JPH0239254B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an ultrasonic diagnostic device, particularly an ultrasonic diagnostic device that measures and analyzes various information such as in-vivo morphological and dynamic conditions or intra-tissue conditions from received ultrasonic waves. This invention relates to an improved ultrasonic diagnostic device.

[Prior art] A method has been developed to analyze received ultrasonic waves from within a living body in order to discover abnormal tissue within a living body and perform morphological analysis and qualitative analysis of the abnormal tissue. .

It has been found that the distribution characteristics of the frequency spectrum of ultrasound waves transmitted into a living body and reflected and scattered within the living tissue change characteristically depending on the attenuation and scattering frequency characteristics of the tissue. In particular, it has been recognized that it is useful to determine the attenuation coefficient of ultrasound waves in tissues in the living body and use this as in-vivo information. Conventionally, this attenuation coefficient has been determined by storing the received ultrasound signal and analyzing it on a computer, but this method has the drawback of requiring a great deal of time and effort, and it is difficult to determine the attenuation coefficient at the exact position within the body. It has not been realized that the information can be understood as information, and it has not been possible to actually use it as effective information.

[Object of the Invention] The present invention has been made in view of the above-mentioned conventional problems, and its purpose is to obtain the average frequency and dispersion of the power spectrum of an ultrasonic reception signal in real time, and at the same time calculate the attenuation coefficient in biological tissue. It is an object of the present invention to provide an ultrasonic diagnostic apparatus that can accurately determine the morphology and properties of tissues in a living body by determining them in real time, and can easily perform tissue diagnosis.

[Structure of the Invention] In order to achieve the above object, the ultrasonic diagnostic apparatus of the present invention mixes a set of complex reference signals having a complex relationship with each other and the received high frequency signal in a certain time domain, and mixes the received high frequency signal. a complex signal converter that converts a signal into a complex signal consisting of a real part and an imaginary part; a complex signal converter that calculates an absolute value of the complex signal and separately divides the real part and the imaginary part by the result of the absolute value calculation;
A normalization circuit normalizes the complex signal so that the maximum value of the amplitude becomes 1, and a delay circuit delays the normalized complex signal by providing a predetermined delay time to calculate the autocorrelation value of the complex signal. It consists of an autocorrelator for calculation, a delay circuit, an addition circuit, and a weighting circuit, and calculates an average autocorrelation value by averaging the real part and imaginary part of the autocorrelation value, respectively, to remove extra signal components. an average autocorrelation circuit comprising: an average autocorrelation circuit comprising an average autocorrelation circuit; an average frequency calculator that calculates a phase angle from the real part and imaginary part of the average autocorrelation value and calculates an average frequency from the phase angle;
a dispersion calculator that squares and adds the real part and imaginary part of the average autocorrelation value, respectively, and calculates a variance value from the autocorrelation function power fluctuation based on the addition result; Find the difference between the average frequency and the average frequency delayed by a predetermined time,
an average attenuation coefficient calculation unit that calculates an attenuation coefficient by dividing by the dispersion value, and stabilizes the attenuation coefficient by an averaging circuit including a delay circuit, an addition circuit, and a weighting circuit to calculate an average attenuation coefficient; analyzing the power spectrum of the received high-frequency signal from within the living body for each fixed time domain, determining the average frequency and dispersion value in real time, and measuring the attenuation coefficient in real time,
It is characterized by displaying changes in the attenuation coefficient.

[Embodiments] Preferred embodiments of the present invention will be described below based on the drawings.

FIG. 1 shows a block circuit of an ultrasonic diagnostic apparatus according to an embodiment of the present invention.

The output of the crystal oscillator 10, which generates a stable high frequency signal, is supplied to a frequency division synchronization circuit 12, and the frequency division synchronization circuit 12 obtains various output signals of desired frequencies. These output signals include a transmission repetition frequency signal 100 for ultrasonic pulse beam transmission, complex reference signals 102 and 104 for complex conversion, and a sweep synchronization signal 106 for displaying ultrasonic diagnostic results.
and a clock signal 1 for synchronizing each part of the device.
Including 08. In the embodiment, the complex reference signals 102 and 104 have a complex relationship with each other.
It has a phase difference of 90°.

The transmission signal 100 is supplied to the probe 18 via the drive circuit 14 and the transmission/reception switching circuit 16, excites the probe 18, and transmits an ultrasonic pulse beam into the subject 20.

The reflected echo from the subject 20 is converted into an electrical signal by the probe 18, and sent from the transmission/reception switching circuit 16 to the high frequency amplifier 22 where it is subjected to the desired amplification action, and then the output of one of the signals is converted into a normal signal. It is supplied to the display section as a B mode or M mode display signal,
The other output is supplied to the arithmetic processing section in order to be computed as information on the morphology, properties, etc. of the living tissue.

The output signal for normal B-mode or M-mode display is sent to a detector 24 and a video amplifier 26.
The light is supplied to the CRT display 30 via the DSC 28, and the surface of the CRT display 30 is modulated in brightness.

The ultrasonic pulse beam of the probe 18 is scanned by mechanical or electrical angle deflection, the object 20 is periodically scanned with the ultrasonic pulse beam, and the scanning is stopped at a desired deflection angle. A scan controller 32 is provided for the purpose of this, and the scan position signal of the scan controller 32 is synchronized with the clock signal 108 and supplied to the probe 18, and at the same time the sweep synchronization signal obtained from the frequency dividing synchronization circuit 12 is synchronized with the clock signal 108. Together with 106, it is supplied to the DSC 28 for display position control.

The characteristic feature of the present invention is that the average frequency or dispersion of the power spectrum of the received ultrasound signal reflected from within the living body is determined in real time, and at the same time, the attenuation coefficient of the living body is calculated in real time from this average frequency and the dispersion value. This is done by setting the ultrasonic reception signal obtained from the living body as e(t) and setting a constant short time interval G to the waveform of e(t), as shown in Figure 2. . Then, the average frequency and dispersion or attenuation coefficient of the power spectrum of the received signal within this G are determined, and then G is shifted little by little to obtain data within each G and these are combined to analyze the ultrasound received signal. will be held.

In order to obtain the average frequency and dispersion value of the power spectrum or the in-vivo attenuation coefficient in real time, a device is used that complex-transforms the received signal and calculates its autocorrelation function. The configuration will be explained below.

Since the other output of the high frequency amplifier 22 is used for autocorrelation calculation, the received high frequency signal obtained from the high frequency amplifier 22 is first supplied to the complex signal converter 36 and converted into a complex signal.

That is, in the embodiment, the complex signal converter 36 includes a pair of mixers 38a, 38a and 38a, including a phase detector.
8b, and in each mixer 38, the received high frequency signal is connected to the complex reference signal 102, 10, respectively.
Since the complex reference signals 102 and 104 are in a complex relationship with a phase difference of 90 degrees from each other as described above, the mixer 38 can output a complex signal corresponding to the received high frequency signal. That is, each mixer 38 outputs a signal having a frequency that is the sum and difference of the frequencies of the received high-frequency signal and the complex reference signal input through mixed detection, and these two signals are supplied to the low-pass filters 40a and 40b. and only the difference frequency components are extracted.

In the mixed detection function of the mixer 38, the complex reference signals 102 and 104 are continuous waves of a single frequency, but the received high frequency signal, which is the other input signal, is a pulse wave containing information in living tissue, so the reduction filter 40 A large number of spectral components will appear in the output. The complex conversion using this complex signal converter is explained in detail in Japanese Patent Application No. 57-70479, and the complex signal obtained by this is expressed by the complex formula Zo=Xo+iYo, where the real part is Xo and the imaginary part is Yo... 1).

Next, the complex signal that has undergone complex conversion is input to the normalizer 44, and the amplitude of the complex signal in equation (1) is normalized by the normalizer 44 so that its maximum value is "1". Ru.

This normalizer 44 is provided to perform highly accurate calculations since digital multiplication calculations are performed in the subsequent autocorrelation calculations.

In other words, in the subsequent autocorrelation calculation, it is necessary to calculate the received signal from a very large signal to a very small signal with the same precision, so if even a small error occurs, it is integrated by multiplication (integration). That will happen. Therefore, in order to eliminate such problems as much as possible, all amplitude information is normalized before the autocorrelator.

This makes it possible to efficiently calculate the phase information necessary to calculate the average frequency later.

Furthermore, by performing such normalization, there is an advantage that the subsequent calculation of the variance value can be performed very stably and accurately.

With such a normalizer 44, the absolute value √ ² + ² of Zo in equation (1) is calculated in an absolute value calculator 46, and the real part and imaginary part are divided into absolute values separately by dividers 48a and 48b. When divided by the value √ ² + ² , the complex signal becomes a signal whose maximum amplitude value is 1, and as a result, the angles when Xo and Yo are expressed by sin and cos contain in-vivo information. It will be. Therefore, the complex signal obtained and normalized by the normalizer 44 is It is indicated by.

The complex signal normalized as described above is then subjected to arithmetic processing by the autocorrelator 50, and the autocorrelation of _Z1 , which is the delay amount T, is determined.

First, the input signal Z ₁ is input to the delay lines 52a, 52
Z ₂ is obtained by delaying one period by b. This output Z ₂ is expressed by the following formula.

Z ₂ = X ₂ + _{iY 2 ...} (3) Then, if Z ₂ ^* ₌
Based on the relationship in the equation, the correlation can be calculated using the following equation.

Z ₁ Z ₂ ^* = (X ₁ + iY ₁ ) (X ₂ − iY ₂ ) = X ₁ X ₂ + Y ₁ Y ₂ + i ₍ X _{2 Y 1} − In order to obtain the correlation, the autocorrelator 50 includes four multipliers 54a, 54b, 56.
a, 56b, and adders/subtractors 58a, 58b are provided to perform the correlation calculation.

Furthermore, if _the output of the adder _/ subtractor _{58a is} R, then R=X _{1 1} −Y ₁ Y ₂ ...(6) is obtained, and the outputs of both adders and subtracters 58 are combined and expressed by the following equation.

S=R+iI...(7) Then, this output S includes signal fluctuation components and noise components generated from the equipment, so in order to remove these noise components, an average is obtained by an averaging circuit, and this average is =+i, and complex correlation is calculated.

The average circuit includes delay lines 60a and 60b.
The output delayed by one round is added to the input signal at the current time by adders 62a and 62b, and the operation of supplying this output to the delay line 60 is repeated, and this addition is performed by, for example, a digital circuit. In this case, the average value can be obtained by outputting the upper bits of the summed output. However, if you simply repeat this operation, as the number of additions increases,
The output value increases successively and finally reaches saturation. Therefore, in the embodiment, the weighting circuits 64a, 6
4b is provided to attenuate the output and add it to the input. In other words, if the attenuation amount is α, the signal ^α10 , which is 10 cycles earlier than the current time signal, will be attenuated and added to the current time signal, so the influence on the output will be small, and the low-pass filter or It becomes possible to perform the same averaging function as a moving average circuit.
Furthermore, by changing the weighting amount of the weighting circuit 64, it is possible to change the degree of averaging.

As described above, the autocorrelation value obtained from the autocorrelator 50 as the correlation of the complex signal is expressed as =+i (8).

As mentioned above, the autocorrelation value obtained in the above manner serves as the basis for analyzing the power spectrum of the received signal within a fixed short time interval G in FIG. This enables real-time processing of power spectrum analysis.

Next, the real part and imaginary part of the autocorrelation value obtained by the autocorrelator 50 are input to an average frequency calculator 66 that calculates an average frequency and a variance calculator 68 that calculates a frequency variance value.

In the average frequency calculator 66, the average frequency is calculated based on the phase angle of the autocorrelation function, and the explanation will be given below.

When the power spectrum of the envelope signal in G is P(f), the autocorrelation function S(τ) is expressed by the following relational expression.

S(τ)=∫ ^∞ _-∞ P(f)e ^j2 〓 ^f 〓df ……(9) Differentiating this, S〓(τ)=S(τ)/dτ=j 2π∫fp(f)e ^j2 〓 ^f
〓df...(10), and if we set π=O here, then S〓(O)=j 2π∫ ^∞ _-∞ fp(f)df. Furthermore, if τ=O for S(τ), then S〓(O)=∫p(f)df. Therefore, S〓(O)/S(O)=j2π∫ ^∞ / _-∞ fp(f)df/∫ ^∞
/ _-∞ p(f)df ≒j2π ...(11), which is obtained as the average frequency.

On the other hand, since the autocorrelation function S(τ) is a complex number, it can be expressed as follows.

S(τ)=A(τ)e ^j 〓 ⁽ 〓 ⁾ ……(12) It is. Also, from the properties of the autocorrelation function, the ratio of S〓(O)/S(O) can be transformed as follows.

S〓(O)/S(O)=jφ〓(O) ……(15) Therefore, from equations (11) and (15), =1/2πφ〓(O) ……(16) and self It can be determined from the phase angle ψ(τ) of the correlation function. However, the differential value φ(O) of the phase angle is φ〓(O)=φ(τ)−φ(O)/τ, and assuming φ(O)=0, the approximate formula φ〓(O)≒φ(τ)/ τ can be found using (17). Therefore, as shown in equation (14), φ(τ) can be obtained from R and , and from equation (17), φ〓
(O) is required, is given by

Here, since the average frequency calculator 66 calculates the calculation result from the autocorrelator 50 including the average circuit based on the equation (18), for example, the frequency fluctuation (phase Fluctuation)
If the error is severe, the error can be reduced and the measurement can be performed with high precision.

On the other hand, in the variance calculation unit 68, a variance value is obtained from the values obtained by the autocorrelator 50, and this variance calculation will be explained below.

The second differential of the autocorrelation function (12) is expressed by the following equation.

S¨(O)=A¨(O)−A(O) {φ〓(O)} ² ...
…(19) Also, the second differential of equation (9) is R¨(O)=(j2π) ² ∫f ² p(f)df≒(j2π) ^{2 2} …(2
0), while the variance value σ ² is generally expressed as σ ² =∫ ^∞ / _-∞ (f - f / -) ² p(f)df / ∫ ^∞ / _-∞ p
(f)df=−() ² ……(21) is defined. Substituting equations (19) and (20) into equation (21) approximately yields σ ² =-1/2π ² τ ² S(O)-|S(τ)|/S(O
)...(22) is obtained. This represents the rate of change of S(τ) with respect to the power S(O), and it is understood from this equation that σ ² is proportional to the power fluctuation.

Here, since S(τ) has already been calculated using the normalized complex signal Z ₁ , the fractional value σ ² can be calculated by the following equation.

That is, in the distributed arithmetic unit 68, the square multiplier 70
R, calculated by the autocorrelator 50 using a, 70b, is squared and becomes ² , ² , and these ² and ²
are added by adder 72 to become ² + ² . This is calculated by the variance value calculator 74 according to equation (23), and the variance value σ ² is calculated.

The calculation of this variance can be approximately performed using equation (23) by the normalizer 44, which has the advantage of being very stable and accurate.

The average frequency and variance value σ ² calculated as above are values within a certain region G of the ultrasonic reception signal,
This region G is moved little by little and σ ² is calculated for each region one by one. The waveform obtained in this way is shown in FIG.

Next, in the embodiment, an attenuation coefficient calculator 76 is provided to calculate the attenuation coefficient in living tissue.
Average frequency and variance value σ ² of the power spectrum
The damping coefficient is calculated by .

The attenuation coefficient can be calculated by dividing the differential value of the average frequency by the variance value. It is obtained by setting a point and dividing the difference between the average frequency _A at point A and the average frequency _B at point B by the variance value σ ² . That is, the average frequency _A − _B is regarded as the differential value of the average frequency f described above, and when Δf= _A − _B , the attenuation coefficient is expressed as α=Δf/σ ² (24).

Therefore, in the attenuation coefficient calculator 76, first, the average frequency _B delayed by the time l by the delay line 78 is calculated and inputted to the attenuator 80 _. The difference Δf is calculated. Next, divider 8
2, the dispersion value σ ² which is the output from the dispersion calculator 68 is input, and the attenuation coefficient α is calculated by dividing Δf obtained by the attenuator 80. This attenuation coefficient α is averaged by an averaging circuit in the same manner as the autocorrelator 50 described above, and the average value is calculated by a delay line 84, an adder 86, and a weighting circuit 88. Fig. 4 shows the signals calculated by each calculation process of the damping coefficient calculator, and the average value of the damping coefficients finally obtained becomes a stable signal as shown in Fig. 4c. is a signal indicating the attenuation coefficient in the living body.

The average frequency and dispersion value σ ² of the power spectrum obtained as described above or the in-vivo attenuation coefficient are converted into image signals and selectively displayed on the screen of the CRT 30. In the embodiment, for example, the attenuation coefficient Colors are determined according to changes in the attenuation coefficient, and changes in the attenuation coefficient at each location within the body can be displayed as changes in color.

[Effects of the Invention] As explained above, according to the present invention, the average frequency and variance of the power spectrum are determined in real time using autocorrelation, and the attenuation coefficient is calculated as necessary. It becomes possible to determine the attenuation coefficient in the body in real time, which makes detailed morphological or qualitative analysis of the tissue at each location in the body extremely easy.

Therefore, it is also possible to perform so-called histological diagnosis, for example, whether abnormal tissue is benign or malignant, even with ultrasound, and it is also easy to display images of the progress status of lesions and other useful information. Alternatively, by using it together with M-mode images, it can contribute to various diagnoses.

[Brief explanation of drawings]

FIG. 1 is a block circuit diagram showing a preferred embodiment of the ultrasonic diagnostic apparatus according to the present invention. FIG. 2, FIG. 3, and FIG. 4 are explanatory diagrams of arithmetic processing of ultrasonic reception signals in each arithmetic unit in this embodiment. 10...Crystal oscillator, 12...Divide period circuit, 18
...Probe, 20...Subject, 30...CRT, 36...
Complex signal converter, 38a, 38b... mixer, 40
...Low pass filter, 44...Normalizer, 46...Absolute value calculator, 48a, 48b...Divider, 50...Autocorrelator, 52a, 52b...Delay line, 54
a, 54b, 56a, 56b...multiplier, 58a,
58b... Adder/subtractor, 66... Average frequency calculator, 6
8... Distributed arithmetic unit, 70a, 70b... Square multiplier, 7
2...Adder, 74...Dispersion value unit, 76...Attenuation coefficient calculator, 78...Delay line, 80...Subtractor, 8
2...Divider.

Claims (1)

[Claims] 1. In an ultrasonic diagnostic apparatus that transmits an ultrasonic pulse beam into a living body at a constant repetition frequency and uses an autocorrelator to determine and display an attenuation coefficient from a received high-frequency signal as a reflected wave, a complex signal converter that mixes a set of complex reference signals having a complex relationship and the received high frequency signal in a certain time domain and converts the received high frequency signal into a complex signal consisting of a real part and an imaginary part; a normalization circuit that calculates the absolute value of a signal, divides the real part and the imaginary part separately by the result of the absolute value calculation, and normalizes the complex signal so that the maximum value of the amplitude is 1; An autocorrelator for calculating the autocorrelation value of the complex signal by delaying the complex signal by a predetermined delay time using a delay circuit, and a delay circuit, an addition circuit, and a weighting circuit to remove excess signal components. an averaging circuit that averages the real part and the imaginary part of the autocorrelation value to calculate an average autocorrelation value in order to remove the real part and the imaginary part of the average autocorrelation value; an average frequency calculator that calculates a phase angle from the phase angle and calculates an average frequency from the phase angle; and an average frequency calculator that calculates and calculates the average frequency from the phase angle; a dispersion calculator that calculates a dispersion value from the power fluctuation of the autocorrelation function based on the above-mentioned information; and a dispersion calculator that calculates a dispersion value from the power fluctuation of the autocorrelation function; and an average attenuation coefficient calculator that calculates the average attenuation coefficient by stabilizing the attenuation coefficient by an averaging circuit including a delay circuit, an addition circuit, and a weighting circuit; Ultrasonic diagnosis characterized by analyzing the power spectrum of a high frequency signal in each fixed time domain to obtain the average frequency and dispersion value in real time, measuring the average attenuation coefficient in real time, and displaying changes in the attenuation coefficient. Device. 2. The apparatus according to claim 1, further comprising an image display selection means for selecting one of the calculation results of the average frequency, dispersion value, and attenuation coefficient, converting it into an image display signal and outputting it, An ultrasonic diagnosis characterized by displaying an ultrasonic tomographic image of the image and a change in the average frequency of the power spectrum, a change in dispersion, or a change in the attenuation coefficient of the received ultrasonic signal simultaneously or separately using different colors. Device.