Method and apparatus for magnetic resonance imaging

Abstract

In a method and apparatus for magnetic resonance imaging, a flip angle and/or inversion time of a spectrum suppression pulse is calculated according to a steady state condition of a longitudinal magnetization component of a spectrum composition suppressed by the spectrum suppression pulse and a zero crossing point condition of the longitudinal magnetization component. Raw magnetic resonance image data are acquired by applying a magnetic resonance imaging sequence that includes the spectrum suppression pulse provided with the flip angle and/or the inversion time.

Claims

1. A method for acquiring magnetic resonance data, comprising: in a processor, automatically calculating a flip angle and/or inversion time of a spectrum suppression pulse according to a steady state condition of a longitudinal magnetization component of a spectrum composition suppressed by a spectrum suppression pulse, to be generated in a magnetic resonance imaging sequence, and a zero crossing point condition of the longitudinal magnetization component; operating a magnetic resonance data acquisition unit to acquire magnetic resonance data by applying said magnetic resonance imaging sequence comprising the spectrum suppression pulse provided with the flip angle and/or the inversion time, to a subject in the data acquisition unit; entering said magnetic resonance data into memory locations in an electronic memory organized as k-space, and making k-space filled with said magnetic resonance data available in electronic form as a data file from said memory in a format suitable for further processing; emitting the spectrum suppression pulse in the magnetic resonance imaging sequence before one or more excitation pulses and using, as the zero crossing point condition of the longitudinal magnetization component, making the longitudinal magnetization component of the spectrum composition zero before an excitation pulse corresponding to a center line of k-space; and the inversion time comprising a first inversion time and a second inversion time, and calculating the flip angle or the inversion time of the spectrum suppression pulse, using a general formula, and the flip angle or the inversion time of the spectrum suppression pulse, wherein said general formula is:
{E+F.Math.C.Math.(D.sup.N-11)/(D1)+F.Math.A.Math.D.sup.N-1}/(1F.Math.B.Math.D.sup.N-1)={C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1}/(B.Math.D.sup.KSpaceCenterLine-1), where, A=1E.sub.Tl1 B=cos().Math.E.sub.Tl1 C=1E.sub.Trep D=cos().Math.E.sub.Trep E=1E.sub.Tr F=cos().Math.E.sub.Tr E.sub.Tl1=exp(Tl1/T1) E.sub.Trep=exp(Trep/T1) E.sub.Tr=exp(Tr/T1) Tl1=Tl1Min+TlFill Tr=TrMin+TrFill is a flip angle of an excitation pulse, is a flip angle of a spectrum suppression pulse, T1 is a longitudinal relaxation time of a spectrum composition suppressed by a spectrum suppression pulse, N is the number of excitation pulses between two spectrum suppression pulses, KSpaceCenterLine is the index number of a the center line of k-space, TI1Min is the minimum time interval between the center of a spectrum suppression pulse and the center of a first excitation pulse following the spectrum suppression pulse, Trep is the time between continuous excitation pulses, TrMin is the minimum time interval between the center of the last excitation pulse in a pulse train and the center of a spectrum suppression pulse in a next pulse string of the magnetic resonance imaging sequence, TIFill is the first inversion time, and TrFill is the second inversion time.

2. A method as claimed in claim 1, further comprising setting the first inversion time TIFill and the second inversion time TrFill as 0, and calculating the flip angle in the step of calculating the flip angle or the inversion time of the spectrum suppression pulse.

3. A method as claimed in claim 1, further comprising: setting the flip angle as 180 degrees; and calculating the flip angle or the inversion time of the spectrum suppression pulse by setting the second inversion time TrFill as 0, and calculating the first inversion time TIFill.

4. A method as claimed in claim 3, wherein, when the calculated first inversion time TIFill is smaller than 0, calculating the flip angle or the inversion time of the spectrum suppression pulse further by: setting the first inversion time TIFill as 0, and calculating the second inversion time TrFill.

5. A method as claimed in claim 1, comprising: setting the flip angle as 180 degrees; and the step of calculating the flip angle or the inversion time of the spectrum suppression pulse by setting the first inversion time TIFill as 0, and calculating the second inversion time TrFill.

6. A method as claimed in claim 5, wherein, when the calculated second inversion time TrFill is smaller than 0, calculating the flip angle or the inversion time of the spectrum suppression pulse further by: setting the second inversion time TrFill as 0, and calculating the first inversion time TIFill.

7. A magnetic resonance apparatus comprising: a magnetic resonance data acquisition unit; a processor configured to automatically calculate a flip angle and/or inversion time of a spectrum suppression pulse according to a steady state condition of a longitudinal magnetization component of a spectrum composition suppressed by a spectrum suppression pulse, to be generated in a magnetic resonance imaging sequence, and a zero crossing point condition of the longitudinal magnetization component; a control unit configured to operate the magnetic resonance data acquisition unit to acquire magnetic resonance data by applying said magnetic resonance imaging sequence comprising the spectrum suppression pulse provided with the flip angle and/or the inversion time, to a subject in the data acquisition unit; said control unit being configured to enter said magnetic resonance data into memory locations in an electronic memory organized as k-space, and to make k-space filled with said magnetic resonance data available in electronic form as a data file from said memory in a format suitable for further processing; said control unit being configured to operate said magnetic resonance data acquisition unit to emit the spectrum suppression pulse in the magnetic resonance imaging sequence before one or more excitation pulses and using, as the zero crossing point condition of the longitudinal magnetization component, making the longitudinal magnetization component of the spectrum composition zero before an excitation pulse corresponding to a center line of k-space; and the inversion time comprising a first inversion time and a second inversion time, and wherein said control unit is configured to calculate the flip angle or the inversion time of the spectrum suppression pulse, using a general formula, and the flip angle or the inversion time of the spectrum suppression pulse, wherein said general formula is:
{E+F.Math.C.Math.(D.sup.N-11) (D1)+F.Math.A.Math.D.sup.N-1}/(1F.Math.B.Math.D.sup.N-1)={C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KspaceCenterLine-1}/(B.Math.D.sup.KSpaceCenterLine-1), where, A=1E.sub.TI1 B=cos().Math.E.sub.TI1 C=1E.sub.Trep D=cos().Math.E.sub.Trep E=1E.sub.Tr F=cos().Math.E.sub.Tr E.sub.TI1=exp(TI1/T1) E.sub.Trep=exp(Trep/T1) E.sub.Tr=exp(Tr/T1) TI1=TI1Min+TIFill Tr=TrMin+TrFill is a flip angle of an excitation pulse, is a flip angle of a spectrum suppression pulse, T1 is a longitudinal relaxation time of a spectrum composition suppressed by a spectrum suppression pulse, N is the number of excitation pulses between two spectrum suppression pulses, KSpaceCenterLine is the index number of a the center line of k-space, TI1Min is the minimum time interval between the center of a spectrum suppression pulse and the center of a first excitation pulse following the spectrum suppression pulse, Trep is the time between continuous excitation pulses, TrMin is the minimum time interval between the center of the last excitation pulse in a pulse train and the center of a spectrum suppression pulse in a next pulse string of the magnetic resonance imaging sequence, TIFill is the first inversion time, and TrFill is the second inversion time.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a schematic diagram of a spoiled gradient echo sequence provided with a fat suppression pulse.

(2) FIG. 2 shows a sequence chart with a number of excitation pulses being applied between continuous fat suppression pulses.

(3) FIG. 3 shows a flow chart of a method for magnetic resonance imaging using an MRI sequence provided with a spectrum suppression pulse according to the embodiments of the present invention.

(4) FIG. 4 shows a block diagram of a device for magnetic resonance imaging using an MRI sequence provided with a spectrum suppression pulse.

(5) FIGS. 5(a) and 5(b) show the simulation results when a spectrum suppression pulse is a FatSat pulse.

(6) FIGS. 6(a) and 6(b) show the simulation results when a spectrum suppression pulse is a SPAIR pulse.

(7) FIGS. 7(a) and 7(b) show images captured using an MRI sequence provided with an optimized spectrum suppression pulse according to the present invention; and FIG. 7(c) shows an image captured using an MRI sequence not provided with a spectrum suppression pulse.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(8) In order to make the object, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail hereinbelow by way of embodiments.

(9) (Embodiment 1)

(10) This embodiment describes a fat suppression pulse as a specific spectrum suppression pulse.

(11) First, MRI sequences involved in the present invention will be briefly described.

(12) FIG. 1 is a schematic diagram of a spoiled gradient echo sequence provided with a fat suppression pulse. Since the radio-frequency pulse interval TR of the spoiled gradient echo sequence is relatively short and a plurality of excitation pulses can be applied between continuous fat suppression pulses, the measurement time is reduced. Therefore, the sequence is suitable for MRI examinations of the abdomen, the chest and so on which need a short sampling time in order to avoid motion artifacts.

(13) As shown in FIG. 1, at first a fat suppression pulse is applied, and the transverse magnetization vector generated in the XY plane due to the application of the fat suppression pulse is eliminated by the spoiled gradient. And then, radio-frequency excitation pulses, slice selection gradients, phase encoding gradients, etc. constituting a gradient echo sequence are applied (for clarity, only one excitation pulse is shown in the figure), and after the excitation pulses are applied, a gradient echo as shown in FIG. 1 is generated.

(14) FIG. 2 illustrates a sequence chart with a number of excitation pulses being applied between continuous fat suppression pulses. To simplify the accompanying drawing, gradient pulses (selective gradient, phase encoding gradient, spoiled gradient, and so on) and the radio-frequency spoil mechanism are not shown in FIG. 2.

(15) As can be seen from FIG. 2, a pulse train in the sequence is constituted by a fat suppression pulse and a number of radio-frequency excitation pulses. The fat suppression pulse is for example, an aforementioned FatSat pulse, SPAIR pulse, STIR pulse and so on. The shown in the figure is the flip angle of the fat suppression pulse, and is the flip angle of the excitation pulses.

(16) As shown in FIG. 2, Tl1 represents the time interval between the center of a fat suppression pulse and the center of the first excitation pulse following the fat suppression pulse, Tr represents the time interval between the center of the last excitation pulse in a pulse train and the center of a fat suppression pulse in a next pulse train of a magnetic resonance imaging sequence, and Trep represents the time between continuous excitation pulses. Furthermore, in the present invention, Tl1Min is used to represent the minimum value of Tl1, and TlFill represents a first inversion time. It is apparent that Tl1 Min+TlFill=Tl1. Similarly, in the present invention, TrMin is used to represent the minimum value of Tr, TrFill represents a second inversion time, and TrMin+TrFill=Tr.

(17) Hereinafter, k-Space and the k-space center line involved in the present invention will be briefly described.

(18) The mathematical domain known as k-space, also referred to as Fourier space, is an organized set of storage locations in an electronic memory that are filled by the raw data of magnetic resonance signals that are acquired while the pulse sequence is being executed. The entered and stored raw data are converted into images after the Fourier transform is carried out thereon. In the process of acquiring two-dimensional image signals, the size and direction of the frequency encoding gradient field of each magnetic resonance signal remain the same, while the intensity of the phase encoding gradient changes in certain steps, the phase encoding of each magnetic resonance signal changes once, and the captured magnetic resonance signals fill in a line of the k-space in a K.sub.y direction. In the present invention, each excitation pulse of a number of excitation pulses corresponds to a line in the k-space. Contributions of data in different locations of the k-space to the final image are different: data in the center of the k-space mainly determines the signal to noise ratio and contrast information of the image, and signals at the edges of the k-space mainly contribute to information in the aspect of resolution capability of the image.

(19) The optimized fat suppression pulse used in the method for magnetic resonance imaging according to this embodiment will be described below. As described above, the energy efficiency of fat suppression depends largely on the flip angle and inversion time of the fat suppression pulse; hence, the optimization of fat suppression can be achieved by optimizing the flip angle or the inversion time of the fat suppression pulse. Therefore, how to determine the optimized flip angle and inversion time becomes essential. In this regard, this embodiment proposes a general solution, which can be used not only to calculate the flip angle of a fat suppression pulse (such as for FatSat) but also to calculate the inversion time of the fat suppression pulse (for inverted pulses including SPAIR and STIR), and which, at the same time, takes into consideration the situation where each fat suppression pulse in the spoiled gradient echo sequence is followed by a plurality of excitation pulses. The general solution is first described below.

(20) 1. Fat Composition Magnetization Vector Expression Deduced from Bloch Equation

(21) Assume that the spoil mechanism of the spoiled gradient echo sequence works ideally, so that the transverse magnetization component in the XY plane is completely spoiled before each excitation pulse (a pulse) and does not contribute to the subsequent excitation pulses. Also assume that the fat suppression pulse ( pulse) only excites the fat composition. In such conditions, according to the classical Bloch equation, the longitudinal magnetization vector representation of the fat composition before each pulse can be as follows:
M.sub.(1)=M.sub.0(1)
M.sub.(1)=M.sub.0(1E.sub.Tl1)+M.sub.(m).Math.cos().Math.E.sub.Tl1(2)
M.sub.(n)=M.sub.0(1E.sub.Trep)+M.sub.(n1).Math.cos().Math.E.sub.Trep(3)
M.sub.(m+1)=M.sub.0(1E.sub.Tr)+M.sub.(N).Math.cos().Math.E.sub.Tr(4)

(22) where:

(23) E.sub.Tl1=exp(Tl1/T1)

(24) E.sub.Trep=exp(Trep/T1)

(25) E.sub.Tr=exp(Tr/T1)

(26) T1: longitudinal relaxation time of fat tissue,

(27) M.sub.0: initial magnetization vector of fat tissue,

(28) M.sub.(n): longitudinal magnetization component of fat before the nth pulse,

(29) M.sub.(m): longitudinal magnetization component of fat before the mth pulse, and

(30) N: the number of pulses between two pulses, i.e. the number of k-space lines filled by each pulse train.

(31) 2. Establishment of general formula

(32) Since M.sub.0 will be offset during calculation, it can be set as 1, i.e. M.sub.0=1.

(33) Therefore, equations (1) to (4) can be written as
M.sub.(1)=1(5)
M.sub.(1)=A+B.Math.M.sub.(m)(6)
M.sub.(n)=C+D.Math.M.sub.(n1)(7)
M.sub.(m+1)=E+F.Math.M.sub.(N)(8)

(34) where

(35) A=1E.sub.Tl1 B=cos().Math.E.sub.Tl1

(36) C=1E.sub.Trep D=cos().Math.E.sub.Trep

(37) E=1E.sub.Tr F=cos().Math.E.sub.Tr

(38) According to equations (6) and (7), the following can be deduced:

(39) $\begin{array}{cc}{M}_{}\left(1\right)=A+B.Math.M_{}\left(m\right)\begin{array}{c}{M}_{}\left(2\right)=C+D.Math.M_{}\left(1\right)\\ =C+A.Math.D+B.Math.D.Math.M_{}\left(m\right)\end{array}\begin{array}{c}{M}_{}\left(3\right)=C+D.Math.M_{}\left(2\right)\\ =C+C.Math.D+A.Math.D^2+B.Math.D^2.Math.M_{}\left(m\right)\end{array}\begin{array}{c}{M}_{}\left(4\right)=C+D.Math.M_{}\left(3\right)\\ =C+C.Math.D+C.Math.D^2+A.Math.D^3+B.Math.D^3.Math.M_{}\left(m\right)\end{array}\mathrm{.Math.}\mathrm{.Math.}\begin{array}{c}{M}_{}\left(n\right)=C+C.Math.D+C.Math.D^2+C.Math.D^3\mathrm{.Math.}+C.Math.D^{n-2}+\\ A.Math.D^{n-1}+B.Math.D^{n-1}.Math.M_{}\left(m\right)\\ =C\left(D^{n-1}-1\right)/\left(D-1\right)+A.Math.D^{n-1}+B.Math.D^{n-1}.Math.M_{}\left(m\right)\end{array}& \left(9\right)\end{array}$

(40) (where, n=1, 2, 3, . . . , and D1, because cos().Math.E.sub.Trep1)

(41) According to equations (8) and (9), the following can be deduced:

(42) $\begin{array}{cc}\begin{array}{c}{M}_{}\left(m+1\right)=E+F.Math.M_{}\left(N\right)\\ =E+F.Math.\{C.Math.\left(D^{N-1}-1\right)/\left(D-1\right)+\\ A.Math.D^{N-1}+B.Math.D^{N-1}.Math.M_{}\left(m\right)\}\\ =E+F.Math.C.Math.\left(D^{N-1}-1\right)/\left(D-1\right)+\\ F.Math.A.Math.D^{N-1}+F.Math.B.Math.D^{N-1}.Math.M_{}\left(m\right)\end{array}& \left(10\right)\end{array}$

(43) The above equations are deduced by introducing a steady state condition and a zero crossing point condition of a longitudinal magnetization component in the present invention.

(44) The steady state condition is known in the art. In particular, when in a steady state, the longitudinal magnetization component of fat before each f3 pulse should be the same; therefore, the steady state condition is:
M.sub.(m +1)=M.sub.(m)(11).

(45) Furthermore, based on the features of the aforementioned k-space and k-space filling line, it is recognized that when the longitudinal magnetization component of fat before an excitation pulse corresponding to one or more k-space filling lines is zero, it is possible to obtain optimized fat suppression. In particular, when the longitudinal magnetization component of fat before an excitation pulse corresponding to a k-space center line is zero, it has the best fat suppression. Therefore, the zero crossing point condition of the longitudinal magnetization component can be set as:
M.sub.a(KSpaceCenterLine)=0(12),

(46) where, KSpaceCenterLine is the index number of a k-space center line.

(47) According to steady state condition (11) and equation (10), the following can be deduced:
M.sub.R(m)=E+F.Math.C.Math.(D.sup.N-11)/(D1)+F.Math.A.Math.D.sup.N-1+F.Math.B.Math.D.sup.N-1.Math.M.sub.(m),
thereby
M.sub.(m)={E+F.Math.C.Math.(D.sub.N-11)/(D1)+F.Math.A.Math.D.sup.N-1}/(1F.Math.B.Math.D.sup.N-1)(13)

(48) In addition, the following can be deduced by combing zero crossing point condition (12) of the longitudinal magnetization component and equation (9):
0=C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1+B.Math.D.sup.KSpaceCenterLine-1.Math.M.sub.(m)
thereby
M.sub.(m)={C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1}/(B.Math.D.sup.KSpaceCenterLine-1) (14)

(49) According to equations (13) and (14), the following can be deduced:
{E+F.Math.C.Math.(D.sup.N-11)/(D1)+F.Math.A.Math.D.sup.N-1}/(1F.Math.B.Math.D.sup.N-1)={C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1}/(B.Math.D.sup.KspaceCenterLine-1) (15)

(50) The final general formula (15) is thereby deduced. With this formula, a person skilled in the art can easily calculate the flip angle and the inversion time TlFill and TrFill. Nevertheless, for clarity, the flip angle of a FatSat pulse and the inversion time TlFill and TrFill of a SPAIR pulse are hereinafter taken as an example, to give examples of calculating the flip angle and the inversion time.

(51) Flip angle of FatSat pulse:

(52) For a spoiled gradient echo sequence provided with a FatSat pulse, the following parameters can be determined in advance according to the conditions of the magnetic resonance imaging hardware and user requirements and so on: Trep, , KSpaceCenterLine, N, Tr=TrMin, Tl1=Tl1Min.

(53) According to equation (15), the following can be deduced:
=arccos(X/E.sub.Tl1)

(54) where

(55) X=G1/(G1.Math.G4G2.Math.G3)

(56) G1=C(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1

(57) G2=D.sup.KSpaceCenterLine-1

(58) G3=E+F.Math.C.Math.(D.sup.N-11)/(D1)+F.Math.A.Math.D.sup.N-1

(59) G4=F.Math.D.sup.N-1

(60) Therefore, can be calculated easily.

(61) First inversion time TlFill and second inversion time TrFill of SPAIR pulse:

(62) For a spoiled gradient echo sequence provided with a SPAIR pulse, the following parameters can be determined in advance according to the conditions of the magnetic resonance imaging hardware and user requirements and so on: Trep, , =180, KspaceCenterLine, N, TrMin, Tl1Min. Also, as described above, Tl1=TlFill+Tl1Min, and Tr=TrFill+TrMin.

(63) The influence of TlFill and TrFill on the position where the longitudinal magnetization component of fat is zero is opposite, i.e., a longer TrFill will increase line indexes with a zero fat signal, while a longer TlFill will decrease line indexes with a zero fat signal. Also, for any parameter set, the result of the inversion time will be (TlFill>=0, TrFill=0) or (TlFill=0, TrFill>=0).

(64) Therefore, the first inversion time TlFill and the second inversion time TrFill of the SPAIR pulse can be calculated in the following steps:

(65) Step A: maintaining TrFill=0, i.e. Tr=TrMin, and calculating TlFill.

(66) According to equation (15), the following can be deduced:
TlFill=Tl1Tl1Min=T1.Math.ln(X/cos())Tl1Min

(67) where

(68) X=(G1+G2)/{G1.Math.G4+G2/cos()G2.Math.G3}

(69) G1=C.Math.(D.sup.KSpaceCenterLine-11)/(D1)

(70) G2=D.sup.KSpaceCenterLine-1

(71) G3=E+F.Math.C.Math.(D.sup.N-11)/(D1)

(72) G4=FD.sup.N-1

(73) Therefore, TlFill can be calculated easily. If TlFill>=0, it means that the current parameter set has calculated the inversion time, where TlFill>=0 and TrFill=0; and if TlFill<0, it means that when TrFill=0, the result set (TlFill, TrFill=0) from the calculation does not meet the requirements, and therefore step B needs to be performed.

(74) Step B: maintaining TlFill=0, i.e. Tl1=TlMin, and calculating TrFill.

(75) According to equation (15), the following can be deduced:
TrFill=TrTrMin=T1 ln(Y/cos())TrMin

(76) where

(77) Y=(G1+G2)/(G1.Math.G4G2.Math.G3)

(78) G1=C.Math.(D.sup.KSpaceCenterLine-11)/(D1)+A.Math.D.sup.KSpaceCenterLine-1

(79) G2=B.Math.D.sup.KSpaceCenterLine-1

(80) G3=C.Math.(D.sup.N-11)/(D1)+A.Math.D.sup.N-11/cos()

(81) G4=B.Math.D.sup.N-1

(82) Therefore, TrFill can be calculated easily. If TrFill>=0, it means that the current parameter set has calculated the inversion time, where TlFill=0 and TrFill>=0; and if TlFill<0, it means that for the current parameter set there is no corresponding inversion time, and the result will be TrFill=0, and TlFill=0.

(83) It should be noted that, although in the above example, TrFill=0 is maintained to calculate the corresponding TlFill in step A, and then in step B TlFill is maintained to calculate the corresponding TrFill, this is only an example. It is also permitted, in step A, to maintain TlFill=0 and calculate the corresponding TrFill, and then if necessary, in step B, maintain TrFill and calculate the corresponding TrFill.

(84) The foregoing has described in detail the general solution for calculating an optimized flip angle and inversion time. A method for magnetic resonance imaging according to this embodiment will be described below with reference to FIG. 3.

(85) As shown in FIG. 3, in step 301, a flip angle or inversion time of a spectrum suppression pulse is calculated according to a steady state condition of a longitudinal magnetization component of a fat composition suppressed by the fat suppression pulse and a zero crossing point condition of the longitudinal magnetization component. In this step, various parameters of the fat suppression pulse are determined in advance. The parameters include, for example, the number N of excitation pulses between two fat suppression pulses, the minimum time interval Tl1Min between the center of a fat suppression pulse and the center of the first excitation pulse following the fat suppression pulse, the time interval TrMin between the center of the last excitation pulse in a pulse train and the center of a fat suppression pulse in the next pulse train of an MRI sequence, time Trep between continuous excitation pulses, flip angle of an excitation pulse, index number KSpaceCenterLine of a k-space center line, and so on. And then, a flip angle or inversion time of the fat suppression pulse based on the various parameters is calculated according to a steady state condition of a longitudinal magnetization component of fat and a zero crossing point condition of the longitudinal magnetization component. In particular, a magnetization vector representation of fat is deduced according to the Bloch equation, and a general formula for calculating a flip angle or inversion time of a fat suppression pulse is deduced by applying a steady state condition of a longitudinal magnetization component of fat and a zero crossing point condition of the longitudinal magnetization component, and then the flip angle or the inversion time is calculated according to the various parameters determined. Applying a steady state condition M.sub.(m+1)=M.sub.(m) of a longitudinal magnetization component of fat and a zero crossing point condition M.sub.a(KSpaceCenterLine)=0 of the longitudinal magnetization component is a key to obtain an optimized flip angle and inversion time. And then, in step 302, magnetic resonance imaging is performed by applying an MRI sequence comprising the fat suppression pulse provided with the flip angle or the inversion time. An exemplary manner for applying fat suppression pulses in the MRI sequence is to follow each fat suppression pulse with one or more excitation pulses. The MRI sequence can be any sequence, as long as it can spoil the residual transverse magnetization components before applying the next excitation pulse. For example, the MRI sequence can be a spoiled fast spin echo (TSE) sequence, a spoiled gradient echo (GRE) sequence, etc.

(86) (Embodiment 2)

(87) The abovementioned embodiment 1 describes a method for optimizing a fat suppression pulse, and the optimization method can be applied to a water suppression pulse.

(88) In magnetic resonance imaging, in addition to fat suppression, water signals of a human tissue may need to be suppressed by applying, for example, a water saturation pulse and a FLAIR (fluid attenuated inversion recovery) pulse. The general formula (15) given in embodiment 1 of the present invention is applicable to both the water saturation pulse and the FLAIR pulse. More specifically, for the water saturation pulse, an optimized flip angle can be calculated using the same method as the solution for FatSat described in embodiment 1, so as to obtain an optimized water suppression pulse, and imaging can be performed by applying the magnetic resonance imaging method described with reference to FIG. 3 in embodiment 1; and similarly, for FLAIR, optimized Tlfill and Trfill values can be calculated using the same method as the solution for SPAIR described in embodiment 1, so as to obtain an optimized water suppression pulse, and the water suppression pulse can be optimized by applying the magnetic resonance imaging method described with reference to FIG. 3 in embodiment 1.

(89) In fact, in addition to the fat suppression pulse and the water suppression pulse, the optimization method described in embodiment 1 of the present invention can be applied to any particular spectrum suppression pulses, and all that needs to be done is to change parameters such as T1 or the center frequency of the suppression pulse and so on. That is to say, the method for magnetic resonance imaging described in embodiment 1 of the present invention can actually be applied to any spectrum suppression pulses.

(90) The method for magnetic resonance imaging according to the present invention has been described in detail above. In the method for magnetic resonance imaging, signals of a specific spectrum composition can be well suppressed by performing imaging using an MRI sequence provided with an optimized spectrum suppression pulse, so that high-quality images can be obtained.

(91) FIG. 4 shows a device 400 for magnetic resonance imaging using an MRI sequence provided with a spectrum suppression pulse. As shown in FIG. 4, the device for magnetic resonance imaging comprises: a calculation component 401 for calculating a flip angle or inversion time of a spectrum suppression pulse according to a steady state condition of a longitudinal magnetization component of a spectrum composition suppressed by the spectrum suppression pulse and an optimized spectrum suppression condition; and an imaging component 402 for performing magnetic resonance imaging by applying a magnetic resonance imaging sequence comprising the spectrum suppression pulse provided with the flip angle or the inversion time.

(92) As described above, the optimization method of a spectrum suppression pulse according to one aspect of the present invention has good universality and can be widely used. Also, the energy efficiency of spectrum signal suppression is improved by optimizing the spectrum suppression pulse. In addition, signals of a specific spectrum composition can be well suppressed by performing imaging using an MRI sequence provided with an optimized spectrum suppression pulse, so that high-quality images can be obtained. MATLAB simulation results and in vivo test results from the application of the abovementioned method of the present invention will be described below with reference to FIGS. 5 to 7.

(93) MATLAB Simulation Results

(94) FIGS. 5(a) and 5(b) show the simulation results when the spectrum suppression pulse is a FatSat pulse. The various parameters are set as follows: using a spoiled 3D gradient echo sequence for imaging, =9, the number of k-space filling lines corresponding to each pulse train=32, KSpaceCenterLine=#12, Trep=3.92 ms, TrMin=12.298 ms, Tl1min=10.482 ms, and T1=290.0 ms(3T). The FatSat flip angle calculated according to the abovementioned general formula (15) of the present invention is: =153.113, wherein the fat signal at line #12 is 0%.

(95) FIG. 5(a) is the simulation result of the fat signals with the abovementioned given parameters and the calculated flip angle. FIG. 5(b) is an enlarged schematic diagram of the curve within the rectangle box shown in FIG. 5(a).

(96) FIGS. 6(a) and 6(b) show the simulation results when the spectrum suppression pulse is a SPAIR pulse. The various parameters are set as follows: using a spoiled 3D gradient echo sequence for imaging. =9, =180, the number of k-space filling lines corresponding to each pulse train=32, KSpaceCenterLine=#12, Trep=3.92 ms, TrMin=14.918 ms, Tl1min=17.342 ms, and T1=290.0 ms(3T). The inversion time calculated according to the abovementioned general formula (15) of the present invention is: TrFill=0, TlFill=0.91 ms, wherein the fat signal at line #12 is 0%.

(97) FIG. 6(a) is the simulation result of the fat signals with the abovementioned given parameters and the calculated TrFill and TlFill. FIG. 6(b) is an enlarged schematic diagram of the curve within the rectangle box shown in FIG. 6(a).

(98) In Vivo Test Results

(99) FIGS. 7(a) and 7(b) show the images formed using an MRI sequence provided with an optimized spectrum suppression pulse according to the present invention. In particular, FIG. 7(a) shows an image captured when the spectrum suppression pulse is a FatSat fat suppression pulse, the number of k-space filling lines corresponding to each FatSat pulse train is 32 lines, and the calculated FatSat flip angle=153; and FIG. 7(b) shows an image captured when the spectrum suppression pulse is a SPAIR fat suppression pulse, the number of k-space filling lines corresponding to each SPAIR pulse train is 32 lines, and it is calculated that TrFill=0 and TlFill=62.3 ms. FIG. 7(c) shows an image captured using an MRI sequence not provided with a spectrum suppression pulse. The images shown in FIGS. 7(a)-7(c) are all obtained using a short T1 spoiled 3D gradient echo sequence, and all require volunteers to hold breath in tests to eliminate motion artifacts.

(100) Arrows below FIGS. 7(a)-7(c) all indicate the subcutaneous area. As shown in FIG. 7(c), when fat suppression is not performed, the subcutaneous area displays fat signals in high luminance. However, the corresponding areas in FIGS. 7(a) and 7(b) are displayed as low dark signals. Therefore, the spectrum suppression pulse optimization method according to the present invention can work well for FatSat and SPAIR techniques.

(101) Although modifications and changes may be suggested by those skilled in the art, it is the intention of the inventor to embody within the patent warranted hereon all changes and modifications as reasonably and properly come within the scope of the inventor's contribution to the art.