip angle spin-echo technique for producing rapid diffusion weighted MR images

A method is described for producing rapid diffusion-weighted images using a modified low flip angle imaging technique. Utility of the method is demonstrated by the quantification of diffusion coefficients in a rat model of focal ischemia. The method may be readily applied to animal research studies using NMR research systems with modest gradient capabilities.

Variable flip angle imaging and fat suppression in combined gradient and spin-echo (GREASE) techniques

Conventional “proton density” and “ T 2-weighted” spin-echo images are susceptible to motion induced artifact, which is exacerbated by lipid signals. Gradient moment nulling can reduce motion artifact but lengthens the minimum TE, degrading the “proton density” contrast. We designed a pulse sequence capable of optimizing proton density and T 2-weighted contrast while suppressing lipid signals and motion induced artifacts. Proton density weighting was obtained by rapid readout gradient reversal immediately after the excitation RF pulse, within a conventional spin-echo sequence. By analyzing the behavior of the macroscopic magnetization and optimizing excitation flip angle, we suppressed T 1 contribution to the image, thereby enhancing proton density and T 2-weighted contrast with a two- to four-fold reduction of repetition time. This permitted an increased number of averages to be used, reducing motion induced artifacts. Fat suppression in the presence of motion was investigated in two groups of 8 volunteers each by (i) modified Dixon technique, (ii) selective excitation, and (iii) hybrid of both. Elimination of fat signal by the first technique was relatively uniform across the field of view, but it did not fully suppress the ghosts originating from fat motion. Selective excitation, while sensitive to the main field inhomogeneity, largely eliminated the ghosts (0.21 ± 0.05 vs. 0.29 ± 0.06, p < 0.01). The hybrid of both techniques combined with bandwidth optimization, however, showed the best results (0.17 ± 0.04, p < 0.001). Variable flip-angle imaging allows optimization of image contrast which, along with averaging and effective fat suppression, significantly improves gradient- and spin-echo imaging, particularly in the presence of motion.

Fast magnetic resonance imaging of intracerebral metastatic disease.

The appearance of intracerebral metastases imaged with a multi-slice fast imaging sequence (FLASH) is described. Images were obtained using a Siemens 2.0 T Magnetom operating at 1.5 T. Results of the FLASH imaging sequence at different values of repetition time, different values of "flip angle" and different numbers of acquisitions are described both qualitatively and quantitatively. At low flip angles, both tumour and oedema appear brighter than surrounding white matter. Increasing the flip angle tends to make oedema brighter than both tumour and white matter and increasing the flip angle still further makes tumour and oedema darker than white matter. A major limitation of this technique is that the low flip angle images, in particular, suffer from low signal-to-noise ratios. High flip angle FLASH images have higher signal-to-noise ratios but show similar contrast behaviour to T1-weighted spin-echo images and are likely to be no better as a screening sequence for intracerebral metastases. All the FLASH sequences showed a sensitivity to changes in magnetic susceptibility. This made small intratumoral haemorrhages and basal ganglia "calcification" easier to detect than on spin-echo images but also caused susceptibility artefacts in images around the base of the skull.

Characteristics of partial flip angle and gradient reversal MR imaging.

Characteristics of partial flip angle and gradient reversal MR imaging.

Retrospective analysis of limited flip angle images

Retrospective analysis of limited flip angle images