Detection of tiny oscillatory magnetic fields using low-field MRI: A combined phantom and simulation study

https://doi.org/10.1016/j.jmr.2020.106828Get rights and content

Highlights

  • Comparison of two spin-lock methods for low-field functional MRI.

  • The analytical solution of the Bloch equation under these sequences.

  • The difference of the methods was clarified experimentally and theoretically.

  • With 0.3-T MRI scanner, we investigated the detectable strength of the two methods.

Abstract

We demonstrated the feasibility of the spin-lock preparation sequence using low-field magnetic resonance (MR) imaging that prevents interference from blood-oxygenation-level-dependent effects. We focused on two spin-lock preparations: spin-lock Mz (SL-Mz) and stimulus-induced rotary saturation (SIRS) and analyzed the magnetization dynamics during the sequences using the Bloch equation. Next, we performed phantom experiments using a loop coil to investigate the MR signal change as a function of the target signal strength and phase. Furthermore, we performed curve fittings to consider the radio frequency, which agreed with the experimental results. Then, we investigated the detectable strength of the magnetic field, and the SL-Mz detected a signal strength of 2.34 nT. In conclusion, our experimental results showed good agreement with the results obtained using the Bloch equation.

Introduction

Functional magnetic resonance imaging (fMRI) is a technique for the non-invasive measurement of brain activity with high resolution. The major fMRI method is blood-oxygenation-level dependent (BOLD)[1], [2], detecting brain activity using hemodynamics. This technique is effective and widely used; however, there is a temporal limitation because of the time lag between the brain activity and hemodynamic response. To overcome this limitation, there are a number of reports on the direct measurements of brain activity, such as the neuronal current MRI [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15] and SQUID NMR/MRI[16], [17].

Among these techniques, we focused on three spin-lock techniques: the stimulus-induced rotary saturation (SIRS) proposed by Witzel et al.[18] and Halpern-Manner et al. [19], the spin-locked oscillatory excitation proposed by J. Sheng et al. [20], [21], [22], and the spin-lock technique proposed by Truong et al.[23], which we call the spin-lock Mz (hereinafter SL-Mz). These methods caused interference of the neural magnetic fields with the magnetization by MR. Therefore, these methods could be used to measure neural activities as MR signals, which helped overcome the BOLD-fMRI limitation.

Furthermore, these methods are independent of the strength of the static magnetic field, i.e., these methods are suitable for low-field MRIs. A low-field MRI has four main advantages. First, it prevents the BOLD effect from contaminating the spin-lock effect because of the BOLD effect arising from magnetic susceptibility changes is expected to be linearly proportional to the static field strength [24]. Assuming the tens milliseconds echo time (TE) and 0.3 T MRI, the signal change caused by the BOLD effect was at the subpercentage level, which is sufficiently small to be ignored. Second, it can alleviate the specific absorption rate (SAR) of the spin-lock preparation sequence. Decreasing the field strength lowers the Larmor frequency, which reduces the RF irradiation energy. Third, it does not always need superconducting magnets, and it has low initial and maintenance costs. Forth, it can decrease the risk of serious magnetic attraction accidents. Naturally, these advantages are common for ultra-low-field fMRI with SQUIDs and optically pumped magnetometers (OPMs).

However, a low-field MRI lowers the signal-to-noise ratio (SNR), which disturbs the measurement of brain functions. Therefore, the objective of this study is to investigate the feasibility of a low-field fMRI with spin-lock preparation sequences. In particular, we focused on the MR signal-change ratio caused by a tiny oscillating magnetic field as a pseudo neural field.

In this study, we first analyze the magnetization dynamics using certain assumptions based on our previous report[25]. Then, we perform phantom experiments to investigate the changes in the MR signal as a function of the target signal strength and phase. Furthermore, we investigate the detectable threshold using statistical tests.

Section snippets

Theory

We describe the magnetization dynamics under spin-lock preparation using the Bloch equation, with Fig. 1, Fig. 2 showing the SIRS and SL-Mz pulse sequences and magnetization dynamics, respectively. SIRS preparation sequence is composed of four parts: 90° pulse, spin-lock pulse, -90° pulse and gradient spoiler. The 90° pulse makes the magnetization flip to x-y planes, and then the spin-lock pulse is applied to the same direction of the magnetization. When there is an oscillating magnetic field

Material and methods

To investigate the MR signal change as a function of the target signal strength, phase, and the minimum detectable signal strength, we performed phantom experiments using a 0.3-T MRI scanner (Hitachi, Ltd.). First, we focused on the detectable strength of the target magnetic fields for each method and examined it using statistical tests. After that, we performed a simulation to determine the current dipole moment that simulated the neural currents.

Phantom experiments

Fig. 5 shows the SIRS and SL-Mz phantom images for varying amplitudes of the oscillating magnetic field. The SIRS images became darker as the amplitude increased. The SL-Mz images were basically noisy, but as the amplitude of the magnetic field increased, the voxels in the loop coil became brighter.

Fig. 6 illustrates the MR signal changes in the SIRS and SL-Mz images as a function of the target signal strength, that is, the amplitude of the magnetic field. The term Son is the pixel value of the

Discussion

Fig. 5 shows that for the SIRS images, the voxels near the signal source become darker as the signal strength becomes large because SIRS reduces longitudinal magnetization in the presence of the oscillating magnetic fields. Similarly, for the SL-Mz images, the voxels near the signal source become bright because the SL-Mz visualizes another magnetization component after the spin-lock pulse, and its magnitude depends on the amplitude of the magnetic fields.

For the MR signal change as a function

Conclusion

We demonstrated the feasibility of SIRS and SL-Mz in the 0.3-T low-field MRI. First, we theoretically revealed that the SL-Mz was suitable in a low-field MRI whose SNR was small because the MR signal-change ratio caused by SL-Mz was larger than that caused by SIRS when the target signal was small. Second, we performed phantom experiments, and their results were consistent with the theoretical calculation based on the Bloch equation. The estimated fitting parameters were reasonable considering

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was partially supported by a Grants-in-Aid for Researches (15H01813, 20K21560) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan and a scholarship from the Iwadare Scholarship Foundation, Japan.

References (34)

  • W. van der Zwaag et al.

    fMRI at 1.5, 3 and 7 T: characterising BOLD signal changes

    NeuroImage

    (2009)
  • H. Ueda et al.

    Dynamics of magnetization under stimulus-induced rotary saturation sequence

    J. Magn. Reson.

    (2018)
  • S. Ogawa et al.

    The sensitivity of magnetic resonance image signals of a rat brain to changes in the cerebral venous blood oxygenation

    Magn. Reson. Med.

    (1993)
  • M. Singh

    Sensitivity of MR phase shift to detect evoked neuromagnetic fields inside the head

    IEEE Trans. Nucl. Sci.

    (1994)
  • H. Kamei et al.

    Neuronal current distribution imaging using magnetic resonance

    IEEE Trans. Magn.

    (1999)
  • S.C. Li et al.

    Comparison of BOLD and direct-MR neuronal detection (DND) in the human visual cortex at 3T

    Magn. Reson. Med.

    (2008)
  • L.M. Parkes et al.

    Inability to directly detect magnetic field changes associated with neuronal activity

    Magn. Reson. Med.

    (2007)
  • Cited by (9)

    View all citing articles on Scopus
    View full text