UTE Imaging in the Brain with RobustLong-T2 Suppression Pulses

Peder E. Z. Larson, John M. Pauly, Dwight G. Nishimura

Department of Electrical Engineering, StanfordUniversity, Stanford, CA, USA

Introduction

Ultra-short echo time (UTE) sequences for MRI can image tissues with very short T2 components that normally have no signal in conventional imaging techniques [1]. Long-T2 suppression is necessary in many UTE applications because the short-T2 components are often obscured and have less signal than the long-T2 components, especially in the brain, where high signal from white and gray matter obscure the myelin [2]. This can be done by subtracting a later echo image, or also with long, low-amplitude suppression pulses that saturate long-T2 spins but leave short-T2 spins relatively unaffected. We have improved the off-resonance characteristics of these pulses [3].

Methods

Figure 1 shows our UTE sequence where a long-T2 suppression pulse and dephasing gradient are applied before the half-pulse excitation. Two acquisitions with alternating slice select gradients are summed for slice selection. Projection reconstruction readout gradients with ramp sampling and a 125 kHz bandwidth were used. The scan parameters were TE = 80 us, TR = 500 ms, 60 flip angle, 5 mm slice thickness, 1 mm in-plane resolution, and 4:15 per image. A transmit/receive head coil was used in a GE Excite 1.5T scanner. A 16 ms rectangular pulse as well as a 40 ms time-bandwidth 2.4 maximum phase saturation pulse, designed with the SLR pulse design algorithm [4], were used for long-T2 suppression. Their simulated T2 and off-resonance responsesare shown in figure 2.

Results

Without long-T2 suppression, the UTE image shows little evidence of any short-T2 components in the cerebrum. When the long-T2 suppression pulses are used, a white matter short-T2 component, which we believe to be associated with the myelin, is clearly visible. We have acquired later TE images, not shown, to confirm that this is a short-T2 component and not unsuppressed white matter. The falx cerebri, another short-T2 tissue, is also more visible with the suppression pulses.

Discussion/Conclusion

Both suppression pulses provide similar T2 contrast but the rectangular suppression contrast varies across the image, with decreased signal in the right anterior portion of the brain and increased signal posteriorly. This poor suppression is likely due to the off-resonance sensitivity of the rectangular pulse, which is not seen when the more robust SLR-designed pulse was used.

UTE brain imagingcan be improved by tailoring the suppression pulses and sequence to the relaxation and resonance characteristics of the short-T2 components, which should be determined withfurther studies, such as T2* and off-resonance mapping. Applications in the spinal cord can also be explored.A short-T2 component we believe to be myelin, which is otherwise completely obscured by the gray and white matter, can be clearly and consistently visualized using UTE imaging with our long-T2 suppression pulses.

References

[1] Gatehouse, P, et al, Clin Radiol 2003, 58:1-19.[3] Larson, PEZ, et al, Proc 13th ISMRM 2005, p. 2345.

[2] Nayak, KS, et al, Proc. 8th ISMRM 2000, p. 509.[4] Pauly, JM, et al, IEEE T-MI 1991, 10: 53-65.

Acknowledgements: We thank Dr. Krishna Nayak for his assistance, GE Healthcare, and NIH grant R01 EB002524.