Directly assessing the integrity of myelin in white matter may be important for diagnosis and assessment of prognosis in multiple sclerosis (MS), a disabling disease of the central nervous system ¹. However, the protons in myelin have extremely short T2s (less than 1 ms) and cannot be directly imaged with conventional clinical MRI sequences which have TEs of several milliseconds or longer ^2-4. As a result, conventional clinical sequences only provide an indirect assessment of myelin. It would be a major achievement to directly image myelin and quantitatively evaluate its MR properties such as T2*, T1 and proton density, and potentially so provide a more specific and sensitive evaluation of the damage to this tissue in MS. Adiabatic inversion recovery prepared ultrashort echo time (IR-UTE) sequences can detect signal from myelin protons and efficiently suppress the signal from water ⁵. In this study we aimed to further validate the IR-UTE technique in sheep brain using a D2O exchange model.The IR-UTE contrast mechanism is shown in Figure 1. The longitudinal magnetization of long T2 white matter is inverted by the adiabatic inversion pulse (duration = 8.64 ms) while the short T2 myelin signal is largely saturated. UTE data acquisition starts after a delay of time to inversion (TI) when the inverted long T2 magnetization approaches the nulling point in order to detect myelin signal recovered during TI without water contamination. Six sheep brain specimens were purchased from a local slaughter house. The brainstem of each sheep brain was dissected into ~5 mm slab and imaged using a GE 3T Signa TwinSpeed MR scanner (GE Healthcare Technologies, Milwaukee, MI) and a 1-inch solenoid coil for signal excitation and reception. The MR protocol included four sequences, including a proton-density weighted fast spin echo (PD-FSE) sequence (TR = 6000 ms) to measure PD of long T2 white matter, an IR-FSE sequence (TR = 2000 ms, TI = 50, 75, 100, 150, 200, 300, 400, 500, 700, 1000, 1500 ms) to measure T1 of long T2 white matter, an IR-UTE sequence (TR=1000ms, TI~240 ms) to measure PD of myelin, and IR-UTE imaging with a series of TEs (TE = 0.01, 0.1, 0.2, 0.4, 1.2, 2.0 ms) to measure T2* of myelin in white matter of the brain stem. Other imaging parameters included a flip angle of 60o, a bandwidth of 62.5 kHz, a FOV of 4 cm, a slice thickness of 2 mm, reconstruction matrix of 192×192. The same MR protocol was applied to each brain stem sample before and after D2O exchange for 11 hours and one week. During the D2O exchange, each brainstem was immersed in 30 ml D2O solution (99.8% purity). Changes in PD of long T2 white matter and myelin were measured as the signal change in PD-FSE and IR-UTE images as a function of exchange time, respectively. T2* was quantified using a mono-exponential decay model. Myelin T2* was also plotted against D2O exchange time.

Figure 2 shows PD-FSE images of a brainstem sample imaged at different stages of D2O exchange. There is a significant signal drop with increased D2O exchange time, consistent with the replacement of H2O with D2O which has no MR signal.

Figure 3 shows an example of IR-UTE imaging of the brainstem at different TEs, and T2* analysis for different stages of D2O exchanges. A short consistent T2* was demonstrated for the brainstem at different stages of D2O exchange, consistent with myelin being detected and quantified.

The IR-UTE sequence allow uniform inversion and nulling of the long T2 components in white matter of the brainstem, and can provide selective imaging of myelin. The IR-UTE sequence can also be used to measure the MR properties such as T2* and tissue properties such as the PD of myelin. These approaches have the a priori potential to detect pathological changes in myelin, and thus provide a new opportunity to characterize MS in a more specific and potentially more sensitive way than has been possible with conventional clinical sequences which are indirect ⁶.