CN1588113A - Receiving method of nuclear magnetic resonance imaging signal - Google Patents

Abstract

The present invention relates to nuclear magnetic resonance imaging method and nuclear magnetic resonance imaging apparatus class, relate to a kind of nuclear magnetic resonance imaging signal receiving method more specifically, this receiving method is to be provided with the counter that is used for measuring each error time on the nuclear magnetic resonance digital receiver , the counter sends the measured error time value to the host, and performs linear phase correction on the spectral lines of the NMR signal within the receiver spectral width. The phase jitter of the nuclear magnetic resonance signal caused by the receiver is very simple and easy to implement. At the same time, due to the use of fast Fourier transform and inverse Fourier transform, the calculation burden of the host is greatly reduced.

Description

A method for receiving nuclear magnetic resonance imaging signals

technical field

The invention relates to a nuclear magnetic resonance imaging method and nuclear magnetic resonance imaging instruments, and more specifically relates to a method for receiving nuclear magnetic resonance imaging signals.

Background technique

With the development of magnetic resonance digital quadrature detection method, the application of digital receiver in imaging system is more and more extensive. Fig. 7 is a schematic diagram of a typical nuclear magnetic resonance digital receiver system. The nuclear magnetic resonance signal collected by the probe (1) is amplified by the low-noise preamplifier (2), enters the mixer (3) for detection, and is amplified by the intermediate frequency signal conditioner (4) to filter out high-frequency components Enter the analog-to-digital converter ADC (5) for digital sampling. According to the Nyquist sampling theorem, the sampling frequency f _s of the ADC is usually at least twice higher than the resonance frequency f _fid of the magnetic resonance signal in order to achieve distortion-free acquisition of the magnetic resonance signal. If in order to further reduce the quantization noise of the ADC, it is necessary to realize the oversampling of the nuclear magnetic resonance signal, and the sampling frequency of the ADC is often ten to dozens of times higher than the intermediate frequency. For now, the high-speed digital signal output from the ADC is difficult to realize the real-time processing of the digital signal by the microprocessor. Some people have used multiple high-performance microprocessors to work in parallel to solve this problem. Obviously, this will greatly increase the complexity and cost of the system. A more convenient approach is to use an application-specific integrated circuit (ASIC) to achieve. As shown in Figure 7, the data sampled by the ADC is sent to the digital down-converter (6), where (7A) and (7B) are digital quadrature mixers with a fixed phase difference of 90 degrees, through which the magnetic resonance signal is down-converted by an intermediate frequency Near zero frequency, and then enter the cascaded comb filter (CIC) (8A), (8B) to perform the first-stage extraction and filtering of the signal to obtain a slower rate stream, programmable finite impulse response extraction filter (DEC1, DEC2) (9A), (9B) then carry out several stages of filtering and extraction to the magnetic resonance signal, and finally store the complex signal (I, Q) that obtains the nuclear magnetic resonance baseband into the host computer (10).

Using a digital receiver can get a magnetic resonance spectrogram without zero-frequency peaks and image peak interference. In addition, due to the use of digital filters in the digital receiver, it also has very good linear phase shift in the passband and out-of-band noise Inhibition. It can be seen from Figure 7 that the digital signal output by high-speed ADC sampling (usually with a bandwidth of tens of megahertz) is processed by a digital receiver, and the output is often a slow narrow-band digital signal (with a bandwidth of usually tens of kilohertz). Fig. 2 is a timing diagram of the output signal of the digital receiver, wherein, A is the nuclear magnetic resonance complex digital signal output by the digital receiver. A' is the data frame signal of the digital receiver, and each signal represents a data sampling point. B is a sampling start signal, which is controlled by the host and starts sampling on its rising edge. C is the actual NMR simulation signal. D represents the N-1 sampling point of the digital receiver and its position on the actual nuclear magnetic resonance signal. D' represents the Nth sampling point of the digital receiver and its position on the actual nuclear magnetic resonance signal. The position corresponding to the sampling start time on the actual nuclear magnetic resonance signal is E. It can be known from Fig. 2 that there will be a t _p time error between E and the output sampling point D of the digital receiving receiver. and have the following relationship:

0≤t _p <t _s

t _s =1/SW

Among them, t _s is the time interval of each sampling; SW is the receiver sampling spectrum width.

This time error is manifested as a random jitter of the phase of the nuclear magnetic resonance signal in a single sampling of the nuclear magnetic resonance signal, thereby affecting the accumulation or cancellation of the nuclear magnetic resonance signal. On magnetic resonance imaging, it is manifested as artifacts in the direction of phase encoding.

In order to reduce the random jitter of this phase, a direct way is to increase the bandwidth of the receiver, so that SW increases and the t _p time error is reduced. However, this will cause the host to decimate and filter the signal output by the receiver again to obtain a suitable field of view (FOV). In addition, with the increase of the output spectral width of the receiver, it is necessary to use a faster computer data bus to transmit data, and use a faster Large memory space to store data will place high demands on the software and hardware of the system.

Contents of the invention

The purpose of the present invention is to provide a method for receiving nuclear magnetic resonance imaging signals according to the shortcomings of the above-mentioned prior art. The method determines the deviation value of the phase by adding a counter, and then corrects the signal to eliminate the phase jitter.

The object of the present invention is realized by the following technical solutions:

A method for receiving a nuclear magnetic resonance imaging signal is characterized in that a counter for measuring each error time is arranged on the nuclear magnetic resonance digital receiver, and in each sampling, after the counter sends the measured error time value to the host computer, and then The linear phase correction is performed on the spectral line of the nuclear magnetic resonance signal within the spectral width of the receiver, thereby eliminating the phase jitter of the nuclear magnetic resonance signal.

Before performing linear phase correction, it is necessary to perform fast Fourier transform (FFT) on the acquired nuclear magnetic resonance signal to transform the nuclear magnetic resonance signal from the time domain to the frequency domain.

The linear phase correction of the spectral lines within the spectral width of the receiver refers to multiplying the data point corresponding to each frequency point in the spectral width range by the phase correction factor α in the frequency domain.

After the linear phase correction of NMR spectral lines, fast inverse Fourier transform is required to transform the NMR signal from the frequency domain to the time domain, so as to obtain the corrected NMR signal.

The invention has the advantages that, compared with the method of reducing the phase jitter of the nuclear magnetic resonance signal by increasing the receiver spectral width, it fundamentally eliminates the phase jitter of the nuclear magnetic resonance signal caused by the digital receiver; and on the premise of minimal hardware changes The phase jitter of the NMR is eliminated, and the implementation method is very simple and easy; at the same time, because the fast Fourier transform and inverse Fourier transform are used to correct the phase jitter of the NMR signal, the computing burden of the host is greatly reduced.

Figure overview

Accompanying drawing 1 is a digital receiver structural block diagram among the present invention;

Accompanying drawing 2 is the schematic diagram of nuclear magnetic resonance signal sampling in the present invention;

Accompanying drawing 3 is the schematic diagram of phase correction of nuclear magnetic resonance signal in the present invention;

Accompanying drawing 4 is the schematic diagram of spin echo pulse sequence in the present invention;

Accompanying drawing 5A is the time-domain schematic diagram of the original nuclear magnetic resonance echo signal;

Accompanying drawing 5B is a schematic diagram of the frequency domain of the nuclear magnetic resonance echo signal;

Accompanying drawing 5C is the time-domain schematic diagram of the echo signal after phase correction;

Accompanying drawing 6A is the raw data of NMR signal of SE sequence (without phase encoding);

Accompanying drawing 6B is the data of the SE sequence (without phase encoding) after phase correction;

Accompanying drawing 7 is the structural block diagram of digital receiver in the prior art;

Specific technical solutions

The features of the present invention and other related features will be further described in detail below in conjunction with the accompanying drawings through embodiments, so as to facilitate the understanding of those skilled in the art:

As shown in Figure 1-7, the labels 1-11 represent: probe (1), low noise preamplifier (2), mixer (3), intermediate frequency signal conditioner (4), ADC (5), digital Downconverter(6), Digital Quadrature Mixer(7A)(7B), Cascaded Comb Filter(CIC)(8A)(8B), Decimation Filter(9A)(9B), Mainframe(10) , counter (11), register (12).

Connect the counter (11) that is used to measure each error time on the nuclear magnetic resonance digital receiver, in each sampling, after this counter (11) gets the error time value of measuring and sends to main frame (10), then in the receiver spectrum The linear phase correction is performed on the spectral line of the nuclear magnetic resonance signal within a wide range, thereby eliminating the phase jitter of the nuclear magnetic resonance signal.

The specific implementation plan is generally divided into three steps:

The first step is to perform fast Fourier transform (FFT) on the acquired nuclear magnetic resonance signal, and transform the nuclear magnetic resonance signal from the time domain to the frequency domain.

In the second step, in the frequency domain, a first-level phase correction is performed on the spectral lines within the spectral width of the receiver, that is, the data point corresponding to each frequency point in the spectral width range is multiplied by the phase correction factor α.

In the third step, fast inverse Fourier transform is performed on the nuclear magnetic resonance spectral lines to obtain corrected nuclear magnetic resonance signals.

It can be seen from Figure 2 that there is a time error of t _p between the data point E corresponding to the sampling start time and the data point C output by the digital receiver. Assuming that the frequency of the NMR signal is f, the actual NMR signal can be expressed as:

A(t)＝A(0)e ^i2πf×t (1)

Assuming that the sampling spectrum width of the digital receiver is SW, the finally obtained nuclear magnetic resonance signal is

${\mathrm{<span\; class="notranslate">\; A</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;f</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; SW</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Comparing Equation 1 and Equation 2, it can be obtained that for each signal sampling, the phase jitter error is

0＜Δφ＜2πf/SW           (3)

For each sampling, if the value of the error time t _p can be measured, we can perform linear phase correction on the NMR signal by introducing a phase correction factor. The expression of the phase correction factor α is as follows:

$\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="i"\; filenames="A20041005315400091.png,A20041005315400101.png"\; state="\{\{state\}\}">i</figure-callout></span>}\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;f</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

Multiplying Equation 2 by Equation 4, we get

A(t)×α＝A(t)         (5)

It can be seen that the phase error caused by the digital receiver can be eliminated by multiplying the phase correction factor on the basis of the original data.

Fig. 3 is a schematic diagram of phase correction of nuclear magnetic resonance signals in the present invention. Wherein the counting clock of the counter (11) is connected to the system clock (the frequency is usually tens of megahertz). The sampling start signal is used as the counting start signal of the counter, and the data frame signal is used as the stop bit of the counter. When sampling starts, the counter starts counting at the rising edge of the sampling start signal, stops counting when the frame signal is high, and stores the value of the counter into the register (12). In Figure 3, the time from the start of the counter to the stop of the counter is t _c , and the counting frequency of the counter is f _clk , then the count value N of the counter and the error time t _p have the following relationship:

t _s −N/f _clk =t _p (6)

Among them, t _s represents the time interval between adjacent data points of the digital receiver, and the value of t _p can be measured by reading the count value N of the counter.

In order to eliminate the influence of the error time t _p on the nuclear magnetic resonance signal, the phase correction factor can also be used to perform a convolution operation on the nuclear magnetic resonance signal in the time domain to eliminate phase jitter.

Taking the spin echo (Spin Echo) sequence of nuclear magnetic resonance imaging as an example, it is shown in Figure 4, where Gs, Gp, and Gr are gradient fields, and their function is to spatially encode the sample. 90° and 180° are the flipping angles of the radio frequency pulse and the magnetization vector of the sample, which are used to excite the nuclear spin system of the sample and generate NMR signals. When the sampling starts, the digital receiver collects the nuclear magnetic resonance signal. For the imaging sequence in Fig. 4, the frequency of the radio frequency pulse is 10.7MHz, the spectral width SW of the digital receiver is 20KHz, the bandwidth of the nuclear magnetic resonance signal is 2KHz, and the number of sampling points TD is 256 points. According to formula 3, the maximum phase error generated by the digital receiver when receiving data is 2π×2KHz/20KHz=0.2π, which is 36°. In Fig. 4, the time interval between 90° and 180° radio frequency pulses is t=5ms, according to the nuclear magnetic resonance theory, the time t′ of the echo peak generated by the signal should also be 5ms. Fig. 5A is the original nuclear magnetic resonance signal in the time domain collected by the digital receiver. Due to the random jitter of the phase of the nuclear magnetic resonance signal, it can be seen from the figure that the peak of the echo is not at the time of 5ms. It can be seen that if the phase correction is not performed on the NMR signal, serious phase random jitter interference will be introduced in the phase encoding direction of the NMR data, which will definitely cause serious image artifacts. To phase correct the NMR signal, the following steps are taken:

In the first step, the nuclear magnetic resonance spectrum in the frequency domain is obtained after performing fast Fourier transform (FFT) on the time domain nuclear magnetic resonance signal, as shown in FIG. 5B .

In the second step, phase correction is performed on the NMR spectrum in the frequency domain as follows:

Since the sampling bandwidth of the receiver is 20KHz and the number of sampling points is 256 points, according to the nature of Fourier transform, the frequency value f of the nth data point on the spectrogram is obtained as

f _n ＝n×78.125Hz (0≤n＜256) (7)

Then the phase correction factor of each data point within the spectral width can be expressed as:

${\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; tp</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>}$

For this embodiment, the phase correction factor can be expressed as:

α _n =e ^{i2π×n×78.125×tp} (9)

Among them, t _p is the time error value of each sampling. For this embodiment, assuming that the spectral width of the digital receiver is 20 KHz, the maximum value of t _p is 50 us. When the system count clock is 50MHz, the maximum count value N is 50us*50MHz=2500. For this sampling, the counting value of the counter is set to 1000, then according to formula 6, the time error value t _p =50us-20us=30us is obtained. Then according to formula 9, multiply each data point in the original spectrogram obtained by the digital receiver by the phase correction factor α _n

${\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 78.125</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 30</span>}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 256</span>}<span\; class="notranslate">\; )</span>}$

After the α _n factor is calculated, the data of each frequency point in the obtained NMR spectrum from 0 Hz to the SW spectrum width is multiplied by the corresponding phase correction factor, so that the corrected NMR spectrum is obtained.

The third step is to perform an inverse Fourier transform on the corrected NMR spectrum to obtain a time-domain NMR signal without phase jitter, as shown in Figure 5C, after the phase correction of the echo vertex A, the phase jitter caused by each sampling is eliminated .

Figure 6A is the original nuclear magnetic resonance K-space data, each row of data in the K-space corresponds to a sampling of the nuclear magnetic resonance signal, and its signal amplitude is represented by grayscale, high brightness indicates a large signal amplitude, and low brightness indicates a small signal amplitude, from In the K space, it can be clearly seen that the phase jitter of the nuclear magnetic resonance signal is caused by each sampling of the digital receiver. Fig. 6B is the nuclear magnetic resonance signal K-space data obtained after being processed by this phase correction method in this embodiment. It can be clearly seen that after the nuclear magnetic resonance signal is corrected by this method, the digital The signal phase jitter per sample caused by the receiver.

Claims (6)

1. A method for receiving a nuclear magnetic resonance imaging signal, characterized in that a timing counter is arranged on the nuclear magnetic resonance digital receiver, and in each sampling, the counter measures each sampling time error value. 2. A method for receiving nuclear magnetic resonance imaging signals according to claim 1, characterized in that the linear phase correction is performed on the nuclear magnetic resonance spectral lines within the receiver spectral width by using the sampling time error value obtained by the counter. 3. A method for receiving nuclear magnetic resonance imaging signals according to claim 1 or 2, characterized in that the phase correction factor α is calculated by using the time error value obtained by the counter. 4. The receiving method of a nuclear magnetic resonance imaging signal according to claim 1 or 2, characterized in that before performing linear phase correction, the collected nuclear magnetic resonance signal needs to be subjected to fast Fourier transform, and the nuclear magnetic resonance signal is transformed from time to time domain to frequency domain. 5. A method for receiving nuclear magnetic resonance imaging signals according to claim 1 or 2, characterized in that linear phase correction is performed on the spectral lines within the spectral width of the receiver, which means that in the frequency domain, for the spectral The data points corresponding to each frequency point in the wide range are multiplied by the phase correction factor α. 6. A method for receiving a nuclear magnetic resonance imaging signal according to claim 1 or 2, characterized in that the nuclear magnetic resonance spectral line after linear phase correction needs to be subjected to fast inverse Fourier transform, and the nuclear magnetic resonance signal is transformed from the frequency domain to the time domain area.

Images