Introduction to SWIFT (Sweep Imaging with Fourier Transformation) for Magnetic Resonance Imaging of Materials
- PDF / 259,811 Bytes
- 5 Pages / 612 x 792 pts (letter) Page_size
- 116 Downloads / 206 Views
0984-MM09-06
Introduction to SWIFT (Sweep Imaging with Fourier Transformation) for Magnetic Resonance Imaging of Materials Curt Corum, Djaudat Idiyatullin, Steen Moeller, Jang-Yeon Park, and Michael Garwood Center for Magnetic Resonance Research, University of Minnesota, 2021 6th Street SE, Minneapolis, MN, 55455
ABSTRACT A novel, fast, and quiet method of magnetic resonance imaging (MRI), called SWIFT (sweep imaging with Fourier transformation) has recently been introduced. In addition to SWIFT's potential for in-vivo MRI, it creates new opportunities for MRI of materials. SWIFT currently operates in 3d radial acquisition mode. A series of segmented hyperbolic secant excitation pulses is accompanied by acquisition in the gaps. After correlation with the pulse, each excitation results in a free induction decay (FID). The spectrum corresponding to the FID is a projection. There is very little “dead time” between excitation and acquisition, making SWIFT useful for imaging of short T2 materials, but in total imaging times comparable to fast gradient echo sequences. We anticipate great interest in this new MRI sequence in the materials MRI community and look forward to exploring SWIFT's advantages and potential relative to existing short T2 imaging techniques. INTRODUCTION The SWIFT sequence[1] consists of a simple series of RF pulses, incremented gradients, and simultaneous data acquisition (see figure 1). In SWIFT, a time-domain signal is acquired during the gaps of a segmented frequency-swept pulse. The field gradient used for spatial encoding is not pulsed on and off; instead, its orientation is stepped in an incremental manner. The gradient waveforms are nearly continuous, resulting in low stress on hardware and quiet operation. With virtually simultaneous excitation and signal acquisition, new possibilities exist for imaging materials consisting of spins with fast transverse relaxation rates, T2, in the 25-250 µs range, such as polymers, macromolecules, semisolids, and quadrupolar nuclei. Due to the frequency sweep and sequential excitation, peak RF power requirements are reduced compared to hard pulse excitation. The gradient creates the space-to-frequency mapping well known from basic MRI. The RF pulse excites transverse magnetization. The time of excitation is largely determined by the frequency sweep of the hyperbolic secant (HS1) pulse [2]. Magnetization is excited when the frequency sweep of the pulse reaches its local resonance offset[3] in the gradient.
The correlation method[4, 5] effectively removes the phase accumulated due to differing excitation times, and recreates an FID as if all the magnetization had been simultaneously excited. Mathematically the excitation process can be represented (in the limit of linearity) as a convolution: d(t) = p(t )* s(t) where d(t) is the acquired time domain data, p(t) is the complex amplitude of the RF pulse and s(t) is the underlying FID if the sample had been excited by a short hard pulse (an impulse response). “*” represents the convolution operation.
pw
τa
Data Loading...
