Iterative optimization method for design of quantitative magnetization transfer imaging experiments

Rapid whole-brain quantitative MT imaging

PURPOSE

To provide a robust whole-brain quantitative magnetization transfer (MT) imaging method that is not limited by long acquisition times.

METHODS

Two variants of a spiral 2D interleaved multi-slice spoiled gradient echo (SPGR) sequence are used for rapid quantitative MT imaging of the brain at 3 T. A dual flip angle, steady-state prepared, double-contrast method is used for combined B₁ and-T₁ mapping in combination with a single-contrast MT-prepared acquisition over a range of different saturation flip angles (50 deg to 850 deg) and offset frequencies (1 kHz and 10 kHz). Five sets (containing minimum 6 to maximum 18 scans) with different MT-weightings were acquired. In addition, main magnetic field inhomogeneities (ΔB₀) were measured from two Cartesian low-resolution 2D SPGR scans with different echo times. Quantitative MT model parameters were derived from all sets using a two-pool continuous-wave model analysis, yielding the pool-size ratio, F, their exchange rate, k_f, and their transverse relaxation time, T_2r.

RESULTS

Whole-brain quantitative MT imaging was feasible for all sets with total acquisition times ranging from 7:15 min down to 3:15 min. For accurate modeling, B₁-correction was essential for all investigated sets, whereas ΔB₀-correction showed limited bias for the observed maximum off-resonances at 3 T.

CONCLUSION

The combination of rapid B₁-T₁ mapping and MT-weighted imaging using a 2D multi-slice spiral SPGR research sequence offers excellent prospects for rapid whole-brain quantitative MT imaging in the clinical setting.

Rapid whole-brain myelin imaging with selective inversion recovery and compressed SENSE

PURPOSE

Quantitative magnetization transfer (QMT) using selective inversion recovery (SIR) can quantify the macromolecular-to-free proton pool size ratio (PSR), which has been shown to relate closely with myelin content. Currently clinical applications of SIR have been hampered by long scan times. In this work, the acceleration of SIR-QMT using CS-SENSE (compressed sensing SENSE) was systematically studied.

THEORY AND METHODS

Phantoms of varied concentrations of bovine serum albumin and human scans were first conducted to evaluate the SNR, precision of SIR-QMT parameters, and scan time. Based on these results, an optimized CS-SENSE factor of 8 was determined and the test-retest repeatability was further investigated.

RESULTS

A whole-brain SIR imaging of 6 min can be achieved. Bland-Altman analyses indicated excellent agreement between the test and retest sessions with a difference in mean PSR of 0.06% (and a difference in mean R_1f of -0.001 s^-1 ). In addition, the assessment of the intraclass correlation coefficient (ICC) revealed high reliability in nearly all the white matter and gray matter regions. In white matter regions, the ICC was 0.93 (95% confidence interval [CI]: 0.88-0.96, p < 0.001) for PSR, and 0.90 (95% CI: 0.83-0.94, p < 0.001) for R_1f . In gray matter, ICC was 0.84 (95% CI: 0.66-0.93, p < 0.001) in PSR, and 0.98 (95% CI: 0.95-0.99, p < 0.001) for R_1f . The method also showed excellent capability to detect focal lesions in multiple sclerosis.

CONCLUSION

Rapid, reliable, and sensitive whole-brain SIR imaging can be achieved using CS-SENSE, which is expected to significantly promote widespread clinical translation.

Macromolecular Proton Fraction as a Myelin Biomarker: Principles, Validation, and Applications

Macromolecular proton fraction (MPF) is a quantitative MRI parameter describing the magnetization transfer (MT) effect and defined as a relative amount of protons bound to biological macromolecules with restricted molecular motion, which participate in magnetic cross-relaxation with water protons. MPF attracted significant interest during past decade as a biomarker of myelin. The purpose of this mini review is to provide a brief but comprehensive summary of MPF mapping methods, histological validation studies, and MPF applications in neuroscience. Technically, MPF maps can be obtained using a variety of quantitative MT methods. Some of them enable clinically reasonable scan time and resolution. Recent studies demonstrated the feasibility of MPF mapping using standard clinical MRI pulse sequences, thus substantially enhancing the method availability. A number of studies in animal models demonstrated strong correlations between MPF and histological markers of myelin with a minor influence of potential confounders. Histological studies validated the capability of MPF to monitor both demyelination and re-myelination. Clinical applications of MPF have been mainly focused on multiple sclerosis where this method provided new insights into both white and gray matter pathology. Besides, several studies used MPF to investigate myelin role in other neurological and psychiatric conditions. Another promising area of MPF applications is the brain development studies. MPF demonstrated the capabilities to quantitatively characterize the earliest stage of myelination during prenatal brain maturation and protracted myelin development in adolescence. In summary, MPF mapping provides a technically mature and comprehensively validated myelin imaging technology for various preclinical and clinical neuroscience applications.

Low duty-cycle pulsed irradiation reduces magnetization transfer and increases the inhomogeneous magnetization transfer effect

Intense off-resonant RF irradiation can lead to saturation of the macromolecular pool magnetization and enhance bound pool dipolar order responsible for the inhomogeneous magnetization transfer (ihMT) effect, but the intensity of RF power in human imaging studies is limited by safety constraints on RF heating. High RF intensities can still be achieved if applied in short pulses with low duty-cycle. Here we investigate the benefits of low duty-cycle irradiation for MT and ihMT studies with both theoretical and experimental methods. Solutions for pulsed irradiation of a two-pool model including dipolar order effects were implemented. Experiments were conducted at 3 T in the brain and through the calf of healthy human subjects. 2D echo planar images were acquired following a preparation of RF irradiation with a 2 s train of 5 ms pulses repeated from between 10 to 100 ms for duty-cycles (DCs) of 50% to 5%, and at varying offset frequencies, and time averaged RF powers. MT and ihMT data were measured in regions of interest within gray matter, white matter and muscle, and fit to the model. RF irradiation effects on signal intensity were reduced at 5% relative to 50% DCs. This reduced RF effect was much larger for single than dual frequency irradiation. 5% DC irradiation reduced single and dual frequency MT ratios but increased ihMT ratios up to 3 fold in brain tissues. Muscle ihMT increased by an even larger factor, depending on the frequency and applied power. The model predicted these changes with duty-cycle. The model fit the data well and constrained model parameters. Low duty-cycle pulsed irradiation reduces MT effects and markedly increases dipolar order effects. This approach is an attractive method to enhance ihMT signal-to-noise ratio and demonstrates a measurable ihMT effect in muscle tissue at 3 T under acceptable specific absorption rates. The effects of duty-cycle changes demonstrated in a separate MT/ihMT preparation provide a route for new applications in magnetization-prepared MRI sequences.

Sensitivity regularization of the Cramér-Rao lower bound to minimize B1 nonuniformity effects in quantitative magnetization transfer imaging

PURPOSE

To develop and validate a regularization approach of optimizing B₁ insensitivity of the quantitative magnetization transfer (qMT) pool-size ratio (F).

METHODS

An expression describing the impact of B₁ inaccuracies on qMT fitting parameters was derived using a sensitivity analysis. To simultaneously optimize for robustness against noise and B₁ inaccuracies, the optimization condition was defined as the Cramér-Rao lower bound (CRLB) regularized by the B₁ -sensitivity expression for the parameter of interest (F). The qMT protocols were iteratively optimized from an initial search space, with and without B₁ regularization. Three 10-point qMT protocols (Uniform, CRLB, CRLB+B₁ regularization) were compared using Monte Carlo simulations for a wide range of conditions (e.g., SNR, B₁ inaccuracies, tissues).

RESULTS

The B₁ -regularized CRLB optimization protocol resulted in the best robustness of F against B₁ errors, for a wide range of SNR and for both white matter and gray matter tissues. For SNR = 100, this protocol resulted in errors of less than 1% in mean F values for B₁ errors ranging between -10 and 20%, the range of B₁ values typically observed in vivo in the human head at field strengths of 3 T and less. Both CRLB-optimized protocols resulted in the lowest σ_F values for all SNRs and did not increase in the presence of B₁ inaccuracies.

CONCLUSION

This work demonstrates a regularized optimization approach for improving the robustness of auxiliary measurements (e.g., B₁ ) sensitivity of qMT parameters, particularly the pool-size ratio (F). Predicting substantially less B₁ sensitivity using protocols optimized with this method, B₁ mapping could even be omitted for qMT studies primarily interested in F.

Optimization of selective inversion recovery magnetization transfer imaging for macromolecular content mapping in the human brain

PURPOSE

To optimize a selective inversion recovery (SIR) sequence for macromolecular content mapping in the human brain at 3.0T.

THEORY AND METHODS

SIR is a quantitative method for measuring magnetization transfer (qMT) that uses a low-power, on-resonance inversion pulse. This results in a biexponential recovery of free water signal that can be sampled at various inversion/predelay times (t_I/ t_D ) to estimate a subset of qMT parameters, including the macromolecular-to-free pool-size-ratio (PSR), the R₁ of free water (R_1f ), and the rate of MT exchange (k_mf ). The adoption of SIR has been limited by long acquisition times (≈4 min/slice). Here, we use Cramér-Rao lower bound theory and data reduction strategies to select optimal t_I /t_D combinations to reduce imaging times. The schemes were experimentally validated in phantoms, and tested in healthy volunteers (N = 4) and a multiple sclerosis patient.

RESULTS

Two optimal sampling schemes were determined: (i) a 5-point scheme (k_mf estimated) and (ii) a 4-point scheme (k_mf assumed). In phantoms, the 5/4-point schemes yielded parameter estimates with similar SNRs as our previous 16-point scheme, but with 4.1/6.1-fold shorter scan times. Pair-wise comparisons between schemes did not detect significant differences for any scheme/parameter. In humans, parameter values were consistent with published values, and similar levels of precision were obtained from all schemes. Furthermore, fixing k_mf reduced the sensitivity of PSR to partial-volume averaging, yielding more consistent estimates throughout the brain.

CONCLUSIONS

qMT parameters can be robustly estimated in ≤1 min/slice (without independent measures of ΔB₀ , B1+, and T₁ ) when optimized t_I -t_D combinations are selected.

Modelling and interpretation of magnetization transfer imaging in the brain

Magnetization transfer contrast has yielded insight into brain tissue microstructure changes across the lifespan and in a range of disorders. This progress has been aided by the development of quantitative magnetization transfer imaging techniques able to extract intrinsic properties of the tissue that are independent of the specifics of the data acquisition. While the tissue properties extracted by these techniques do not map directly onto specific cellular structures or pathological processes, a growing body of work from animal models and histopathological correlations aids the in vivo interpretation of magnetization transfer properties of tissue. This review examines the biophysical models that have been developed to describe magnetization transfer contrast in tissue as well as the experimental evidence for the biological interpretation of magnetization transfer data in health and disease.

An optimized framework for quantitative magnetization transfer imaging of the cervical spinal cord in vivo

PURPOSE

To develop a framework to fully characterize quantitative magnetization transfer indices in the human cervical cord in vivo within a clinically feasible time.

METHODS

A dedicated spinal cord imaging protocol for quantitative magnetization transfer was developed using a reduced field-of-view approach with echo planar imaging (EPI) readout. Sequence parameters were optimized based in the Cramer-Rao-lower bound. Quantitative model parameters (i.e., bound pool fraction, free and bound pool transverse relaxation times [ T2F, T2B], and forward exchange rate [k_FB ]) were estimated implementing a numerical model capable of dealing with the novelties of the sequence adopted. The framework was tested on five healthy subjects.

RESULTS

Cramer-Rao-lower bound minimization produces optimal sampling schemes without requiring the establishment of a steady-state MT effect. The proposed framework allows quantitative voxel-wise estimation of model parameters at the resolution typically used for spinal cord imaging (i.e. 0.75 × 0.75 × 5 mm³ ), with a protocol duration of ∼35 min. Quantitative magnetization transfer parametric maps agree with literature values. Whole-cord mean values are: bound pool fraction = 0.11(±0.01), T2F = 46.5(±1.6) ms, T2B = 11.0(±0.2) µs, and k_FB  = 1.95(±0.06) Hz. Protocol optimization has a beneficial effect on reproducibility, especially for T2B and k_FB .

CONCLUSION

The framework developed enables robust characterization of spinal cord microstructure in vivo using qMT. Magn Reson Med 79:2576-2588, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Brain tissue compartment density estimated using diffusion-weighted MRI yields tissue parameters consistent with histology

We examined whether quantitative density measures of cerebral tissue consistent with histology can be obtained from diffusion magnetic resonance imaging (MRI). By incorporating prior knowledge of myelin and cell membrane densities, absolute tissue density values were estimated from relative intracellular and intraneurite density values obtained from diffusion MRI. The NODDI (neurite orientation distribution and density imaging) technique, which can be applied clinically, was used. Myelin density estimates were compared with the results of electron and light microscopy in ex vivo mouse brain and with published density estimates in a healthy human brain. In ex vivo mouse brain, estimated myelin densities in different subregions of the mouse corpus callosum were almost identical to values obtained from electron microscopy (diffusion MRI: 42 ± 6%, 36 ± 4%, and 43 ± 5%; electron microscopy: 41 ± 10%, 36 ± 8%, and 44 ± 12% in genu, body and splenium, respectively). In the human brain, good agreement was observed between estimated fiber density measurements and previously reported values based on electron microscopy. Estimated density values were unaffected by crossing fibers.

Quantitative magnetization transfer imaging of human brain at 7 T

Quantitative magnetization transfer (qMT) imaging yields indices describing the interactions between free water protons and immobile macromolecular protons. These indices include the macromolecular to free pool size ratio (PSR), which has been shown to be correlated with myelin content in white matter. Because of the long scan times required for whole-brain imaging (≈20-30 min), qMT studies of the human brain have not found widespread application. Herein, we investigated whether the increased signal-to-noise ratio available at 7.0 T could be used to reduce qMT scan times. More specifically, we developed a selective inversion recovery (SIR) qMT imaging protocol with a i) novel transmit radiofrequency (B(1)(+)) and static field (B(0)) insensitive inversion pulse, ii) turbo field-echo readout, and iii) reduced TR. In vivo qMT data were obtained in the brains of healthy volunteers at 7.0 T using the resulting protocol (scan time≈40 s/slice, resolution=2 × 2 × 3 mm(3)). Reliability was also assessed in repeated acquisitions. The results of this study demonstrate that SIR qMT imaging can be reliably performed within the radiofrequency power restrictions present at 7.0 T, even in the presence of large B(1)(+) and B(0) inhomogeneities. Consistent with qMT studies at lower field strengths, the observed PSR values were higher in white matter (mean±SD=17.6 ± 1.3%) relative to gray matter (10.3 ± 1.6%) at 7.0 T. In addition, regional variations in PSR were observed in white matter. Together, these results suggest that qMT measurements are feasible at 7.0 T and may eventually allow for the high-resolution assessment of changes in composition throughout the normal and diseased human brain in vivo.