OPTIMALLY-SHAPED RF PULSE FOR MRI APPLICATIONS

Abstract

A method for shaping an RF pulse for use with an MRI system includes shaping an RF pulse for use with an MRI system that uses an RF coil. The RF pulse is shaped to reduce changes in B1 amplitude and in an off-resonance effect with respect to Larmor frequency as a function of distance from the RF coil.

Claims

1. A method for shaping an RF pulse for use with a magnetic resonance imaging (MRI) system, the method comprising: shaping a radio-frequency (RF) pulse for use with an MRI system that uses an RF coil, wherein said RF pulse is shaped to reduce changes in B.sub.1 amplitude and in an off-resonance effect with respect to Larmor frequency as a function of distance from said RF coil.

2. The method according to claim 1, wherein said RF pulse has a non-flat-shaped spectrum.

3. The method according to claim 2, wherein there is a quadratic increase in the B.sub.1 amplitude versus frequency around a center of the spectrum.

4. The method according to claim 2, wherein there is a linear increase in the B.sub.1 amplitude versus frequency around a center of the spectrum.

5. The method according to claim 1, wherein said RF coil comprises a spiral shaped RF coil and said MRI system uses a nuclear magnetic resonance (NMR) signal in a Carr-Purcell-Meiboom-Gill (CPMG) sequence.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:

[0011] FIGS. 1A and 1B are graphical illustrations of a prior art RF pulse envelope and its spectrum, respectively;

[0012] FIG. 2 is a graphical illustration of the B.sub.1(x,y) field (in Tesla) vs. z of a spiral antenna like the one used in the prior art version of the CLEAR SIGHT system of ClearCut Medical Ltd., wherein x and y are in the plane of the antenna and z is perpendicular to the x-y plane;

[0013] FIG. 3 is a graphical illustration of the Larmor frequency (in MHz) vs. y and z of the spiral antenna of FIG. 2;

[0014] FIGS. 4A and 4B are graphical illustrations of an RF pulse envelope and its spectrum, respectively, in accordance with a non-limiting embodiment of the present invention, with a quadratic increase in amplitude vs frequency around the center of the spectrum; and

[0015] FIGS. 5A and 5B are graphical illustrations of another RF pulse envelope and its spectrum, respectively, in accordance with a non-limiting embodiment of the present invention, this time with a linear increase in amplitude vs frequency around the center of the spectrum.

DETAILED DESCRIPTION OF EMBODIMENTS

[0016] Reference is now made to FIGS. 4A and 4B, which illustrate an RF pulse envelope and its spectrum, respectively, in accordance with a non-limiting embodiment of the present invention.

[0017] The RF pulse is shaped to overcome two effects: [0018] 1. The change in the B.sub.1 amplitude in space with Larmor frequency (i.e., with distance from the RF coil), to counteract the effect of the change in the B.sub.1 field. [0019] 2. The off-resonance effect, which also changes with Larmor frequency (i.e., with distance from the RF coil), to counteract the fact that the pulse is less effective off-resonance than on-resonance.

[0020] The invention encompasses different pulses or families of pulses (with amplitude varying as a function of frequency) that optimize the NMR signal obtained in a CPMG sequence with a spiral shaped RF coil in the presence of a strong linear gradient. FIGS. 4A and 4B illustrate an example of such a pulse.

[0021] Optimization of the pulse may be done by defining the pulse frequency response and testing the efficacy at exciting magnetization using an NMR simulation. A non-limiting way of simulation is that shown in the article by Y. Zur, An algorithm to calculate the NMR signal of an multi spin-echo sequence with relaxation and spin-diffusion, J Magn. Reson. 171 :97-106, 2004. One begins with a flat spectrum as a reference, calculating the B.sub.1 pulse (i.e. B.sub.1(t)) as the inverse Fourier transform of the specified frequency domain vector. Next one starts to modify the Fourier coefficients, increasing those that are most efficient at spins with lower values of the B.sub.1 field (e.g. because they are farther from the antenna). Each time one modifies the spectrum, one calculates the B.sub.1 pulse (i.e., B.sub.1(t)) as the inverse Fourier transform of the specified frequency domain vector and tests its efficacy at exciting magnetization using an NMR simulation.

[0022] The inventors have tested increasing the amplitude vs frequency using various B.sub.1 field polynomial models (linear vs offset, quadratic, etc.), and have “tweaked” the amplitude manually at individual frequencies to optimize the performance. In the pulse shown in FIGS. 4A-4B, the response is roughly quadratic vs off-resonance, while in FIGS. 5A-5B the response is roughly linear vs off-resonance. This improves the performance (as compared with the reference pulse shown in FIGS. 1A-1B) because spins at large off-resonance get “extra” energy, which compensate (at least partly) for the fact that the B.sub.1 field at those positions is weaker.

[0023] Accordingly, the invention uses a pulse whose spectrum is substantially non-flat, to compensate as much as possible for the two factors mentioned above - the change of the B.sub.1 field in space and the variation of the Larmor frequency in space.

[0024] The use of such an RF pulse can improve the sensitivity of the system, since the spin echo condition will be more closely obeyed at more points in the voxel under study.