Electricity: measuring and testing – Particle precession resonance – Spectrometer components
Reexamination Certificate
1999-05-18
2001-08-14
Oda, Christine (Department: 2862)
Electricity: measuring and testing
Particle precession resonance
Spectrometer components
Reexamination Certificate
active
06275040
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to designing and applying spectral-spatial pulses for magnetic resonance imaging.
BACKGROUND OF THE INVENTION
Frequency selective pulses are well known in the art. One type of frequency selective pulses is a standard volume (non-spectral-spatial) pulse for fat saturation, which is applied non-spatially selective to an entire image volume, for example by applying a narrow-band pulse without a slice selection gradient. Another type of frequency selective pulse is a spectral-spatial pulse, which is used to selectively excite spins at a spatial location z and a spectral frequency &ugr;. A particular important application of frequency selective pulses is selective imaging of water-based tissue in the presence of fat-based tissue. In this application, it is desirable that, over a very small frequency range (e.g., the difference between the frequency bands of fat and water), there be a high contrast between excited and non-excited tissue types. One technique is to selectively saturate the fat tissue, so that when the entire region is excited, the fat, which is pre-saturated, will not become excited and will not generate a significant signal. Another technique is to selectively excite the water tissue so that only it generates a signal.
A benefit of spectral-spatial pulses is that they are relatively immune to the effects of magnetic field inhomogenity, especially those caused by changes in magnetic susceptibility. In human body imaging, there are many susceptibility changes between image slices, so a spectral-spatial pulse, which can be adapted separately for each slice, is inherently better than a volume saturation pulse for exciting only a specific frequency range.
Another advantage is reduced cross-talk between fat and water magnetization. In an imaging procedure when n slices are imaged, a volume fat saturation pulse may be applied as many as “n” times, for each TR (the time between repeated successive excitations of the same slice), thus, the entire volume is “saturated” n times. If the saturation is not exact, for example due to field inhomogenities, some of the water may also be excited or saturated. In contrast, the spectral-spatial pulse saturates a particular portion only once every TR, thus, any adverse magnetization effects, which are signal, do not accumulate and/or they decay in time.
It should be noted that the “fat” frequency is typically broad, in many cases, there is even a significant overlap between the broad fat band and the narrow water band. Thus, any frequency shift caused by inhomogenity is quite likely to move the target frequency of the pulse and saturate the water signal by mistake (in a fat saturation pulse) or excite the fat (in a a selective water excitation pulse).
Spectral-spatial excitation pulses are used, for example in short TR Gradient echo imaging. A Fat saturation pulse applies a large tip angle very often. Thus, even a small frequency shift (due to field inhomogenity) can unintentionally saturate the nearby frequency band of water. On the other hand, the small tip angle for water excitation will not usually cause a significant amount of fat signal, even if the field is a little inhomogeneous.
A spectral-spatial pulse comprises a train of sub-pulses, the train being applied in synchrony with an oscillating slice selection gradient. Each sub-pulse acts only on the slice at a location z by the slice selection gradient, while the accumulation of phase along the train results in the saturation for a particular frequency &ngr;. More precisely, the end-magnetization comprises a series of magnetization lobes in &ngr; separated by 1/&tgr; Hz, where &tgr; is the delay between adjacent sub-pulses. The width of the slice s determined by the slice selection gradient. In selective fat/water suppression/excitation, the fat-water frequency separation is matched to the spectral-spatial pulse such that one of the frequencies of water and fat are inside a magnetization lobe and the other is outside a magnetization lobe.
As indicated above, a spectral-spatial pulse may be used to excite (saturate) an off-center slice portion. If the applied gradient is G(t), the frequency shift of the slice is &ggr;G(t)z
i
, where z
i
is the slice location. Thus, the applied frequency f(t) is f(t)=&ggr;G(t) z
i
+&ngr;
0
. A spin with frequency &ngr;
1
can be excited (saturated) by setting &ngr;
0
=&ngr;
1
.
An excitation pulse excites a desired set of spins to achieve an M
xy
magnetization vector. Saturation works by achieving an M
z
magnetization vector. The advantage of some excitation pulses is that they can be used to apply small tip angles, which pulses are less likely to excite undesirable tissue (with other frequencies). A disadvantage of excitation pulses is that a spectral-spatial excitation pulse is longer than a non-frequency selective excitation, so that it affects the TE in short TE sequences. These sequences are thus made more sensitive to patient movement and field inhomogeneity. A saturation pulse does not affect the TE, since it is applied before the pulse, so longer pulses with a sharper transition can be used. For example, saturation pulses are used with spin echo (SE) and fast spin echo FSE) sequences. However, the saturation pulse is sensitive to RF field inhomogenity. Variations in B
1
change the tip angle and cause the saturation to be incomplete.
Spectral-spatial pulses are usually classified as type I or as type II pulses.
FIGS. 1A and 1B
show a type I pulse, in which RF sub-pulses are applied (with a spacing&tgr;) only during one polarity of the oscillating gradient.
FIGS. 1C and 1D
show a type II pulse in which RF sub-pulses are applied during both polarities of the gradient field. If &ggr;G&Dgr;z>>&Dgr;&ngr;, where &ggr;G&Dgr;z is the frequency bandwidth along the z axis and &Dgr;&ngr; is the frequency bandwidth along &ngr;, the pulse sequence of
FIG. 1A
(and
FIG. 1C
) can be considered equivalent to that of
FIG. 1B
(and FIG.
1
D), in which inherently refocused sub-pulses arc replaced by delta-type sub-pulses at spacing &tgr;, for example as described in “A Linear Class of Large Tip-Angle Selective Excitation Pulses”, by J. Patly, D. Nishirura and A. Macovski in
J. Mag. Res
. Vol. 82 pp. 571-587 (1989), the disclosure of which is incorporated herein by reference.
As noted above, it is desirable for one of frequencies of the two differentiated tissues (fat, water) to be in a magnetization lobe and for the other one to be outside a magnetization lobe. The locations of the magnetization lobes are very sensitive to field variations, so the two frequencies are usually positioned so one is in the middle of a lobe and one is in the center area between the lobe and an adjacent lobe. This means that ½&tgr;≧&Dgr;&ngr;
wf
. The fat-water separation increases with the field strength, so a smaller &tgr; is required in higher field strengths to obtain a same quality of fat water separation. However, the slice thickness also constrains &tgr;. The slice thickness (&Dgr;z) and the time-bandwidth (TB) product are determined by &ggr;S=TB/&Dgr;z, where S is the effective area of the gradient during the sub-pulse,
S
=
∫
0
TRF
G
(
u
)
ⅆ
u
and TRF is the duration of the RF pulse. In type I pulses, TRF is about equal to &tgr;/2, since there is no RF over half of the gradient (the negative part). In type II pulses, TRF is about equal to &tgr;. Thus, S for type II pulses can be about three times as large as for a type I pulse using same maximnum gradient amplitudes and slew rates. Consequently, the slice width of a type II pulse can be about ⅓ the width of a comparable type I pulse. This is a requirement in high-field MRI, in which minimum slices widths are desired, even as the fat-water frequency separation increases.
A limitation of type II spectral-spatial pulses is that they have many magnetization lobes.
FIG. 2
illustrates the M
xy
magnetization profile of a prior art type II selective water excitation pulse. However, this pu
Cowan Liebowitz & Latman P.C.
Dippert William H.
Oda Christine
Vargas Dixomara
LandOfFree
Designing spectral-spatial pulses does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Designing spectral-spatial pulses, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Designing spectral-spatial pulses will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2508264