2288
Multi-exponential decay of dipolar order in spinal cord and its correlation to spin diffusion1School of Chemistry, Tel Aviv University, Tel Aviv, Israel, 2SQITS/NICHD/, National Institutes of Health, Bethesda, MD, United States
Synopsis
Images obtained by inhomogeneous magnetization transfer were analyzed using models involving Zeeman and dipolar orders. These orders are affected by spin diffusion and relaxation. In the current study a method that enables the measurement of these processes in spinal cord by the same pulse sequence is presented. Magnetization transfer time (MT, spin diffusion) between CH2 groups is ~0.1ms while the MT between them and CH3 groups is much slower (~2ms). The dipolar order is found to decay by three exponentials process (0.11, 0.86, 11.4ms). The fastest decaying component is compatible with the spin diffusion between the CH2.
Introduction:
Inhomogeneous magnetization transfer (ihMT) images that report myelin content of nerves were analyzed using models involving Zeeman and dipolar orders (spherical tensors T1,0 and T2,0, respectively)1. These models describe the dipolar order decay by a single exponential and do not incorporate effects of spin diffusion. Direct measurement of these two types of processes can be helpful in modeling and interpreting ihMT1 imaging data.Materials and Methods
The above processes were studied by a new pulse sequence that combines double (DQ) and zero (ZQ) quantum filters with magnetization transfer (DQF-ZQF-MT): 90°-τ-90°-tDQ-90°-tZQ-90°-tDQ-90°-τ-90°-tLM-90°-acq, (90° pulse was 17μs) τ is the time evolution of the second-rank single-quantum tensor (T2,1), tDQ, tZQ are the DQ and ZQ evolution periods, and during the period tLM Zeeman order is exchanging between spins (spin diffusion). This approach enables the measurement of Zeeman and dipolar orders dynamics using the same pulse sequence. Spectra were obtained using a 9.4T Bruker spectrometer with an Avance III HD console. Porcine spinal cords were obtained from freshly sacrificed animals and soaked in saline.Results
In Fig. 1 we show dependence on tLM while keeping τ=15μs and tZQ=2.5μs. While the lineshape of widths larger than ~2kHz varies, the integrals of the spectra are constant, proving that the observed processes is magnetization transfer (MT, spin diffusion). Two types of spectra are evident: one with width of ~20 kHz2-4 and its maximum close to the water peak (clamped peak) and a narrower one with width of ~2kHz and peak position at 3.8ppm from the water3. At longer tLM values MT to water was observed3 In Fig. 2 we show the spectral dependence on tZQ while keeping τ=15μs and tDQ=2.5μs. The lineshape is unchanged while the amplitude can be described by a three exponential decay.Discussion
The spectrum at tLM =1μs in Fig. 1 consists of two peaks separated by ~20kHz, typical of dipolar interaction between protons in the CH2 group. At tLM =100μs the splitting collapses while its integral is similar to the one obtained for tLM=1μs, indicating spin diffusion that was modeled, based on three assumptions: (1) there are two types of spins, one having dipolar interaction that yield splitting of ~20kHz2-4 and are selected by setting τ=15μs (Fig. 1), and another that has smaller dipolar interactions and whose magnetization is suppressed by the short value of τ(15μs); (2) there is dipolar interaction, ωD, between the selected and the suppressed spins; (3) initial condition for the density matrix, ρ(tLM=0)=I1z to yield:
ρ(tLM)=Iz1cos2(21/2ωDtLM/2)+Iz2sin2(21/2 ωDtLM/2)-i21/2T1,0(two spins)sin(21/2ωDtLM)
for the tLM dependence of the density matrix, where I1z and I2z are the z components of the selected and the suppressed spins respectively and ωD=γ2h(3cos2θ-1)/4πr3, r is the norm of the internuclear vector and θ is its orientation with respect to the magnetic field, γ is the gyromagnetic ratio. This expression describes spin diffusion since the expectation value of I1z+I2z (tr[ρ(I1z+I2z)]) is constant. This result is consistent with the experimental results (Fig. 1). For tLM values in the range of 1-5 ms there is a buildup of a peak with resonance frequency typical of CH3 groups, indicating MT between CH2 and CH3 groups. The slower rate of the latter process is probably the result of CH3 rotation around its axis. Previous studies1 suggested that under r.f. irradiation of one of the two peaks resulting from dipolar interactions5 the Zeeman order can evolve into dipolar one, while irradiation of both peaks will suppress the process1. Thus, spin diffusion is expected to reduce the efficiency of yielding dipolar order in these experiments since saturation of one of the peaks will be transferred to the other peaks. In Fig, 2 the dynamics of dipolar order are shown. It was best fit to three exponentials. The shortest decay time TD1 is compatible with spin diffusion shown in Fig. 1, and is therefore consistent with the interpretation above of the role that spin diffusion plays in reducing the dipolar order. The longest time TD3 is compatible with decay time of dipolar order measured for CH3 peak (results not shown), by setting t=160μs and varying tZQ, which is consistent with the values of the dipolar decay time used to interpret the ihMT imaging sequence1.
Conclusion
Spin diffusion and dipolar order decay can be measured by the same pulse sequence. MT (spin diffusion) between CH2 groups is very fast (~0.1ms) while the MT between them and CH3 groups are much slower (~2ms). Spin diffusion is an important source of reducing observed dipolar order and should be taken into account when analyzing the ihMT imaging sequence or any measurement that involves those orders.Acknowledgements
This research was partially supported by a US-Israel Binational Science Foundation grant. PJB was supported by the Intramural Research Program of the NICHD, NIHReferences
1. Varma G, Duhamel G, de Bazelaire C, and Alsop DC. Magnetization Transfer From Inhomogeneously Broadened Lines: A Potential Marker for Myelin. Magn. Reson. Med. 2015:73;614-622.
2. Wilhelm MJ, Ong HH, Wehrli SL, Li C, Tsai PH, Hackney DB, Wehrli FW. Direct Magnetic Resonance Detection of Myelin and Prospects for Quantitative Iimaging of Myelin Density. Proc. Natl. Acad. Sci. 2012:109;9605-9610.
3. Eliav U, Basser PJ, Navon G, Magnetization Transfer Among Non-Aqueous Species and Between Them and Water in Spinal Cord. ISMRM, Paris, 2018, poster 838.
4. Yang W, Lee JS, Windschuh J, Leninger M, Traaseth N, Jerschow A. Magnetization Transfer in Lipids - Role of Exchangeable Groups and Water Binding. ISMRM, Paris, 2018, poster 5120.
5. Eliav U, Naumann C, Navon G, Kuchel PW, Double Quantum Transition as the Origin of the Central Dip in the Z-spectrum of HDO in Variably Stretched Gel. J. Magn. Reson. 2009:198;197-205.
Figures
