CN110780249B - Magnetic resonance imaging method for measuring RF B1 field distribution using adiabatic RF pulses - Google Patents

Abstract

The invention discloses a magnetic resonance imaging method for measuring radio frequency B1 field distribution by using adiabatic radio frequency pulse, which comprises the following steps of designing adiabatic pulse, setting frequency offset and adiabatic pulse time, calculating K value of the adiabatic pulse, obtaining initial magnetization vector size without applying adiabatic pulse sampling, and respectively applying frequency offset of △ omega_rfAnd- △ omega_rfAnd the magnitude of the magnetization vector M obtained by sampling_z1And M_z2(ii) a Calculation of B_1,obsAnd calculating the deviation of B1 measurement caused by B0 field deviation, and applying large △ omega to facilitate calculation and analysis_rfThe values are good enough to avoid the effect of B0 field offset on the B1 measurement.

Description

Magnetic resonance imaging method for measuring RF B1 field distribution using adiabatic RF pulses

technical field

The invention belongs to the technical field of nuclear magnetic resonance imaging, in particular to a magnetic resonance imaging method for measuring radio frequency B1 field distribution using adiabatic radio frequency pulses. The invention is suitable for measuring the distribution of radio frequency B1 field in different tissues or samples when the surface coil, body coil or phased array coil transmits in magnetic resonance imaging, and is used for image reconstruction and B1 field correction.

Background technique

Magnetic resonance imaging is an important clinical method for diagnosing and evaluating disease progression. It has the advantages of no ionizing radiation, non-invasiveness, high spatial resolution, imaging at any level, and high tissue contrast. It is invaluable in the diagnosis of tumors, organs and soft tissue lesions. substitute role.

The B1 field represents the intensity of the radio frequency pulse when the magnetic resonance coil is fired. In magnetic resonance imaging, due to the irregular structure of the tissue and the non-uniform emission field of the coil, the radio frequency B1 field is unevenly distributed in space. As the magnetic field strength of the current magnetic resonance spectrometer is getting higher and higher, the problem of radio frequency B1 field inhomogeneity is becoming more and more serious. In the current methods of parallel imaging and chemical exchange saturation transfer (CEST) imaging commonly used in magnetic resonance, the radio frequency B1 field distribution is also required for image reconstruction and correction. The development of a magnetic resonance method to measure the radio frequency B1 field distribution can solve the above problems.

The B0 field represents the magnitude of the magnetic field induced at a certain point in space during magnetic resonance imaging, and is closely related to the nuclear spin frequency. The B0 field offset frequency is used to represent the difference between the RF frequency and the nuclear spin frequency during RF excitation. The B0 field offset will affect the measurement of the B1 field distribution, mainly when the B0 field offset is large, and the magnetic resonance signal is related to both B0 and B1 under the off-resonance effect. Therefore, the influence of the B0 field offset cannot be ignored when measuring the B1 field distribution.

The existing magnetic resonance methods for measuring the B1 field distribution are mainly based on the double flip angle method based on the signal amplitude, for example, Insko EK et al. -85.] Proposed B1 field measurement method based on different flip angles, and improvements based on this method, for example, Cunningham CH et al. [Sturated double-angle method for rapid B1+mapping.Magn.Reson.Med.2006.55:p .1326-1333.] proposed a faster B1 field measurement method based on different flip angles. However, the excitation waveforms of the pulses at different flip angles are not linear, and the method is sensitive to B0 field shifts.

AFI (Actualflip-angle imaging in the pulsed steady state:amethod for rapid three-dimensional mapping of the transmitted radio frequencyfield.Magn.Reson.Med.2007.57:p.192-200.] by Yarnykh VL et al. ) method, using the ratio of the signal intensities of two different recovery times TR1 and TR2 under the same excitation angle to calculate the B1 field distribution, this method is fast, but is sensitive to the B0 field shift and motion.

The method based on signal phase mainly includes the measurement method proposed by Sacolick LI et al. proportional. The phase reconstruction in the phase method is more complicated, and it may be accompanied by the problem of phase wrapping (the phase exceeds 360°).

In addition, other methods include the use of off-resonance effects simultaneously proposed by Patrick Schuenke et al. Method for fitting B0 and B1 field distributions. This method is applied to chemical exchange saturation transfer imaging, which requires the acquisition of 10 or more images for curve fitting.

The present invention proposes a magnetic resonance imaging method using adiabatic radio frequency pulses to measure the radio frequency B1 field distribution, which is a B1 measurement method that is insensitive to the B0 field offset and can be widely used in different pulse sequences, and is the measurement of the radio frequency B1 field. And provide technical support for image reconstruction and correction.

SUMMARY OF THE INVENTION

In order to solve the above problems existing in the prior art, the present invention provides a magnetic resonance imaging method for measuring the field distribution of radio frequency B1 using adiabatic radio frequency pulses. The present invention is based on the following ideas:

In the rotating coordinate system, the magnetization vector can be considered to precess along the effective B1 field, the _Beff field. Among them, B _eff =[(γB1) ² +(Δω) ² ] ^1/2 , γ represents the gyromagnetic ratio of the nucleus, Δω represents the magnitude of the B0 field shift, B _eff represents the intensity of the B _eff field, and B1 represents the The strength of the B1 field. Precession frequency ω=γB _eff . The included angle θ=arctan(γB1/|Δω|) between the B _eff field and the Z axis (ie, the direction of the main magnetic field).

Adiabatic pulses are radio frequency pulses under special conditions. Under adiabatic conditions, the magnetization vector parallel to the initial _Beff field will be spin-locked by the adiabatic pulse and always stay in line with the _Beff field.

The adiabatic pulse needs to meet the adiabatic conditions. By defining K=γ|B _eff |/|dθ/dt|, K>>1 is required to ensure adiabaticity, that is, the precession frequency around the B _eff field is much larger than that of the B _eff field and the Z axis. change in angle. To satisfy the adiabatic condition, a large frequency offset is set for the RF pulse, defined as Δω _rf , to ensure that Δω _rf > Δω. The addition of the bias value not only increases the B _eff value, but also reduces the variation of θ with the B1 field, ensuring the adiabatic properties of the RF pulse.

Select a radio frequency pulse with an initial intensity of 0 and a maximum intensity at the end, and set the maximum intensity of the radio frequency _pulse to be B _{1max .} △ω, the angle between the magnetization vector and the Z axis at the beginning of the RF pulse is 0, and the angle between the magnetization vector and the Z axis at the end of the RF pulse is θ ₁ =arctan[γB _1max /|(△ω+△ω _rf )|] .

Suppose the initial magnetization vector is M ₀ , after the adiabatic pulse ends, the angle between the magnetization vector and the Z-axis direction is θ ₁ =arctan[γB _1max /|(△ω+△ω _rf )|], applying the damage gradient field to damage the transverse direction For the magnetization vector, only the projection M _z1 of the initial magnetization vector M ₀ on the Z-axis direction is retained, then M _z1 =M ₀ cos(θ ₁ ). In the same way, change the frequency offset to -Δω _rf , then the angle between the magnetization vector and the Z axis θ ₂ =arctan[γB _1max /|(Δω-Δω _rf )|], M _z2 =M ₀ cos(θ ₂ ).

Let B _{1, obs} be the calculated value of the B1 field, and its calculation is publicized as follows:

In formula 2, the term Δω ² /(Δω _rf ² -Δω ² ) is the deviation between the observed B _1,obs and the actual B _1max . When Δω _rf is selected to be 4000 Hz, the deviation of Δω in the range of -400 to 400 Hz is less than or equal to 1.01%. In order to ensure the adiabatic properties of the pulse, and the measurement results are not sensitive to Δω, the larger the value of Δω _rf , the better, and 2000Hz or above can meet the experimental conditions.

A magnetic resonance imaging method for measuring RF B1 field distribution using adiabatic RF pulses, comprising the following steps:

Step 1, design an adiabatic pulse, the initial intensity of the adiabatic pulse is zero, and the adiabatic pulse is gradually increased to the maximum intensity B _1max . Set frequency offset △ω _rf and adiabatic pulse time;

Step 2: Calculate the K value of the adiabatic pulse to ensure that the K value satisfies the adiabatic condition. The calculation of the K value is given by the formula K=γ|B _eff |/|dθ/dt|, and the adiabatic condition is satisfied in the case of K>100. The larger the maximum intensity B _1max , the smaller the K value; the larger the △ω _rf , the larger the K value; the pulse time should be as short as possible under the condition of satisfying the K value to reduce the influence of relaxation, among which, B _eff characterizes B The strength of the _eff field, γ is the gyromagnetic ratio of the nucleus, θ is the angle between the B _eff field and the Z-axis (ie, the direction of the main magnetic field), θ=arctan(γB1/|△ω|), B1 represents the strength of the B1 field, △ω is the magnitude of the B0 field offset;

Step 3, no adiabatic pulse is applied or the maximum intensity B _1max of the adiabatic pulse is set to zero, and the initial magnetization vector size M ₀ is obtained by sampling;

Step 4, applying an adiabatic pulse with a frequency offset of Δω _rf and a damage gradient, and sampling to obtain the magnitude of the magnetization vector M _z1 ;

Step 5, applying an adiabatic pulse with a frequency offset of _{-Δω rf} and a damage gradient, and sampling to obtain the magnitude of the magnetization vector M _z2 ;

Step 6, use the following formula to calculate B _1,obs ,

Among them, M ₀ is the initial magnetization vector.

Divide B _1,obs by the maximum intensity B _1max for normalization to obtain the normalized radio frequency B1 field distribution;

Step 7, according to the selected frequency offset Δω _rf , according to Δω ² /(Δω _rf ² -Δω ² ) to calculate the deviation caused by the B0 field offset to the B1 measurement, Δω _rf >Δω .

Compared with the prior art, the present invention has the following beneficial effects:

1. In this method, the adiabatic pulse intensity and the signal intensity show a simple trigonometric function relationship under adiabatic conditions, which is convenient for calculation and analysis.

2. The calculation deviation of this method due to the B0 field offset △ω is △ω ² /(△ω _rf ² -△ω ² ). By applying a large △ω _rf value, the effect of the B0 field offset on B1 measure the impact.

3. In this method, the adiabatic pulse and the damage gradient are applied before sampling and can be integrated into a sub-module, which is not limited by the sampling method.

Description of drawings

Fig. 1 is the flow chart of the inventive method;

Figure 2(a) shows a simple linearly increasing pulse shape with a B _1max of 40 μT and a duration of 2 ms,

Figure 2(b) shows the variation of K value with pulse time when the bias Δω _rf of the pulse shown in Figure 2(a) is 4000Hz. The K value is much greater than 1 in the whole process, which satisfies the adiabatic condition. The pulses used in the present invention are not limited to this pulse shape;

3 is a pulse sequence diagram of using adiabatic pulses to measure the B1 field distribution magnetic resonance imaging method; this figure only introduces the pulse sequence diagram of using spin echo for data acquisition, and the data acquisition method used in the present invention is not limited;

Fig. 4 is the evolution diagram of the magnetization vector with the pulse intensity under adiabatic conditions, the initial magnetization vector M ₀ is flipped to the direction of B _eff by the pulse, at this time, it is projected on the Z-axis direction as M _z1,2 ;

FIG. 5 is a normalized B1 field distribution diagram measured by the method of the present invention for a 3% agarose solution sample;

Figure 6 shows the measurement error caused by the offset Δω of the B0 field when the magnetic resonance imaging method using adiabatic pulses is used to measure the B1 field distribution.

Detailed ways

In order to facilitate the understanding and implementation of the present invention by those of ordinary skill in the art, the present invention will be further described in detail below with reference to the embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, but not to limit the present invention.

Example:

The sample used in this example is a 3% agarose solution, which is placed in an NMR sample tube with an outer diameter of 10 mm. Using a Bruker 400M wide-chamber small animal imager with a saddle coil with an inner diameter of 10 mm, the temperature was controlled at 300 K. Under this condition, the T1 relaxation time of the sample is about 3400ms, the T2 relaxation time is about 40ms, and the TR time of the pulse sequence is set to 5s. The test object of this method can be a sample, an animal or a human. The radio frequency coil can be a surface coil, a body coil or a phased array coil or the like.

A magnetic resonance imaging method for measuring RF B1 field distribution using adiabatic RF pulses, comprising the following steps:

Step 1. As shown in the flowchart of Fig. 1, the magnetic resonance imaging method using adiabatic pulse to measure B1 field distribution must first determine the waveform of the adiabatic pulse so that the initial intensity is 0, and the maximum intensity B _1max is reached before the end of the adiabatic pulse. In this embodiment, a simple linearly varying pulse shape is used, as shown in FIG. 4 . Set the maximum intensity B _1max to 40 μT, set the frequency offset Δω _rf to 4000 Hz, and set the adiabatic pulse time to 2 ms. The pulse shape of the adiabatic pulse used in this method is not limited to a linear pulse, but can also be a pulse of various shapes satisfying an increment from 0 to B _1max .

Step 2. Calculate the adiabatic degree, that is, the K value. First, the adiabatic pulse in this embodiment consists of 1000 dots. Calculate the B _eff value corresponding to each point separately from the formula B _eff =[(γB ₁ ) ² +(Δω _rf ) ² ] ^1/2 , and then use the formula θ=arctan[γB ₁ /|(Δω _rf )| ] Calculate the angle θ corresponding to each time point separately. The ratio of the difference between two adjacent points θ and the time difference between adjacent points is dθ/dt. Finally, the K value can be calculated using the formula K=γ|B _eff |/|dθ/dt|. The time distribution of the K value of the pulse used in this embodiment is shown in Fig. 2(b), which satisfies the adiabatic condition of K>>1.

The pulse sequence used in this embodiment is shown in FIG. 3 . After the adiabatic pulse in step 2 is applied, the initial magnetization vector will be flipped to the final B _eff direction, as shown in FIG. 4 . At this time, 1ms of damage gradients in the three directions of X/Y/Z are immediately applied to destroy the transverse magnetization vector, and then the data acquisition is started. The data acquisition method used in this embodiment is the fast spin echo method. The present invention is not limited to the sampling method. After applying the adiabatic pulse and destroying the gradient, it is also feasible to use the gradient echo or plane echo and other sampling methods.

Step 3, no adiabatic pulse is applied or the maximum intensity B _1max of the adiabatic pulse is set to zero, and the initial magnetization vector size M ₀ is obtained by sampling;

Step 4, applying an adiabatic pulse with a frequency offset of Δω _rf and a damage gradient, and sampling to obtain the magnitude of the magnetization vector M _z1 ;

Step 5, applying an adiabatic pulse with a frequency offset of _{-Δω rf} and a damage gradient, and sampling to obtain the magnitude of the magnetization vector M _z2 ;

Step 6, use the following formula to calculate B _1,obs ,

Among them, M ₀ is the initial magnetization vector.

Divide B _1,obs by the maximum intensity B _1max for normalization to obtain the normalized radio frequency B1 field distribution;

First acquire the first image when no adiabatic pulse is applied (or the maximum intensity of the pulse is 0), and then acquire two images of the adiabatic pulse with RF bias Δω _rf and -Δω _rf applied, respectively, the second image and third image.

Background segmentation is performed on the first image, the second image, and the third image, respectively, so that the image backgrounds of the first image, the second image, and the third image are zero. Then, point-to-point calculation is performed on the first image, the second image, and the third image. Take any point on the first image, set its signal strength as M ₀ , and take the second image and the third image with the RF bias Δω _rf and -Δω _rf applied respectively at the same position as the point on the first image. point, and the signal strengths are set as M _Z1 and M _Z2 , respectively. Then, the measured B _1,obs can be calculated by formula (1).

Through step 6, the B1 _,obs corresponding to each point on the image are calculated, and the B1 distribution map is obtained. Then divide each point of the B1 distribution map by the B _1max value set by the system to obtain the normalized B1 distribution map. The normalized B1 distribution calculated in this embodiment is shown in FIG. 5 , and the sample tube has a local B1 field inhomogeneity in the area close to the coil.

Step 7, according to the selected frequency offset Δω _rf , according to Δω ² /(Δω _rf ² -Δω ² ) to calculate the deviation caused by the B0 field offset to the B1 measurement, Δω _rf >Δω .

The calculation error caused by the B0 field offset Δω is Δω ² /(Δω _rf ² -Δω ² ). Figure 6 shows the relationship between the calculation error and Δω under different Δω _rf . The Δω _rf used in this embodiment is 4000 Hz. When Δω is 400 Hz, the calculation error is about 1%, and the calculation result still has good accuracy under a certain B0 field offset.

It should be pointed out that the specific embodiments described in the present invention are only for illustrating the spirit of the present invention. Those skilled in the art to which the present invention pertains can make various modifications or additions to the described specific embodiments or substitute in similar manners, but will not deviate from the spirit of the present invention or go beyond the definition of the appended claims range.

Claims (2)

1. A magnetic resonance imaging method using adiabatic radio frequency pulses to measure a radio frequency B1 field distribution, comprising the steps of:

step 1, designing adiabatic pulse, wherein the initial intensity of the adiabatic pulse is zero, and the adiabatic pulse is gradually increased to the maximum intensity B_1maxSetting the frequency offset △ omega_rfAnd adiabatic pulse time;

step 2, calculating the K value of the adiabatic pulse, wherein K is gamma | B_eff|/|dθ/dt|，B_effCharacterization B_effThe intensity of the field, gamma being the gyromagnetic ratio of the nucleus, theta being B_effAngle of field to main magnetic field direction, B_effThe field is the effective B1 field;

step 3, maximum intensity B of adiabatic pulse is not applied or is applied_1maxSet to zero, the sample obtains the initial magnetization vector magnitude M₀；

Step 4, applying frequency offset of △ omega_rfAnd the magnitude of the magnetization vector M obtained by sampling_z1；

Step 5, applying frequency bias to- △ omega_rfAnd the magnitude of the magnetization vector M obtained by sampling_z2；

Step 6, calculate B using the following formula_1,obs，

Wherein M is₀In order to be the vector of the initial magnetization,

step 7, according to △ omega²/(△ω_rf ²-△ω²) To calculate the deviation of B1 measurement caused by B0 field deviation, △ omega is the size of B0 field deviation, B_1,obsIs the calculated value of the B1 field.

2. The use of adiabatic radio frequency pulsing as set forth in claim 1A magnetic resonance imaging method for measuring the distribution of radio frequency B1 field is characterized in that the frequency is offset △ omega_rfGreater than the magnitude △ ω of the B0 field offset.

Images