JP3418652B2 - Ultrasonic device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ultrasonic device, and more particularly to an ultrasonic device capable of measuring a pulse wave propagation velocity with high accuracy.

[0002]

2. Description of the Related Art Arteriosclerosis causes narrowing or blockage of the inner wall of blood vessels, which may lead to serious diseases such as ischemic heart disease and cerebrovascular accident. Is becoming. Therefore, it is necessary to diagnose the degree of arteriosclerosis at an early stage, determine the degree of severity, and perform quantitative follow-up after treatment. Moreover, arteriosclerosis is relatively common in major organ arteries, rather than developing uniformly in systemic blood vessels.
It is also known that the degree of lesion locally varies. Therefore, it is strongly pointed out that the measurement at the deep part of the blood vessel is important.

On the other hand, as a method for diagnosing arteriosclerosis in clinical practice at present, (a) an angiography method in which a contrast medium is injected into a blood vessel and an X-ray is taken is generally used.
This is an invasive method. Further, (b) evaluation of morphological information and calcification of blood vessel wall by simple X-ray, ultrasonic wave, X-ray CT and nuclear magnetic resonance tomography, and determination of degree of hardening of cerebral blood vessel by optical fundus examination. However, since morphological information is extracted in each case, early diagnosis is often difficult.

On the other hand, in order to non-invasively quantitatively evaluate the blood vessel characteristics, (c) pulse wave (deformation of the blood vessel wall due to the pressure of blood ejected from the heart is transmitted along the blood vessel. (Pressure pulse waveform)) propagation velocity (hereinafter referred to as pulse wave velocity) is measured, and it is studied to use this as an index of arteriosclerosis degree. Here, in the most basic pulse wave velocity measuring method, the pressure pulse wave is recorded at two points on the body surface that are far apart from each other, such as the carotid artery position and the hip artery position, and the average value is obtained by the propagation time of the pressure pulse waveform between them. The pulse wave velocity is measured. However, the average pulse wave velocity measured by this method cannot be locally measured because it is an average velocity in a long section, and since the detectable depth of the pressure pulse waveform is shallow, deep measurement is also possible. It will be difficult.

Therefore, the present inventor has devised a method for measuring the propagation time of a pulse wave by complex cross-correlation processing of blood flow Doppler signals measured at two points (Japanese Patent Application No. 4-26).
1111, Japanese Patent Application No. 5-51669). This processing is a method of performing correlation processing on blood flow Doppler signals without converting them into a flow velocity waveform.

[0006]

In the method of measuring the propagation velocity of the pulse wave by measuring the delay time between the blood flow Doppler signals, the signal processing for reducing the correlation between the blood flow Doppler signals is performed. Will be required. As one of the methods, a configuration for performing a moving average of phase differences has been proposed.
The measurement accuracy is insufficient. An object of the present invention is to provide an ultrasonic device that enables highly accurate pulse wave velocity measurement in blood vessels.

[0007]

According to the present invention, signal processing is performed by a novel structure to improve measurement accuracy. That is, a receiver for receiving the Doppler signals of the target object from a plurality of positions of the inspection target, a means for measuring the phase of each received signal by the receiver, a means for measuring the phase difference of each received signal, A means for deriving an approximate phase difference function that approximates a time change of the phase difference, a means for measuring a mutual time delay of the approximate phase difference function, a means for measuring a distance between two receiving points, a time delay and a distance. Is characterized in that the pulse wave propagation velocity between the two receiving points is measured.

[0008]

In the ultrasonic device of the present invention, a function approximating the phase difference of each Doppler signal measured at two points is obtained by the least square method, and the time delay between the approximating functions is measured. Can measure the wave propagation velocity.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, the signal processing method used in the ultrasonic device of the present invention will be described in detail. Using the two transducers TD1 and TD2 shown in FIG. 1, the reflected Doppler signals a (t) and b (t) from the blood cells A and B in the sample volumes SV1 and SV2 are obtained by the pulse Doppler method. Here, U1 (t) and U2 (t) are the moving velocities of blood cells A and B (blood flow velocity components in the normal direction of the beam), and this is called a flow velocity waveform. In addition, v is the propagation velocity of the flow velocity waveform, L is SV1 and SV2
If the distance is between, the propagation time Tp of the flow velocity waveform is L / v,

[0010]

## EQU1 ## U2 (t) = U1 (t-Tp) = U1 (tL / v) (Equation 1). Furthermore, when k (= 2π / λ) is the wave number of the ultrasonic wave in blood, λ is the wavelength of the ultrasonic wave, and Θa and Θb are the initial phases of the reflected Doppler signals, the reflected Doppler signals a (t), b
(T) is shown by (Equation 2) and (Equation 3), respectively. (Equation 3) is transformed from (Equation 4) to (Equation 7).

[0011]

[Equation 2] a (t) = exp {jΘa + 2jk∫U1 (τ) dτ}… (Equation 2) The range of integral ∫ in (Equation 2) is 0 to t.

[0012]

[Equation 3] b (t) = exp {jΘb + 2jk∫U2 (τ) dτ}… (Equation 3) b (t) = exp {jΘb + 2jk∫U1 (τ-Tp) dτ}… (Equation 4) b (t) = exp {jΘb + 2jk∫U1 (τ) dτ}… (Equation 5) b (t) ＝ exp {jΘb + 2jk∫U1 (τ) dτ + 2jk∫U1 (τ) dτ}… (Equation) 6) b (t) = exp (jΘ) exp {2jk∫U1 (τ) dτ} = exp (jΘ) a (t-Tp) (Equation 7) Integral ∫ in (Equation 3) and (Equation 4) Range is 0 to t,
The range of integral ∫ in (Equation 5) is -Tp to t-Tp, the range of first term integral ∫ in (Equation 6) is -Tp to 0, the range of second term integral ∫ is 0 to t-Tp, ( The range of integral ∫ in equation 7) is from 0 to t-Tp. here,

[0013]

## EQU8 ## Θ = Θb−Θa + 2jk∫U1 (τ) dτ (Equation 8), and the phase angles Θa, Θb and Θ are all constant. The range of integral ∫ in (Equation 8) is from -Tp to 0. Thus, the time difference between the two waveforms indicates the propagation time Tp (= L / v) of the flow velocity waveform. Here, for simplification, the sample volumes SV1 and SV2 of the ultrasonic waves are large, and the times T1 and T2 during which the blood cells are present in the beam are longer than the time T0 during which the blood cells are accelerated. The case where the sample volume is small will be described later. On the other hand, it is well known that the flow velocity waveforms U1 (t) and U2 (t) correspond to the pressure pulse wave waveform, and the direct correlation of the Doppler signal indicates the propagation time of the pressure pulse wave.

In the method of measuring the propagation time of the pulse wave using the blood flow Doppler signal waveform, when the sample volume is small, it is necessary to perform signal processing as a measure for reducing the degree of correlation between blood flow Doppler signals. This problem will be described in detail below. It is assumed that the three blood cell positions existing at time t = 0 are a, b, and c in FIG. 2, respectively. In pulse Doppler, the reflection signal is measured at each time interval t0, but the blood cell position at each measurement time is indicated by a circle in FIG. 2, and the number written together with the circle corresponds to the measurement time. Here, since the systole (a time phase in which blood cells are accelerated) is assumed, the intervals indicated by circles in FIG. 2 are gradually expanded. The reflected signal is generated corresponding to the period during which each blood cell is present in the beam, and the reflected Doppler signal (referred to as a partial signal) from each blood cell is shown in FIG. 3 (a), FIG. 3 (b), and FIG.
As shown in (c). The frequencies of these partial signals monotonically increase with the acceleration of blood cells, but their phase relationships are independent. Therefore, the received Doppler signal in the pulse Doppler blood flow meter obtained as a superposition of these partial signals has an irregular waveform shown in FIG. Such a correlation function between waveforms shows multimodality, and precise time measurement is impossible.

As described above, each partial signal appears corresponding to the time when each blood cell is present in the beam. Therefore, the initial phase, which is the phase at the time when the blood cell corresponding to the partial signal existing at time t reaches the ultrasonic beam, is n (t), and the phase change due to the blood cell movement from that position is s (t). Then, the phase φ (t) of the received signal is

[0016]

## EQU00009 ## .phi. (T) = n (t) + s (t) (Equation 9). Here, the initial phase n (t) is determined by the initial position where each blood cell enters the ultrasonic beam, and is independent among blood cells. Therefore, the signal phase φ (t) jumps discontinuously when the first partial signal comes out of the ultrasonic beam and another partial signal enters the ultrasonic beam, and the waveform is not continuous. Change to rules (Figure 3). The phase difference Δφ (t) between adjacent times of such a received signal is

[0017]

[Formula 10] Δφ (t) = φ (t + t0) −φ (t) = n (t + t0) −n (t) + s (t + t0) −s (t) = Δn (t) + Δs ( t) (Equation 10) Here, Δn (t) differs depending on the situation of the difference, Δn (t) = 0 in the same blood cell partial signal, and Δn (t) between different blood cell partial signals.
n (t) ≠ 0. Here, it is considered that the phase difference Δφ (t) is approximated to the least square by the function c (t). here,

[0018]

(Equation 11) c (t) = c0 + c1t + c2t2 + ... (Equation 11) If the discrete time is tm = m × t0, the mean square value Q of the difference between Δφ (t) and arbitrary function c (t) is

[0019]

[Equation 12] Q = Σ {Δφ (tm) -c (tm)} ² / M = Σ {Δn (tm) + Δs (tm) -c (tm)} ² / M (Equation 12) Here, if d (t) = Δs (t) -c (t),

[0020]

[Equation 13] Q = Σ {Δn (tm) + d (tm)} ² / M = Σ {(Δn (tm)) ² + (d (tm)) ² + 2Δn (tm) d (tm)} / M ... (Equation 13) In addition, in (Equation 12) and (Equation 13), addition Σ is performed for m = 1, ..., M. Where Δn (t) and d (t)
Has no correlation, the third term on the right side is zero. For this reason,

[0021]

[Expression 14] Q = Σ {(Δn (tm)) ² + (d (tm)) ² } / M (Expression 14). When the phase difference function c (t) that minimizes this Q is csq (t), csq (t) is a least-squares approximation function of the phase difference Δφ (t), and the minimum value Qmin of Q is

[0022]

[Equation 15] Qmin = Σ {(Δn (tm)) ² } / M = <(Δn (t)) ² > ... (Equation 15) In addition, in (Equation 14) and (Equation 15), addition Σ is performed for m = 1, ..., M. Here, <> means averaging processing. Since the relation of this (Equation 15) is established in d (t) ≡0,

[0023]

## EQU16 ## d (t) = Δs (t) -csq (t) ≡0 (Equation 16) Therefore, csq, which is the least-squares approximate phase difference function
(t) gives correct phase difference information Δs (t). Therefore, integrating this csq (t),

[0024]

[Equation 17] Csq (t) = ∫csq (t) dt = ∫Δs (t) dt = s (t) (Equation 17) and the correct phase signal s (t) is obtained, and the irregular phase fluctuation is obtained. The corrected signal waveform h (t) is calculated by Csq (t),

[0025]

[Expression 18] h (t) = exp {jCsq (t)} = exp {js (t)} (Expression 18) The range of the integral ∫ on each side in (Equation 17) is from 0 to t.

In the present invention, there are two Doppler signals a
Least-squares approximate phase difference function csq for (t) and b (t) respectively
Find (t) and call these ca (t) and cb (t). These phase difference functions are

[0027]

[Formula 19] ca (t) = a0 + a1t + a2t2 + ... (Equation 19)

[0028]

[Equation 20] cb (t) = b0 + b1t + b2t2 + ... (Equation 20) Is obtained.

In the first configuration of the present invention,
Two signal waveforms ha (t) and hb (t) corresponding to the signal waveform h (t) are obtained from the least-squares approximated phase difference functions ca (t) and cb (t). Then, the time delay is measured by complex correlation processing of those signal waveforms ha (t) and hb (t). Where the complex correlation function R
(Τ) is

[0030]

(21) R (τ) = | ∫ha (t) [hb (t + τ)] dt | ... (Equation 21) In (Equation 17), [] represents a conjugate complex number, the range of the integral ∫ is from 0 to T0, and T0 is the time length of the wideband signal portion in which the frequency greatly changes in FIG. This R (τ) is unimodal because the phase discontinuity has been corrected, and since τ = Tp has the maximum value, accurate delay time measurement is possible.

In the second configuration of the present invention, 2
The time delay is directly measured from the least-squares approximated phase difference functions ca (t) and cb (t) obtained for the Doppler signals a (t) and b (t) at various locations. Since ca (t) and cb (t) change linearly in normal arterial blood flow, in the following, ca (t) and cb (t) are assumed to be linear functions of t, and ca (t) = Let a0 + a1t and cb (t) = b0 + b1t.

Here, the least-squares approximation phase difference functions ca (t) and cb
As the first method to directly measure the time delay from (t),
The configuration using the least squares method will be described again. Function ca (t)
QQ is the mean squared value of the difference between cb (t) and

[0033]

[Equation 22] QQ (τ) = ∫ {ca (t) -cb (t-τ)} ² dt / T0 = ∫ {(a0-b0 + b1τ) + (a1-b1) t} ² dt / T0 = ^{∫ {x + yt} 2 dt} / T0 = ∫ {x 2 + 2xyt + y 2 t 2} dt / T0 = x 2 + y 2 T0 2/3 + xyT0 ... ( Equation 22) sides in equation (22) The integral ∫ of is in the range 0 to T0. Here, the minimum value is given by dQQ / dτ = 0, so

[0034]

DQQ / dτ = 2b1 (a0-b0) + (a1-b1) b1T0 + 2b1 ² τ = 0 (Equation 23) Here, since the time difference τ for matching the functions ca (t) and cb (t) is Tp,

[0035]

The propagation time Tp of the pulse wave is known from Tp = {(b0-a0) + (b1-a1) T0 / 2} / b1 (Equation 24). Where the signal a (t)
Since it is possible to measure the time difference even if the b and t are exchanged,

[0036]

[Equation 25] Tp ′ = {(b0-a0) + (b1-a1) T0 / 2} / a1 (Equation 25) similarly measures the propagation time of the pulse wave. With this Tp
Originally, Tp 'should match, but a1 and b1 do not completely match when there are differences in the pulse wave conditions at the two locations. Therefore, by the average gradient (a1 + b1) / 2,

[0037]

[Equation 26] Tp "= {2 (b0-a0) + (b1-a1) T0} / (a1 + b1) (Equation 26) Thus, the pulse wave propagation time is similarly measured.

Next, the least-squares approximation phase difference function ca (t), cb
As a second method of directly measuring the time delay from (t), a method of measuring the delay time by setting the average value of the errors to 0 will be described. If QM is the average value of the difference between the least squares approximate phase difference function ca (t) and cb (t),

[0039]

[Equation 27] QM (τ) = ∫ {ca (t) -cb (t-τ)} dt / T0 = ∫ {(a0-b0 + b1τ) + (a1-b1) t} dt / T0 = (a0 -b0 + b1τ) T0 + (a1 -b1) T0 2/2 ... ( number 27) (the range of the integral ∫ of each side in the number 27) is from 0 T0. Here, since QM = 0 gives the delay time,

[0040]

[Number 28] is a (a0-b0 + b1τ) T0 + (a1-b1) T0 2/2 = 0 ... ( number 28), from the time difference τ is the propagation time difference Tp,

[0041]

The propagation time Tp of the pulse wave can be obtained by Tp = {(b0-a0) + (b1-a1) T0 / 2} / b1 (Equation 29). This relationship is consistent with the result of the first direct measurement method using the least square method (Equation 24).

The above description has been made by measuring the time difference independently for each heartbeat, but since the time delay Tp is constant for each heartbeat, the R due to each heartbeat is
A configuration is also possible in which (τ), (b0-a0), (b1-a1), (a1 + b1), etc. are averaged and a highly accurate Tp is obtained from the averaged measurement values.

FIG. 4 shows a specific structural example of the ultrasonic device according to one embodiment of the present invention. The array-type ultrasonic wave transmitter / receiver TD is driven by the transmitting unit indicated by TX (transmission repetition period is t0
And). By using the ultrasonic beams B1 and B2 by the transmitter / receiver, the reflected signals from the blood cells in the corresponding region are obtained. From this reflected signal, the Doppler measurement unit DD
Extract the Doppler signal component. Thus obtained,
The Doppler signal at each observation point is stored in the waveform storage units M1 and M2. The signal processing unit SP performs signal processing of the contents described below for the waveforms stored in M1 and M2, and selectively extracts a wideband signal portion such as a region in the acceleration time phase of blood cells. The delay time calculator TE calculates the delay time between signals. The velocity calculation unit VE calculates the velocity v (= L / Tp) from this delay time and the distance L between the observation points obtained by the distance determination unit RE. The configuration of the signal processing unit SP is shown in FIG. 5, where a (t)
And b (t), ca (t) and cb (t) are determined by least-squares approximation. Various configurations of the delay time calculation unit TE are shown in FIG. Figure 6
In (a), the least-squares approximation phase difference functions ca (t), cb (t)
From the phase integrator PI corresponding to the signal waveform h (t)
Signal waveforms ha (t) and hb (t) are obtained, and then the complex correlator C
The time delay is measured by C by complex correlation processing of those signal waveforms ha (t) and hb (t). In FIG. 6 (b), a minimum of 2
The square calculation value QQ of the difference between the functions ca (t) and cb (t) is obtained by the multiplication calculator LSQ, and then the pulse wave propagation time Tp is measured under the condition of minimizing QQ. In FIG. 6C, the average difference calculation unit DM calculates Q from the phase difference functions ca (t) and cb (t).
M is calculated, and the pulse wave propagation time Tp is obtained from the condition that QM is 0.

[0044]

With the ultrasonic device of the present invention, a function that approximates the phase difference of each Doppler signal measured at two points by the least squares method is obtained, and the time delay between these approximate functions is measured to achieve high accuracy. Enables pulse wave velocity measurement.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a time relationship between a measurement position of a reflected signal and a pulse wave according to the present invention.

FIG. 2 is an explanatory diagram showing changes in the position of blood cells measured by the present invention.

FIG. 3 is a diagram illustrating a relationship between a reflected Doppler signal from each blood cell and a received Doppler signal in the present invention.

FIG. 4 is a diagram illustrating a configuration example of an ultrasonic device according to the present invention.

FIG. 5 is a diagram illustrating a configuration example of a signal processing unit of the ultrasonic apparatus of the present invention.

FIG. 6 is a diagram illustrating various configuration examples of a delay time calculation unit of the ultrasonic device of the present invention.

[Explanation of symbols]

TX ... Transmitter, TD ... Linear ultrasonic transducers, B1, B2 ...
Ultrasonic beam, DD ... Doppler measurement unit, M1, M2 ... Waveform storage unit, SP ... Signal processing unit, TE ... Delay time calculation unit, V
E ... Velocity calculation unit, RE ... Distance determination unit, PI ... Phase integrator,
CC ... Complex correlator, LSQ ... Least squares calculator, DM ... Average difference calculator.

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Shoichi Senda 3738-36, 8 wards, Kita-cho, Takamatsu-shi, Kagawa (72) Inventor Hirohide Matsuo 5319-70, Anji-cho, Kida-gun, Kagawa (56) References 6-261899 (JP, A) JP 6226050 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) A61B 8 / 00-8 / 15

Claims (6)

(57) [Claims] 1. An ultrasonic device for transmitting an ultrasonic signal to an inspection target and receiving a reflection signal from the inspection target, wherein the receiver receives Doppler signals of a target object from a plurality of positions of the inspection target. A means for measuring the phase of each received signal by the receiver, a means for measuring each phase difference of the received signals, a means for deriving an approximate phase difference function approximating a time change of the phase difference, A means for measuring the mutual time delay of the approximate phase difference functions, a means for measuring the distance between the two receiving points, and a pulse wave propagation velocity between the two receiving points from the time delay and the distance. An ultrasonic device characterized by measuring. 2. The ultrasonic device according to claim 1, wherein the means for measuring the mutual time delay of the approximate phase difference function and the means for integrating the approximate phase difference function and the cross-correlation of the result of the integration. An ultrasonic device, comprising: a calculating unit for calculating. 3. The ultrasonic device according to claim 1, wherein the means for measuring the mutual time delay of the approximate phase difference functions approximates the difference between the approximate phase difference functions by least square approximation. And ultrasonic equipment. 4. The ultrasonic device according to claim 1, wherein the means for measuring the mutual time delay of the approximate phase difference functions minimizes the average value of the differences of the approximate phase difference functions. Ultrasonic device. 5. The ultrasonic device according to claim 1, wherein the order of the approximate phase difference function is first order. 6. The ultrasonic device according to claim 1, wherein the means for measuring the time delay between the approximate phase difference functions measures the time delay for each heartbeat and averages the measured values by each heartbeat. An ultrasonic device characterized by obtaining a pulse wave propagation velocity from an averaged measured value.