JPH038215B2 - - Google Patents

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention relates to nuclear magnetic resonance (nuclear magnetic resonance).
The present invention relates to an improvement in the method of measuring NMR signals in an NMR tomographic imaging device that utilizes resonance (hereinafter abbreviated as NMR) phenomenon to determine the distribution of specific atomic nuclei within a subject from outside the subject.

(Prior Art) There are two methods of observing NMR signals in an NMR tomography apparatus: a method of observing an FID signal (free induction decay signal) and a method of observing an echo signal. Figure 5 is a diagram showing the pulse sequence when observing the FID signal, and Figure 6 is a diagram showing the pulse sequence when observing the echo signal. In Figure 6, the illustration of the gradient magnetic field application is omitted. It has been done.

In the case of FID signal observation, apply a Z gradient magnetic field as shown in Figure 5B, and apply a 90° pulse as shown in Figure 5A to excite spins in a predetermined slice (selective excitation). . Next, the Y gradient magnetic field Gy
(the size changes for each view) and performs phase encoding. Subsequently, an X-gradient magnetic field for a readout gradient is applied, and this time point is set as the time origin (t=0) of the FID signal, and the FID signal is observed over T time.

On the other hand, in the case of echo signal observation, the process from the application of the 90° pulse to the point where the NMR signal is generated is the same as in Figure 5, but the 180° pulse is applied when τ time has elapsed after the application of the 90° pulse. , flips the spin during dispersion by 180°. As a result, the spins begin to focus, generating an echo signal that reaches its maximum at the time τ after the application of the 180° pulse. Observation of the echo signal generated in this way is carried out for a period of T _n from immediately after the application of the 180° pulse until the time when the echo signal substantially disappears, as shown in the figure.

(Problems to be Solved by the Invention) Incidentally, since FID signals have several points that are difficult to use compared to echo signals, conventionally, observation of echo signals has been mainly used. However, when compared in terms of S/N per unit time, it can be seen that the FID signal is superior to the echo signal in that there is no signal drop due to T ₂ relaxation and no wasted time due to 180° pulses.

For these reasons, it is preferable to use the FID signal, but there are other problems with using the FI signal.

When performing two-dimensional Fourier transformation, it is necessary to accurately capture the peak value of the FID signal and its position (this position becomes the time origin when observing the FID signal) and process the data. When using an echo signal, there is no problem as the DC component in the two-dimensional Fourier transform of the NMR image can be accurately obtained as the peak of the echo signal (point P in the figure), as shown in Figure 6. When using an FID signal, although the DC component corresponds to the first data, it is difficult to find the exact position of the peak in this part, and the waveform is disturbed by the rise of the observation slope, making it difficult to determine the exact amplitude. There was also the problem that it was difficult to obtain.

In view of these points, an object of the present invention is to be able to know the exact position of the time origin of the FID signal,
An object of the present invention is to provide an FID signal processing FID signal measurement method that can obtain an accurate reconstructed image based on detection data from an FID signal.

(Means for solving the problem) In order to achieve such an objective, the present invention
In a nuclear magnetic resonance signal measurement method in which a gradient magnetic field for a readout gradient is applied after selective excitation and the generated FID signal is observed, prior to applying a gradient magnetic field for a readout gradient, a magnetic field of opposite sign to that of the gradient magnetic field is applied. Apply the gradient for an appropriate time, generate the FID signal a little before the origin at that time, and then
The system detects the signal, estimates the accurate position of the time origin from multiple sampling points near the time origin, and calculates a reconstructed tomographic image based on the detected data of the FID signal after the time origin and the estimated position of the time origin. It is characterized by the following.

(Example) The present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram of the main parts of an NMR tomography apparatus for carrying out the method of the present invention. In the figure, 1 is a magnet assembly, inside which a space (hole) is provided for inserting an object, and a main magnetic field coil that surrounds this space and applies a constant magnetic field to the object. and a gradient magnetic field coil for generating a gradient magnetic field (x configured to be able to individually generate a gradient magnetic field)
gradient magnetic field coils, y gradient magnetic field coils, and z gradient magnetic field coils), RF transmitting coils that provide high-frequency pulses to excite the spins of atomic nuclei within the object, and receiving coils that detect NMR signals from the object, etc. is located.

The main magnetic field coil is connected to the main magnetic field power supply 2, Gx, Gy, Gz
Each gradient magnetic field coil is connected to a Gx, Gy, Gz gradient magnetic field drive circuit 3, and the RF transmitting coil is connected to an RF power amplifier 4.
The coils for receiving NMR signals are each connected to a preamplifier 5. 10 is a sequence storage circuit that controls the generation sequence of gradient magnetic fields and high-frequency magnetic fields, and also controls the timing when A/D converting the obtained NMR signal.

6 is a gate modulation circuit, and 7 is an RF oscillation circuit that generates a high frequency signal. The gate modulation circuit 6 modulates the high frequency signal output by the RF oscillation circuit 7 using the timing signal from the sequence storage circuit 10,
Generates high frequency pulses. This high frequency pulse
RF power amplifier 4.

8 is a phase detector which refers to the output signal of the RF oscillation circuit 7, detects it with a receiving coil, and sends it to the preamplifier 5.
Detects the phase of the NMR signal sent via the

11 is an A/D converter, which converts the NMR signal (analog) obtained through the phase detector 8 into a digital signal. 13 is a processing device including a computer, which sequences scan conditions to realize various scans. It has a function of supplying data to the memory circuit 10, a calculation processing function of reconstructing information on spin density distribution, etc. into an image from observation data input from an A/D converter, a function of sending and receiving information to and from the operation console 12, etc. .

The reconstructed image obtained by the processing device 13 is displayed on the display device 9.
Displayed in .

The operation in such a configuration will be explained next.

In the conventional sequence shown in FIG. 5, the moment when the readout gradient (X gradient magnetic field) is applied corresponds to the time origin. However, in reality, it takes time for the readout gradient to rise, and the spin phase changes during this time, so by the time the gradient has completely risen (and therefore the signal can be observed), the time origin has already been reached. It has drifted away from it. Further, immediately after the gradient rises, the received signal may be disturbed due to the coupling state of the gradient magnetic field coil and the receiving coil, and the accuracy of the data is not guaranteed.

In the present invention, as shown in Fig. 2, accurate data near the time origin is obtained by scanning in a sequence in which the readout gradient (X gradient magnetic field) is applied differently from the conventional sequence, and as shown in Fig. 3. The accurate time origin is estimated and the image is reconstructed based on the FID signal and the time origin. That is, prior to FID observation, a readout gradient is applied slightly in the opposite direction. This sequence allows the FID signal to be observed from a little before the time origin, and as shown in Figure 3, the readout gradient is applied in the opposite direction. By calculating from multiple sampling points, it becomes possible to determine the time origin as the peak of the signal. Furthermore, if ΔT is set to an appropriate length, signal disturbances due to the rise of the gradient, etc. will be contained within this time, and a signal with accurate amplitude can be observed. ΔT may be several hundred microseconds to several milliseconds, depending on the device, and the increase in signal observation time caused by this can be ignored.

Such a sequence is stored in the sequence memory circuit 1.
0, and by controlling the gradient magnetic field drive circuit 3 by the sequence storage circuit 10, application of the gradient magnetic field as described above is achieved.

Now, the time origin is the position where the signal peaks, but generally that point does not always coincide with the sampling point. Therefore, as shown in Fig. 3, several points S ₁ , S ₂ , and S ₃ near the peak are approximated by an appropriate function (for example, a quadratic function) Fn to find the true peak Pt.
It is desirable to seek. Furthermore, in the Fourier transform method, since the time origin can be detected only in a view with a warp amount of zero, values determined in a view with a warp amount of zero are used in other views.

Accurate FID data obtained in this way is
Since the data is half of the frequency space, it is possible to use the conjugate symmetry of the Fourier transform of a real function to create the remaining half of the data and reconstruct the image, or to perform the Fourier transform with the remaining half of the data as zero and then calculate its real part. The image can be created by an appropriate method such as reconstructing the image using . However, whichever method is used, it is important to obtain an accurate position of the time origin in order to suppress phase distortion in image reconstruction processing that uses data in the positive time domain, that is, half of the frequency space. There are various imaging methods, and one method for obtaining an image with fewer artifacts using Fourier reconstruction is, for example, the following method.

By embedding a zero value in the negative time domain of the FID signal, or by assuming that there is a zero value, one is left as is, and the readout time axis is reversed depending on the direction of the magnetic field gradient when measuring the FID signal. The complex image is Fourier-transformed as shown in FIG. By adding complex numbers to the two images obtained in this way, the influence of the main magnetic field inhomogeneity is reduced,
A real value image is obtained by image processing such as absolute value processing or phase rotation.

Note that the method of the present invention is not limited to the Fourier method.
It can be similarly used for projection method (PR method).
An example of the pulse sequence at that time is shown in FIG.
In the PR method, scanning is performed by changing the direction of the projection gradient for each view (by changing the X gradient magnetic field and the Y gradient magnetic field). It is sufficient to apply a gradient with the same angle and opposite sign.
That is, for example, when applying G _pr1 as the Y gradient magnetic field, -kG _pr1 is applied before signal observation, and when applying G _pr2 as the X gradient magnetic field, -kG _pr2 is applied before signal observation. However, k is a positive constant, and G _pr1 and G _pr2 are the magnetic field strengths that change for each view.

(Effects of the Invention) As explained above, according to the present invention, an accurate position near the time origin of the FID signal can be obtained. Therefore, it is possible to obtain an image with a good S/N ratio per unit time by using the FID signal, which is data in the positive time domain, that is, data in half of the frequency space, and to obtain a highly accurate image without phase distortion. Can be done.

[Brief explanation of the drawing]

Figure 1 shows NMR for carrying out the method of the present invention.
FIG. 2 is a diagram showing a pulse sequence for carrying out the method of the present invention, and FIG. 3 is a diagram for explaining how the peak value of an FID signal is approximated by a curve. Figure 4 shows the pulse sequence when the method of the present invention is applied to the PR method, Figure 5 shows the conventional pulse sequence for FID signal observation, and Figure 6 shows the pulse sequence for echo signal observation. FIG. 3 is a diagram showing such a conventional pulse sequence. DESCRIPTION OF SYMBOLS 1... Magnet assembly, 2... Main magnetic field power supply, 3... Gradient magnetic field drive circuit, 4... RF power amplifier, 5... Preamplifier, 6... Gate modulation circuit, 7... RF oscillation circuit, 8... Phase detector, 9
... Display device, 10 ... Sequence memory circuit, 1
1... A/D converter, 12... Operation console,
13... Processing device.

Claims (1)

[Claims] 1. After selectively exciting a desired slice plane of the object, a gradient magnetic field for a readout gradient is applied, and the FID signal is measured and processed to obtain a reconstructed tomographic image of the object. In the resonance signal processing method, prior to applying the gradient magnetic field for the readout gradient, a gradient with the opposite sign to the gradient magnetic field is applied for an appropriate time, and the FID signal is generated slightly before the time origin. Detects the FID signal, estimates the accurate position of the time origin from multiple sampling points near the time origin, and reconstructs a tomographic image based on the FID detection data after the time origin and the estimated position of the time origin. A method of processing nuclear magnetic resonance signals as required.