2825

In-vivo Magnetization Transfer neuroimaging at 46 mT

Beatrice Lena¹, Thomas O'Reilly¹, and Andrew Webb¹
¹C.J. Gorter MRI Center, Radiology Department, Leids Universitair Medisch Centrum, Leiden, Netherlands

Synopsis

Keywords: Low-Field MRI, Magnetization transfer, Low Field MRI

Motivation: Here, we exploited the Low SAR in low-field MRI to study magnetization transfer (MT) , and potential to enhance brain tissue contrast, in particular between white and gray matter (challenging at low-field).

Goal(s): The study aimed to identify suitable MT settings and explore MT contrast in-vivo at 46 mT.

Approach: MT settings were optimized to minimize direct saturation and maximize MT contrast. With this setup, we scanned five volunteers at 46 mT and acquired additional 3T scans for two participants

Results: Optimal MT settings were defined in the phantom and in-vivo scans demonstrated the expected MT effects, and improved tissue contrast at 46 mT.

Impact: Magnetization transfer could be achieved off-resonance in-vivo at low-field MRI. It improves brain tissue contrast, in particular between white matter and gray matter, and has minimal direct saturation.

Introduction

The advancements in both software and hardware, along with improved image reconstruction and denoising techniques, have brought renewed interest in the application of low-field MRI in Point-of-care (POC) settings, thanks also to its reduced financial costs, portability, and accessibility[1]. Nevertheless, there are still potential areas not yet exploited. One significant advantage of low-field MRI is its low specific absorption rate (SAR)[2]. This allows the use of long off-resonance RF pulses for strong magnetization transfer (MT) saturation to achieve enhanced contrast in tissue and lesions[3]. For accurate off-resonance measurements, relatively high homogeneity is required. We previously showed that a Halbach-magnet array POC scanner with optimized ring diameters and magnet positions/orientations was sufficiently homogenous for this application[4]. In this study we characterized the settings for the MT experiment in a phantom and we explored the MT contrast in-vivo at 46 mT. Moreover, we qualitatively compared the low-field images with 3T scans.

Methods

Theory
The RF energy of the MT pulses is applied off-resonance, where it saturates some of the longitudinal magnetization of the macromolecular pool, while having hardly any effect on the liquid pool. Ideally, the MT pulse should not directly saturate any of the protons in the liquid pool, as that would reduce the SNR. The direct saturation Mdir and the saturation due to the true MT effect MMT together define the total saturation Msat=Mdirect+MMT. The degree of direct saturation of the liquid pool can be estimated assuming a Gaussian line shape: $$M_{direct}/M_0=exp(-(\gamma B_1\Delta\nu)^2 T_1/T_2 /2)$$
for a block pulse, with γ gyromagnetic ratio, B₁ and $$$\Delta\nu$$$ the amplitude and frequency offset of the MT pulse, and T_1,2 the relaxation times of the tissue of interest[5, 6].
Acquisition
Images were obtained using a 46 mT Halbach-magnet based MRI system using a Magritek Kea2 spectrometer[7]. A single transmit and receive solenoid and head coil were used for the phantom and in-vivo study, respectively. MT was implemented by applying off-resonance RF energy continuously, followed by a spoiler gradient.
For the phantom study, M_sat and M₀ were measured acquiring 9 spectra after the MT pulse with 1 kHz increment in the frequency offset in water, gel, cross-linked BSA[8] at 14% and 24% w/v. The first 2 tubes were used to check the direct saturation of the liquid pool, as they should show no M_MT, and the BSA tubes should mimic the ratio T₁/T₂ and MTR expected in white and grey matter at low MR field. For the in-vivo study, 5 volunteers were scanned with the Halbach-magnet MRI 46 mT. The protocol involved two scans, with and without MT pulse, the specifics of which were chosen using the phantom study (hard pulse with B₁=12 $$$\mu T$$$, $$$\Delta\nu=2$$$ kHz, MT pulse length =100 ms). After this premodule, a 3D turbo spin-echo (TSE) readout was used, with the following parameters: TR/TE: 1000/19 ms, echo train length: 7, 1.5x1.5x10 mm³ resolution, and acquisition bandwidth: 25 kHz. For comparison, two volunteers had an additional MT scan with a clinical 3T scanner (Philips Achieva), with MT pulse specifics: sinc-gauss pulse with 12 $$$\mu T$$$, $$$\Delta\nu=2$$$ kHz, MT pulse length.

Results

With a maximum of 10% of direct saturation, the optimal settings for the MT experiment in the phantom using a 15 $$$\mu T$$$ pulse was: $$$\Delta\nu=2$$$ kHz and MT pulse length=100 ms (Figure 1). Applied off-resonance, though, the effective B₁ amplitude of the MT-pulse changes (Figure 2) due to the narrow bandwidth of the RF coil, resulting in this case in 12 $$$\mu T$$$, instead. In-vivo (Figure 3), signal reduction after the MT pulse was evident: the computed MT ratio map resulted in $$$28 \pm 2.5%$$$ in WM, $$$21 \pm 3.2%$$$ in GM and $$$6 \pm 3%$$$% in CSF (in line with the phantom experiment). Comparing 46 mT and 3T scans (Figure 4) in two volunteers, finer structures as deep grey matter appear more evidently on the MT scan in both cases. In general, MT scans enhance WM/GM contrast in all volunteers, also when artefacts were present (Figure 5).

Discussions and Conclusions

In this study, we demonstrated how an appropriately selected MT prepulse before a standard sequence can enhance WM/GM tissue contrast. This is relevant especially at low-field MRI, where obtaining high WM/GM contrast can be challenging. It should be noted that MTR maps may be less useful at this field strength due to lower SNR and low-field MRI and more challenging computations (image registration is difficult with such a coarse slice thickness).

Acknowledgements

This work was partly funded by the Dutch Science Foundation Open Technology 18981.

References

1. Hori M, Hagiwara A, Goto M, et al (2021) Low-Field Magnetic Resonance Imaging Its History and Renaissance. https://doi.org/10.1097/RLI.0000000000000810

2. Marques JP, Simonis FFJ, Webb AG (2019) Low-field MRI: An MR physics perspective. Journal of Magnetic Resonance Imaging 49

3. de Boer RW (1995) Magnetization transfer contrast Part 2 : Clinical applications. Medica Mundi (Philips Healthcare)

4. Lena B, Najac C, Václavu L, et al (2023) T1ρ and magnetization transfer experiments on a point-of-care 46 mT MRI system. In: Proceedings of ISMRM ISMRT 32nd Annual Meeting

5. Morrison C, Stanisz G, Henkelman RM (1995) Modeling Magnetization Transfer for Biological-like Systems Using a Semi-solid Pool with a Super-Lorentzian Lineshape and Dipolar Reservoir. J Magn Reson B 108:. https://doi.org/10.1006/jmrb.1995.1111

6. Hajnal J V., Baudouin CJ, Oatridge A, et al (1992) Design and implementation of magnetization transfer pulse sequences for clinical use. J Comput Assist Tomogr 16:. https://doi.org/10.1097/00004728-199201000-00003

7. O’Reilly T, Webb A (2021) The design of a low-weight homogenous Halbach helmet for imaging the adult brain. In: Proceedings of 31st annual meeting of ISMRM & ISMRT

8. Koenig SH, Brown RD, Ugolini R (1993) Magnetization transfer in cross‐linked bovine serum albumin solutions at 200 MHz: A model for tissue. Magn Reson Med 29:. https://doi.org/10.1002/mrm.1910290306

Figures

Figure 1. MT spectra measured in the tubes with an MT pulse of amplitude 15 , different offset frequencies and duration (round markers). The dotted line indicates the threshold for the maximum degree of direct saturation allowed (10%). The bold line shows the degree of direct saturation estimated from Eq.1 for the white matter (T1=275 ms, T2=102 ms) with the specifics of this MT pulse. The best combination of duration and frequency offset that maximize the MT contrast and minimize direct saturation is highlighted in yellow and are used for the in-vivo experiments.

Figure 2. Because the MT pulse is applied off-resonance, the power optimization should be repeated applying the proper frequency offset. This leads to a decrease in the effective transmitted RF power as the frequency offset increases due to the limited RF bandwidth of the RF coil.

Figure 3. A. Scan images at 46 mT in one volunteer. In the MT scan, a significant reduction of signal amplitude can be observed, especially in the white matter. This decrease was expected, considering the white matter is mostly composed of axons coated with myelin, a mixture of proteins and lipids. Grey matter, consisting mainly of neuronal cell bodies, showed less signal reduction after the MT pulse. WM/GM contrast is significantly improved in the MT scan images. B. Example of a MT ratio map that can be calculated from the two scans in A.

Figure 4. Scan images at 46 mT and 3T in two volunteers. Beside the higher resolution and image quality of 3T, at both field strength the MT scan allows for better identification of finer image structures (white arrows).

Figure 5. Scan images at 46 mT in other two volunteers. Beside two artefacts, due to wrong positioning of the subjects in the coil and suboptimal power optimization (red arrow), WM/GM contrast is visibly improved in the scans with MT pulse.

Proc. Intl. Soc. Mag. Reson. Med. 32 (2024)

2825

DOI: https://doi.org/10.58530/2024/2825