Myocardial T measurements in iron‐overloaded thalassemia: An in vivo study to investigate optimal methods of quantification

Reproducible and accurate myocardial T measurements are required for the quantification of iron in heart tissue in transfused thalassemia. The aim of this study was to determine the best method to measure the myocardial T from multi‐gradient‐echo data acquired both with and without black‐blood preparation. Sixteen thalassemia patients from six centers were scanned twice locally, within 1 week, using an optimized bright‐blood T sequence and then subsequently scanned at the standardization center in London within 4 weeks, using a T sequence both with and without black‐blood preparation. Different curve‐fitting models (monoexponential, truncation, and offset) were applied to the data and the results were compared by means of reproducibility. T measurements obtained using the bright‐ and black‐blood techniques. The black‐blood data were well fitted by the monoexponential model, which suggests that a more accurate measure of T can be obtained by removing the main source of errors in the bright‐blood data. For bright‐blood data, the offset model appeared to underestimate T values substantially and was less reproducible. The truncation model gave rise to more reproducible T measurements, which were also closer to the values obtained from the black‐blood data. Magn Reson Med 60:1082–1089, 2008. © 2008 Wiley‐Liss, Inc.     

Performance of sampling density‐weighted and postfiltered density‐adapted projection reconstruction in sodium magnetic resonance imaging

Sampling density‐weighted apodization projection reconstruction sequences are evaluated for three‐dimensional radial imaging. The readout gradients of the sampling density‐weighted apodization sequence are designed such that the locally averaged sampling density matches a Hamming filter function. This technique is compared with density‐adapted projection reconstruction with nonfiltered and postfiltered image reconstruction. Sampling density‐weighted apodization theoretically allows for a 1.28‐fold higher signal‐to‐noise ratio compared with postfiltered density‐adapted projection reconstruction sequences, if T decay is negligible compared with the readout duration T_RO. Simulations of the point‐spread functions are performed for monoexponential and biexponential decay to investigate the effects of T decay on the performance of the different sequences. Postfiltered density‐adapted projection reconstruction performs superior to sampling density‐weighted apodization for large T_RO/T ratios [>1.36 (monoexponential decay); >0.35 (biexponential decay with T/T = 10)], if signal‐to‐noise ratio of point‐like objects is considered. In conclusion, it depends on the readout parameters, the T relaxation times, and the dimensions of the subject which of both sequences is most suitable. Magn Reson Med, 2013. © 2012 Wiley Periodicals, Inc.     

Feasibility of T mapping for the evaluation of hip joint cartilage at 1.5T using a three‐dimensional (3D), gradient‐echo (GRE) sequence: A prospective study

This study defines the feasibility of utilizing three‐dimensional (3D) gradient‐echo (GRE) MRI at 1.5T for T mapping to assess hip joint cartilage degenerative changes using standard morphological MR grading while comparing it to delayed gadolinium‐enhanced MRI of cartilage (dGEMRIC). MRI was obtained from 10 asymptomatic young adult volunteers and 33 patients with symptomatic femoroacetabular impingement (FAI). The protocol included T mapping without gadolinium‐enhancement utilizing a 3D‐GRE sequence with six echoes, and after gadolinium injection, routine hip sequences, and a dual‐flip‐angle 3D‐GRE sequence for dGEMRIC T₁ mapping. Cartilage was classified as normal, with mild changes, or with severe degenerative changes based on morphological MRI. T₁ and T findings were subsequently correlated. There were significant differences between volunteers and patients in normally‐rated cartilage only for T₁ values. Both T₁ and T values decreased significantly with the various grades of cartilage damage. There was a statistically significant correlation between standard MRI and T (T₁) (P < 0.05). High intraclass correlation was noted for both T₁ and T. Correlation factor was 0.860 to 0.954 (T‐T₁ intraobserver) and 0.826 to 0.867 (T‐T₁ interobserver). It is feasible to gather further information about cartilage status within the hip joint using GRE T mapping at 1.5T. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Ultrashort T relaxometry for quantitation of highly concentrated superparamagnetic iron oxide (SPIO) nanoparticle labeled cells

A new method was developed to measure ultrashort T relaxation in tissues containing a focal area of superparamagnetic iron oxide (SPIO) nanoparticle‐labeled cells in which the T decay is too short to be accurately measured using regular gradient echo T mapping. The proposed method utilizes the relatively long T₂ relaxation of SPIO‐labeled cells and acquires a series of spin echo images with the readout echo shifted to sample the T decay curve. MRI experiments in phantoms and rats with SPIO‐labeled tumors demonstrated that it can detect ultrashort T down to 1 ms or less. The measured T values were about 10% higher than those from the ultrashort TE (UTE) technique. The shorter the TE, the less the measurements deviated from the UTE T mapping. Combined with the regular T mapping, this technique is expected to provide quantitation of highly concentrated iron‐labeled cells from direct cell transplantation. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Myocardial T measurement in iron‐overloaded thalassemia: An ex vivo study to investigate optimal methods of quantification

Myocardial T measurement has been increasingly used for iron quantification to assess the risk of cardiac complications in thalassemia patients. In this study the noise effects were evaluated along with different curve‐fitting models on an iron overloaded ex vivo heart in order to determine the optimal method of T measurement and to help understand issues affecting reproducibility and accuracy. Gradient multiecho short axis images were acquired with differing numbers of excitations to generate varying signal‐to‐noise ratio (SNR) images. A noise correction method was implemented; linear and nonlinear curve‐fitting algorithms were compared and different curve‐fitting models (monoexponential, truncation, baseline subtraction, and offset) were evaluated. This study suggests that the T decay curve in an ex vivo heart can be fitted by a monoexponential model and accurate T measurements can be obtained with proper noise correction. With MRI noise, T is generally overestimated by including late low SNR data points, but underestimated by the offset or baseline subtraction models, which are in fact equivalent. In this situation the truncation model proves to be reproducible and more accurate than the other models. The study also shows that the nonlinear algorithm is preferred in T curve fitting. Magn Reson Med 60:350–356, 2008. © 2008 Wiley‐Liss, Inc.     

Rapid single‐scan T‐mapping using exponential excitation pulses and image‐based correction for linear background gradients

A method for fast quantitative T mapping based on multiple gradient‐echo (multi‐GE) imaging with correction for static magnetic field inhomogeneities is described, using an exponential excitation pulse. Field gradient maps are obtained from the phase information and modulus data are subsequently corrected, allowing for simple monoexponential T fitting. Echoes with long echo times suffering from major signal losses due to field inhomogeneities are excluded from the analysis. The acquisition time for a matrix size of 256 × 256, 1 mm in‐plane resolution, and 2 mm slice thickness amounts to 15 s per slice. An additional correction for in‐plane field gradients further improves accuracy. Phantom experiments show that the method provides accurate T values for field gradients up to 200 μT/m; for gradients up to 300 μT/m errors do not exceed 15%. In vivo T values acquired on healthy volunteers at 3T are in excellent agreement with results from the literature. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Regional variations of T2* in healthy and pathologic achilles tendon in vivo at 7 Tesla: Preliminary results

The aim of this study was to investigate T in the Achilles tendon (AT), in vivo, using a three‐dimensional ultrashort time echo (3D‐UTE) sequence, to compare field strength differences (3 and 7 T) and to evaluate a regional variation of T in healthy and pathologic tendon. Ten volunteers with no history of pain in the AT and five patients with chronic Achilles tendinopathy were recruited. 3D‐UTE images were measured with the following echo times, at echo time = [0.07, 0.2, 0.33, 0.46, 0.59, 0.74, 1.0, 1.5, 2.0, 4.0, 6.0, and 9.0 ms]. T values in the AT were calculated by fitting the signal decay to biexponential function. Comparing volunteers between 3 and 7 T, short component T was 0.71 ± 0.17 ms and 0.34 ± 0.09 ms (P < 0.05); bulk long component T was 12.85 ± 1.87 ms and 10.28 ± 2.28 ms (P < 0.05). In patients at 7 T, bulk T was 0.53 ± 0.17 ms (P = 0.045, compared to volunteers), T was 11.49 ± 4.28 ms (P = 0.99, compared to volunteers). The results of this study suggest that the regional variability of AT can be quantified by T in in vivo conditions. Advanced quantitative imaging of the human AT using a 3D‐UTE sequence may provide additional information to standard clinical imaging. Finally, as the preliminary patient data suggest, T may be a promising marker for the diagnosis of pathological changes in the AT. Magn Reson Med, 2012. © 2012 Wiley Periodicals, Inc.     

Dynamically applied B1+ shimming solutions for non‐contrast enhanced renal angiography at 7.0 tesla

The purpose of this study was to detail a strategy for performing non‐contrast enhanced renal magnetic resonance angiography studies at 7.0 T. It is demonstrated that with proper B management, these studies can be successfully performed at ultrahigh field within local specific absorption rate constraints. An inversion prepared gradient echo acquisition, standard for non‐contrast renal magnetic resonance angiography studies, required radiofrequency pulse specific B shimming solutions to be dynamically applied to address the field dependent increases in both B₀ and B inhomogeneity as well as to accommodate limitation in available power. By using more efficient B shimming solutions for the inversion preparation and more homogeneous solutions for the excitation, high quality images of the renal arteries were obtained without venous and background signal artifacts while working within hardware and safety constraints. Finite difference time domain simulations confirmed in vivo measurements with respect to B distributions and homogeneity for the range of shimming strategies used and allowed the calculation of peak local specific absorption rate values normalized by input power and B. Increasing B homogeneity was accompanied by decreasing local specific absorption rate per Watt and increasing maximum local specific absorption rate per [B]², which must be considered, along with body size and respiratory rate, when finalizing acquisition parameters for a given individual. Magn Reson Med, 2013. © 2012 Wiley Periodicals, Inc.     

B(1) mapping with selective pulses

Knowledge of B distribution is crucial for many applications, such as quantitative MRI. A novel method has been developed to improve the accuracy of the conventionally applied double‐angle method for B mapping. It solves the remaining issues raised by the use of selective pulses for slice selection to accelerate the acquisition process. A general approach for reconstructing B maps is presented first. It takes B‐induced slice profile distortions over off‐resonance frequencies into account. It is then shown how the ratio between the prescribed flip angles can be adjusted to reach a compromise between the level of noise propagated onto B maps and the width of the range in which the field can be mapped. Lastly, several solutions are proposed for reducing the B‐dependent pollution of regions distal to the image slice which participates significantly in the inaccuracy of B mapping. These methods were experimentally tested by comparison with gold standard B maps obtained on a phantom using a non‐selective and thus much slower technique. As they are independent and lead to significant improvements, these solutions can be combined to achieve high precision and fast B mapping using spin‐echo DAM. Magn Reson Med, 2012. © 2012 Wiley Periodicals, Inc.     

Postprocessing correction for distortions in T2* decay caused by quadratic cross‐slice B0 inhomogeneity

Relaxometric measurement of the effective transverse relaxation rate R plays an important role in the quantitative evaluation of brain function, perfusion, and tissue iron content. However, accurate measurement of R is prone to macroscopic background field inhomogeneity. In clinical applications and systems, postprocessing correction techniques are more flexible in implementation than unsupported protocol or hardware modifications. The current postprocessing correction approach assumes the cross‐slice background field inhomogeneity can be approximated by a linear gradient and corrects for a sinc modulation function. The importance of the high‐order terms in background field inhomogeneity has increased with the fast development of high‐ and ultrahigh‐field scanners in recent years. In this study, we derived an analytical expression of the free induction decay signal modulation in the presence of a quadratic cross‐slice background field inhomogeneity. The proposed quadratic correction method was applied to phantom and volunteer studies and demonstrated to be superior to the classic monoexponential model, monoexponential‐plus‐constant model, and the linear sinc correction method in recovering background field inhomogeneity‐induced. R overestimations with visual inspection of R parametric maps and a statistical model selection technique. We also tabulated 7‐T T/R measurements of several human brain structures and MnCl₂ solutions with various concentrations for fellow researchers' reference. Magn Reson Med 63:1258–1268, 2010. © 2010 Wiley‐Liss, Inc.     

Bolus‐tracking MRI with a simultaneous T1‐ and T‐measurement

The aim of this study was to propose and evaluate a methodology to analyze simultaneously acquired T‐weighted dynamic susceptibility contrast (DSC) MRI and T₁‐weighted dynamic contrast enhanced (DCE) MRI data. Two generalized models of T‐relaxation are proposed to account for tracer leakage, and a two‐compartment exchange model is used to separate tracer in intra‐ and extravascular spaces. The methods are evaluated using data extracted from ROIs in three mice with subcutaneously implanted human colorectal tumors. Comparing plasma flow values obtained from DCE‐MRI and DSC‐MRI data defines a practical experimental paradigm to measure T‐relaxivities, and reveals a factor of 15 between values in tissue and blood. Comparing mean transit time values obtained from DCE‐MRI and DSC‐MRI without leakage correction, indicates a significant reduction of susceptibility weighting in DSC‐MRI during tracer leakage. A one‐parameter gradient correction model provides a good approximation for this susceptibility loss, but redundancy of the parameter limits the practical potential of this model for DSC‐MRI. Susceptibility loss is modeled more accurately with a variable T‐relaxivity, which allows to extract new parameters that cannot be derived from DSC‐MRI or DCE‐MRI alone. They reflect the cellular and vessel geometry, and thus may lead to a more complete characterization of tissue structure. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

B interferometry for the calibration of RF transmitter arrays

Multiple‐channel RF transmission holds great promise for MRI, especially for human applications at high fields. For calibration it requires mapping the effective RF magnetic fields, B, of the transmitter array. This is challenging to do accurately and fast due to the large dynamic range of B and tight SAR constraints. In the present work, this problem is revisited and solved by a novel mapping approach relying on an interference principle. The B fields of individual transmitter elements are measured indirectly by observing their interference with a SAR‐efficient baseline RF field. In this fashion even small RF fields can be observed in the B ‐sensitive large‐flip‐angle regime. Based on a set of such experiments B maps of the individual transmitter channels are obtained by solving a linear inverse problem. Confounding relaxation and off‐resonance effects are addressed by an extended signal model and nonlinear fitting. Using the novel approach, 2D mapping of an 8‐channel transmitter array was accomplished in less than a minute. For validation it is demonstrated that mapping results do not vary with T₁ or parameters of the mapping sequence. In RF shimming experiments it is shown that the measured B maps accurately reflect the linearity of RF superposition. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Characterization of paramagnetic effects of molecular oxygen on blood oxygenation level‐dependent‐modulated hyperoxic contrast studies of the human brain

In hyperoxic contrast studies modulated by the blood oxygenation level‐dependent effect, it is often assumed that hyperoxia is a purely intravascular, positive contrast agent in T‐weighted images, and the effects that are not due to blood oxygenation level‐dependent contrast are small enough to be ignored. In this study, this assumption is re‐evaluated and non‐blood oxygenation level‐dependent effects in T‐weighted hyperoxic contrast studies of the human brain were characterized. We observed significant negative signal changes in T‐weighted images in the frontal lobes; B₀ maps suggest that this effect was primarily due to increased intravoxel dephasing from increased static field inhomogeneity due to susceptibility changes from oxygen in and around the upper airway. These static field effects were shown to scale with magnetic field strength. Signal changes observed around the brain periphery and in the ventricles suggest the effect of image distortions from oxygen‐induced bulk B₀ shifts, along with a possible contribution from decreased T due to oxygen dissolved in the cerebrospinal fluid. Reducing the concentration of inhaled oxygen was shown to mitigate negative contrast of molecular oxygen due to these effects, while still maintaining sufficient blood oxygenation level‐dependent contrast to produce accurate measurements of cerebral blood volume. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.     

RF pulse optimization for Bloch‐Siegert B mapping

The Bloch–Siegert (B–S) method of B mapping has been shown to be fast and accurate, yet has high SAR and moderately long TE. These limitations can lengthen scan times and incur signal loss due to B₀ inhomogeneity, particularly at high field. The B–S method relies on applying a band‐limited off‐resonant B–S radiofrequency pulse to induce a B‐dependent frequency‐shift for resonant spins. A method for optimizing the B–S radiofrequency pulse is presented here, which maximizes B–S B measurement sensitivity for a given SAR and T₂. A 4‐ms optimized pulse is shown to have 35% less SAR compared with the conventional 6‐ms Fermi pulse while still improving B map angle‐to‐noise ratio by 22%. The optimized pulse performance is validated both in phantom and in vivo brain imaging at 7 T. Magn Reson Med, 2012. © 2011 Wiley Periodicals, Inc.     

Early detection of lung inflammation: Exploiting T1‐effects of iron oxide particles using UTE MRI

At high magnetic fields diagnostic proton MRI of the lung is problematic, because of fast T relaxation. The application of superparamagnetic contrast agents and the exploitation of the corresponding T effect is inefficient with conventional MRI methods, which limits the early detection of lung diseases. However, a simple theoretical treatment shows that in the lung, by the use of ultra‐short echo time sequences, T effects can be neglected while T₁ shortening effects can be used for signal detection. In our study, we have applied a theoretically and experimentally optimized 3D ultra‐short echo time sequence to lung phantoms and to a mouse model of lung inflammation, which was induced by systemic bacterial infection. Following the systemic application of very small superparamagnetic iron oxide nanoparticles, a significant signal increase in the lung of infected animals was detected already at 24 h postinfection, compared to control mice (17%, P < 0.001). Iron accumulation in the lung parenchyma as consequence of the host immune response was histologically confirmed. By conventional T‐ and T₂‐weighted imaging, neither structural changes nor formation of substantial edema were observed. Magn Reson Med, 2012. © 2012 Wiley Periodicals, Inc.     

Fast CPMG‐based Bloch‐Siegert B1+ mapping

A novel method for B mapping based on the Bloch‐Siegert (BS) shift was recently presented. This method applies off‐resonant pulses before signal acquisition to encode B₁ information into the signal phase. BS‐based methods possess significant advantages in measurement time and accuracy compared to magnitude‐based B methods. This study extends the idea of BS B mapping to Carr, Purcell, Meiboom, Gill (CPMG)‐based multi‐spin‐echo (BS‐CPMG‐MSE) and turbo‐spin‐echo (BS‐CPMG‐TSE) imaging. Compared to BS‐based spin echo imaging (BS‐SE), faster acquisition of the B information was possible using the BS‐CPMG‐TSE sequence. Furthermore, signal loss by T₂* effects could be minimized using these spin echo‐based techniques. These effects are critical for gradient echo‐based BS methods at high field strengths. However, multi‐spin‐echo‐based BS B₁ methods inherently possess high specific absorption rates. Thus, the relative specific absorption rate of BS‐CPMG‐TSE sequences was estimated and compared with the specific absorption rate produced by BS‐SE sequences. Magn Reson Med, 2012. © 2011 Wiley Periodicals, Inc.     

Multiecho water‐fat separation and simultaneous R estimation with multifrequency fat spectrum modeling

Multiecho chemical shift–based water‐fat separation methods are seeing increasing clinical use due to their ability to estimate and correct for field inhomogeneities. Previous chemical shift‐based water‐fat separation methods used a relatively simple signal model that assumes both water and fat have a single resonant frequency. However, it is well known that fat has several spectral peaks. This inaccuracy in the signal model results in two undesired effects. First, water and fat are incompletely separated. Second, methods designed to estimate T in the presence of fat incorrectly estimate the T decay in tissues containing fat. In this work, a more accurate multifrequency model of fat is included in the iterative decomposition of water and fat with echo asymmetry and least‐squares estimation (IDEAL) water‐fat separation and simultaneous T estimation techniques. The fat spectrum can be assumed to be constant in all subjects and measured a priori using MR spectroscopy. Alternatively, the fat spectrum can be estimated directly from the data using novel spectrum self‐calibration algorithms. The improvement in water‐fat separation and T estimation is demonstrated in a variety of in vivo applications, including knee, ankle, spine, breast, and abdominal scans. Magn Reson Med 60:1122–1134, 2008. © 2008 Wiley‐Liss, Inc.     

Comparison of amyloid plaque contrast generated by T2‐weighted,T‐weighted,and susceptibility‐weighted imaging methods in transgenic mouse models of Alzheimer's disease

One of the hallmark pathologies of Alzheimer's disease (AD) is amyloid plaque deposition. Plaques appear hypointense on T₂‐weighted and T‐weighted MR images probably due to the presence of endogenous iron, but no quantitative comparison of various imaging techniques has been reported. We estimated the T₁, T₂, T, and proton density values of cortical plaques and normal cortical tissue and analyzed the plaque contrast generated by a collection of T₂‐weighted, T‐weighted, and susceptibility‐weighted imaging (SWI) methods in ex vivo transgenic mouse specimens. The proton density and T₁ values were similar for both cortical plaques and normal cortical tissue. The T₂ and T values were similar in cortical plaques, which indicates that the iron content of cortical plaques may not be as large as previously thought. Ex vivo plaque contrast was increased compared to a previously reported spin‐echo sequence by summing multiple echoes and by performing SWI; however, gradient echo and SWI were found to be impractical for in vivo imaging due to susceptibility interface–related signal loss in the cortex. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Broadband slab selection with B mitigation at 7T via parallel spectral‐spatial excitation

Chemical shift imaging benefits from signal‐to‐noise ratio (SNR) and chemical shift dispersion increases at stronger main field such as 7 Tesla, but the associated shorter radiofrequency (RF) wavelengths encountered require B mitigation over both the spatial field of view (FOV) and a specified spectral bandwidth. The bandwidth constraint presents a challenge for previously proposed spatially tailored B mitigation methods, which are based on a type of echovolumnar trajectory referred to as “spokes” or “fast‐k_z”. Although such pulses, in conjunction with parallel excitation methodology, can efficiently mitigate large B inhomogeneities and achieve relatively short pulse durations with slice‐selective excitations, they exhibit a narrow‐band off‐resonance response and may not be suitable for applications that require B mitigation over a large spectral bandwidth. This work outlines a design method for a general parallel spectral‐spatial excitation that achieves a target‐error minimization simultaneously over a bandwidth of frequencies and a specified spatial‐domain. The technique is demonstrated for slab‐selective excitation with in‐plane B mitigation over a 600‐Hz bandwidth. The pulse design method is validated in a water phantom at 7T using an eight‐channel transmit array system. The results show significant increases in the pulse's spectral bandwidth, with no additional pulse duration penalty and only a minor tradeoff in spatial B mitigation compared to the standard spoke‐based parallel RF design. Magn Reson Med 61:493–500, 2009. © 2009 Wiley‐Liss, Inc.     

Role of very high order and degree B(0) shimming for spectroscopic imaging of the human brain at 7 tesla

With the advent of ultrahigh field systems (7T), significant improvements in spectroscopic imaging (SI) studies of the human brain have been anticipated. These gains are dependent upon the achievable B₀ homogeneity, both globally (σB, over the entire regions of interest or slice) and locally (σB, influencing the linewidth of individual SI voxels within the regions of interest). Typically the B₀ homogeneity is adjusted using shim coils with spatial distributions modeled on spherical harmonics which can be characterized by a degree (radial dependence) and order (azimuthal symmetry). However, the role of very high order and degree shimming (e.g., 3rd and 4th degree) in MRSI studies has been controversial. Measurements of σB and σB were determined from B₀ field maps of 64 × 64 resolution. In a 10 mm thick slice taken through the region of the subcortical nuclei, we find that in comparison to 1st–2nd degree shims, use of 1st–3rd and 1st–4th degree shims reduces σB by 29% and 55%, respectively. Using a SI voxel size of ～1cc with an estimate of σB from 3 × 3 × 3 B₀ map pixels in this subcortical region, the number of pixels with σB of less than 5 Hz increased from 24 to 59% with 1st–3rd and 1st–4th over 1st–2nd degree shims, respectively. Magn Reson Med, 2012. © 2011 Wiley Periodicals, Inc.