JP2000310621A - Signal sampling device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To well extract the characteristics of an original waveform before re-sampling by suppressing not only the lowering of brightness generated by point sampling but also the lowering of resolving power generated by max. value sampling. SOLUTION: The original signal data row inputted from an A/D converter 2 is scanned by a window. A max. value selecting circuit 6 and a min. value selecting circuit 8 output the max. and min. values of the data value in the window. On the basis of the code of the difference between adjacent data calculated by a subtractor 12, an inflaction point detecting circuit 18 detects the inflection point of the data value. When the inflection point detecting circuit 18 detects the max. and min. values, the max. and min. values from the max. value selecting circuit 6 and the min. value selecting circuit 8 are outputted to an intermediate signal data row. In a case other than this, the data value in the window at a predetermined position is outputted to the intermediate signal data row. The intermediate signal data row is re-sampled at a constant interval to obtain a sampling result wherein the max. and min. values are necessarily caught.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal sampling apparatus for thinning out and sampling an original signal data sequence and generating a data sequence with a reduced number of data.

[0002]

2. Description of the Related Art For example, in an ultrasonic diagnostic apparatus, an ultrasonic wave is transmitted to a subject and a tomographic image is generated based on the intensity of reflected waves from each point of the subject. For example, by emitting one ultrasonic pulse from the probe, an A-mode signal including a reflection signal from each part of the subject on a line scanned by the ultrasonic beam is generated. The A-mode signal is represented as a single linear image on the screen, and each point on the line is displayed with a luminance corresponding to the amplitude of the reflected signal. What is called a B-mode image is a tomographic image obtained by sequentially and gradually moving the position and angle of the probe.
The B-mode image is a two-dimensional image having an axis in the ultrasonic wave transmission / reception direction and an axis in the moving direction of the probe included in the A mode.

[0003] The reflected signal obtained as an analog signal is
The signal is converted into a digital signal by an A / D (analog to digital) conversion circuit, and signal processing is performed using the digital signal. The sampling in this A / D conversion is performed at a relatively high frequency corresponding to the baseband frequency of the transmitted ultrasonic wave. However, the spatial frequency corresponding to the pixel resolution of the image display device is lower than the sampling frequency that can be performed by the A / D conversion circuit. Therefore, A /
When trying to display an image of a signal sampled at a high frequency by the D conversion circuit, it is necessary to resample at a frequency corresponding to the number of display pixels of the display device.

In a conventional method, a reflected signal of an ultrasonic wave is passed through an anti-aliasing filter having a frequency corresponding to the number of display pixels as a Nyquist frequency before A / D conversion. Sampling processing is performed. This sampling method includes point sampling and maximum value sampling. FIG.
FIG. 2 is a schematic diagram showing an example of sampling by these conventional methods. FIG. 3A shows a data string input from the A / D conversion circuit to a circuit that performs resampling. The horizontal axis is a time axis, and the vertical direction corresponds to a data value. FIGS. 7B and 7C are diagrams respectively showing the results of point sampling and maximum value sampling for the input data. In this example, the resampling is performed every time t ₃ after 4 data.

In the point sampling shown in FIG. 1B, each time t _i, which is the output timing of resampling, is used.
Is output as a result value of the resampling. On the other hand, in the maximum value sampling shown in FIG. 9C, the maximum value of the input data after the previous resampling timing, that is, t _i-1 <t ≦ t _{i is} obtained as the resampling result value at each time t _i . The maximum value of the input data in the time axis range is output. These resampled data are held as values representing the reflection signal until the next resampling timing. By this resampling process, the input data sequence is thinned out according to the resolution of the display device, and a general change in the reflected signal is extracted.

[0006]

All of the above-mentioned resampling methods have a problem that the deviation between the original reflected signal waveform and the resampled reflected signal waveform becomes large.

First, in the point sampling method in which the band is limited at the Nyquist frequency and resampling is performed at regular intervals, input data during the resampling timing is discarded. That is, fine information obtained by transmission and reception of ultrasonic waves is not effectively used.
Specifically, if input data such as the maximum value or the minimum value is located between resampling timings, there is a problem that important input data representing such a characteristic of the reflected signal is not sampled. So, for example,
When a tomographic image is generated by converting a reflected signal into luminance, the maximum value is not sampled, causing a decrease in luminance of the image, resulting in an inconvenient image.

On the other hand, in the maximum value sampling method, the maximum value of the input data at each resampling interval is held, so that the luminance of the image does not decrease. However, as shown in FIG. It causes towing. This tailing occurs because the timing at which the data reaches the maximum value and the output timing thereof are not synchronized, and thus the occurrence time difference between adjacent resampled data and its output interval fluctuate. This phenomenon causes a problem that the resolution in the sampling direction is reduced and the apparent cutoff frequency becomes lower than the original Nyquist frequency.

FIGS. 7 and 8 are explanatory views for understanding the problem of the conventional resampling process from another aspect. FIG. 7 shows a state where a sine wave is point-sampled at three different phases. FIG. FIG.
FIG. 8 is a schematic diagram showing a state where a sine wave is sampled at the maximum value at three different phases as in FIG. 7. FIG.
(A) and FIG. 8 (a) show that the resampling timing is
7 (b) and 8 (b) when the sine wave phase is positioned at θ = 90 ° and 270 °, FIGS. 7 (c) and 8 (c) when θ = 0 ° and 180 ° ) Is θ = 45 °, 225 °
Is the case.

FIG. 7A shows the point sampling.
In this case, the maximum and the minimum of the sine wave are captured and there is no problem. However, the phase shifts from this position,
For example, at the position shown in FIG. 3C, the amplitude of the sine wave is underestimated and the contrast is reduced. The case shown in FIG. 3B is an extreme case, in which no amplitude appears and the signal becomes flat.

Regarding the maximum value sampling, the maximum value of the sine wave can be correctly detected at any phase, but the amplitude becomes small. For example, when the resampling cycle and the sine wave cycle are the same as shown here, the amplitude after resampling is at most の of the original sine wave (FIG. 8).
(B), and in an extreme case, FIG.
As shown in (1), the amplitude does not appear and the signal becomes flat.

From these facts, it is understood that in the conventional resampling process, the amplitude of the hold value greatly differs depending on the resampling phase even for the same input data sequence. Such phenomena occur more or less not only in sine waves but also in general waveforms.
Depending on the sampling phase, the signal waveform after resampling is relatively affected. Therefore, there has been a problem that information obtained by the ultrasonic probe is difficult to be faithfully reproduced on the screen.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and it is possible to accurately detect the maximum and minimum of an original waveform, suppress a decrease in luminance caused by point sampling, and reduce a signal near the Nyquist frequency. The object of the present invention is to provide a signal sampling device capable of stably capturing regardless of the phase of resampling, suppressing degradation in resolution caused by maximum value sampling, and satisfactorily extracting characteristics of an original waveform before resampling. And

[0014]

A signal sampling apparatus according to the present invention is a signal sampling apparatus for dividing an original signal into a plurality of sections and obtaining a sampling value for each of the sections. It has an extreme value detecting means for detecting, and a sampling value determining means for determining a value based on the extreme value as the sampling value for the section having the extreme value.

According to the present invention, the extreme value detecting means checks whether or not the original signal has a maximum or a minimum in each section. If there is a local maximum point or a local minimum point in the section, the sampling value determining means determines the value of the original signal at any of those points,
That is, the sampling value corresponding to the section is determined and output based on the maximum value or the minimum value. Therefore, a value reflecting the maximum value and the minimum value of the original signal is captured in the sampling, so that the contrast does not decrease. Also,
The maximum and minimum values of the original signal near the Nyquist frequency are detected, so that a sampling data sequence fluctuating at the same frequency as the original signal is obtained. Here, the original signal may be an analog signal or a digital data string. In addition, the sampling value corresponding to the section having the extreme value can be the value itself when there is a single extreme value included in the section, and can be calculated based on a predetermined rule from the extreme value. You can also. Further, when a plurality of extreme values, that is, a local maximum value and a local minimum value, simultaneously exist in the section, a predetermined one can be simply selected as a sampling value, and the signal value at both ends of the section or the adjacent one can be selected. Based on a predetermined criterion that also takes into account the signal waveform of the section, it is also possible to determine which more appropriately represents the characteristics of the original signal, and select the suitable extreme value as the sampling value.

In another aspect of the signal sampling apparatus according to the present invention, the sampling value determining means determines, for the section having no extreme value, the value of the original signal at a predetermined position in the section. It is characterized by being defined as.

According to the present invention, for an interval in which no extreme value exists, the sampling value determining means determines the value of the original signal at a predetermined position in the interval as the sampling value. By performing sampling at a fixed position in each section, the relationship between the time interval between adjacent sampling values on the original signal and their output timing difference is kept constant unlike the maximum value sampling. That is, in the present invention, basically, similarly to the point sampling, the time at which the original signal is sampled and the timing at which it is output are synchronized, so that a reduction in resolution such as tailing is suppressed.

In a preferred aspect of the present invention, the sampling value determining means determines, as the sampling value, a value of the original signal at a predetermined position in the section having the plurality of extreme values. .

Still another preferred aspect of the present invention is the signal sampling device, wherein the predetermined position is one end of the section. One end of the section is preferable in that it is specified by processing that is easier than other positions. Another preferred embodiment of the present invention is the signal sampling device, wherein the predetermined position is a center of the section. In this aspect, the difference between the position of the sampled original signal data in the case where the extremum does not exist in the section and the position in the section where the extremum exists does not exceed half the section width. Is suppressed.

Another preferred embodiment of the present invention is a signal sampling device, wherein the sampling value determining means determines the extreme value as the sampling value for the section having a single extreme value. . Another preferred embodiment of the present invention is a signal sampling device, wherein the extreme value is a maximum value.

Another signal sampling apparatus according to the present invention is a signal sampling apparatus which divides an original signal into a plurality of sections having a fixed width and obtains a sampling value for each section, and detects an extreme value of the original signal. Extremum detection means, signal conversion means for generating an intermediate signal by deforming the original signal based on the extremum, and sampling the intermediate signal at intervals of the constant width to obtain a sampling value for each section. And a sampling value determining means for determining.

In the above-mentioned conventional sampling method, an extreme value is appropriately reflected on a waveform after sampling or, on the contrary, a signal represented by a sampling value sequence in which no extreme value is reflected on a sampling value in accordance with a sampling phase. The problem is that the difference between the waveform and the original signal becomes large. According to the present invention, an intermediate signal is generated by transforming an original signal based on an extreme value. This intermediate signal can be configured such that the extreme value is reflected in the sampling value regardless of the sampling phase. Therefore, even if the sampling value determination means simply performs sampling at a constant interval, the extreme value in the section is reflected on the sampling value.

In a preferred aspect of the present invention, the intermediate signal is:
The signal sampling device is characterized in that the signal value is maintained at the extreme value in a range having the constant width and including the position of the extreme value.

In another preferred aspect of the present invention, there is provided a window setting means for moving a window having a width corresponding to the section by the original signal, wherein the extreme value detecting means is provided in the window at each window position. The signal conversion means detects the extreme value of the original signal included in the above, when the extreme value is detected, sets the extreme value as a representative value at the window position, and when the extreme value is not detected, A signal sampling apparatus, wherein an original signal value at a predetermined position in the window is set as a representative value, and the intermediate signal is configured by the representative value.

According to another preferred aspect of the present invention, the extreme value detecting means detects a change rate of the original signal within the window, and a change rate detecting means for detecting a change rate of the original signal in the window. An extreme value position determining means for determining the position of the extreme value based on the signal value.

[0026]

[Embodiment 1] Next, an embodiment of the present invention will be described with reference to the drawings. The embodiment of the present invention described here is an ultrasonic diagnostic apparatus having a function of displaying a B-mode image. The ultrasonic diagnostic apparatus scans a subject with ultrasonic waves emitted from an ultrasonic probe, and the reflected signals of the ultrasonic waves are converted into luminance of a display device and displayed as an image.

FIG. 1 is a block diagram showing a main circuit configuration of the ultrasonic diagnostic apparatus according to the embodiment of the present invention.
A reflected signal 1 which is an analog signal output from an ultrasonic probe (not shown) is digitally sampled by an A / D converter 2. From the A / D converter 2, the reflected signal 1
Is output as time-series digital data representing the change in the amplitude value of. The sampling in the A / D converter 2 is performed at a relatively fast frequency depending on the transmission frequency of the ultrasonic wave.

The signal sampling processor 4 according to the present invention
Is input with the digital data sequence generated by the A / D converter 2 as an original signal. This original signal data string is divided into a maximum value selection circuit 6, a minimum value selection circuit 8, a data delay circuit 10,
The signals are input to the subtractor 12 and the selection circuit 14, respectively. Window width data is input to the maximum value selection circuit 6 and the minimum value selection circuit 8. This window width is a parameter that determines the resampling cycle. That is, as will be described later, the original signal data for each window width is output as a resampling value in principle, so that, for example, the screen scanning rate of the image display device (not shown) of the ultrasonic diagnostic apparatus and the pixel resolution thereof The window width is determined according to.

The maximum value selection circuit 6 is a circuit for outputting to the selection circuit 14 the maximum value from the original signal data traced back by the window width to the latest original signal data. On the other hand, the minimum value selection circuit 8 selects the minimum value from the original signal data traced back by the window width to the latest original signal data.
4 is a circuit for outputting the result.

The data delay circuit 10 is a circuit capable of delaying the input original signal data by one cycle of the sampling clock of the A / D converter 2 and outputting it. The delayed data output from the data delay circuit 10 is input to a subtractor 12.

The subtracter 12 receives the latest original signal data input without delay and the 1 input from the data delay circuit 10.
The difference from the previous original signal data is obtained, and sign data representing the sign of the difference is output to the shift register 16. The inflection point detection circuit 18 monitors the arrangement of the code data in the shift register 16 and detects whether or not a change in the polarity of the change rate (inflection point) exists in the original signal data sequence cut out according to the window width. I do.

When the inflection point detection circuit 18 detects the inflection point from the increase of the original signal data to the decrease, it instructs the selection circuit 14 to select the maximum value data from the maximum value selection circuit 6. When the inflection point from the decrease to the increase is detected, an instruction to select the minimum value data from the minimum value selection circuit 8 is issued. If no inflection point is detected, an instruction to select the latest original signal data is issued.

In response to this control signal, the selection circuit 14
One of three data, the latest original signal data input from the / D converter 2 and the data respectively input from the maximum value selection circuit 6 and the minimum value selection circuit 8, is selected and output. The data string output from the selection circuit 14 is A
The rate is the same as the output of the / D converter 2. The sampling circuit 20 is a circuit for resampling the high-speed data string from the selection circuit 14 at a cycle corresponding to the window width.

Next, an operation example of the signal sampling processing section 4 will be described with reference to the drawings. FIG. 2 is a schematic diagram illustrating a process of generating an intermediate signal data sequence which is an output of the selection circuit 14 from the original signal data sequence. 3A to 3K, the horizontal axis represents time t (the right direction is positive), and the vertical axis represents the data value D.
(Upward positive) axis. FIG. 7A shows an original signal data sequence, and FIG. 7K shows an intermediate signal data sequence. Black circles represent data. Hereinafter, the original signal data at time t,
The intermediate signal data is represented by D _O (t) and D _P (t), respectively. FIGS. 7B to 7J show data cut out by windows (the window width is 5 data in this case) as black circles.

Here, a case where the resampling interval of the signal sampling processing unit 4 is once for four data of the original signal data sequence will be described. That is, the signal sampling processing section 4 divides the original signal data into four sections and outputs one sampling value for each section.

In this apparatus, a difference between adjacent data is obtained to detect an extreme value. Due to this, the window width, which is the number of data used for processing, is set to be one larger than the section width. That is, "5" is designated as the window width here in correspondence with the resampling section width "4". Based on this window width, the maximum value selection circuit 6 and the minimum value selection circuit 8 target the past five data including the latest data of the original signal data string for the maximum value selection processing or the minimum value selection processing.

In addition, the past 5 including the latest original signal data
There are four code data generated from the data. Therefore, the inflection point detection circuit 18 sets the window width to “5”.
Is specified, the past four are monitored including the latest one of the code data held in the shift register 16. Here, the sign data is a value "1" when the subtraction result in the subtractor 12 is positive, that is, when the rate of change of the original signal data is positive (+), and conversely, the subtraction result is negative, that is, the original signal data If the rate of change is negative (-), it is defined as a value "0". Under this definition, each code data is represented by one bit. Shift register 16 when a new code data is input to left shift the storage contents of the existing stores latest code data to the least significant bit b _0. The inflection point detection circuit 18 obtains a 4-bit code data string “b ₃ b ₂ b ₁ b ₀ ” including the latest code data by reading the lower 4 bits of the shift register 16. Then, the inflection point detection circuit 18 detects an inflection point in the window based on the bit pattern.

For example, if the code data string is "1111",
When "0000", the original signal data is monotonically increasing and monotonically decreasing within the window, and has no inflection point. On the other hand, if the code data string is “1110”, “110”
In the case of "0" or "1000", there is only one inflection point in which the rate of change changes from positive to negative, that is, only one maximum point.
In the case of “0011” or “0111”, there is only one inflection point at which the rate of change changes from negative to positive, that is, only one minimum point. The inflection point detection circuit 18 controls the selection circuit 14 based on the detection result of the inflection point as described above.

In the case of other bit patterns, such as "0110", there are a plurality of inflection points in the window. In this case, the present apparatus needs a weighted judgment criterion as to which inflection point should be selected and sampled. Whether or not to perform processing for determining such a weighting criterion to capture an extreme value can be arbitrarily selected depending on the application. In general, resampling will not be performed at frequencies where such inflection points occur frequently within one window. In consideration of this, in the present apparatus, when a plurality of inflection points are detected, they are treated in the same manner as when there is no inflection point, thereby reducing the processing load and simplifying the circuit configuration.

[0040] Now, at time t _5, the window original signal data at time t ₁ ~t ₅ is cut (Fig. 2
(B)). The maximum value selection circuit 6 and the minimum value selection circuit 8 respectively provide a maximum value D _O (t ₅ ) and a minimum value D _O (t ₅ ) in this window.
₁ ) is output. The data delay circuit 10 outputs original signal data that is one older than the latest original signal data, that is, D _O (t ₄ ). The subtractor 12 calculates “D _O (t ₅ ) −D _O (t ₄ )”,
New code data is generated and output to the shift register 16. At this point, since the code data string is “1111”, the inflection point detection circuit 18 determines that there is no inflection point. The inflection point detection circuit 18 controls the selection circuit 14 to select the latest original signal data.
_O (t ₅ ) is output as intermediate signal data D _P (t ₅ ).

At time t ₆ , no inflection point exists in the window (FIG. 2C). Therefore, the latest original signal data D _O (t ₆ ) is selected by the same operation as at time t _5, and is output as intermediate signal data D _P (t ₆ ).

At time t ₇ , the original signal data at times t _{3 to} t ₇ is cut out in the window (FIG. 2).
(D)). In this case, a value smaller than the immediately preceding D _O (t ₆ ) is input as the latest original signal data D _O (t ₇ ). Therefore, the output of the subtractor 12 changes to “0”, and the code data string output to the inflection point detection circuit 18 becomes “1110”. The inflection point detection circuit 18 is
Determines that there is a local maximum in the window. Then, the inflection point detection circuit 18 controls the selection circuit 14 so as to select the output from the maximum value selection circuit 6. The maximum value selection circuit 6 outputs the maximum value D _O (t ₆ ) at times t _{3 to} t ₇ . Therefore, the selection circuit 14 outputs this D _O (t ₆ ) to the intermediate signal data D _P
Output as (t ₇ ).

At times t ₈ and t ₉ shown in FIGS. 2E and 2F, code data “0” is sequentially input to the shift register 16 by newly input original signal data. As a result, the code data strings are respectively “110”.
0 "," it will be changed to 1000 ", both containing inflection point from positive to negative. In other words, state that the maximum point is present in the window has not changed from the time t _7. Thus, these times In, the output from the maximum value selection circuit 6 is selected by the same operation as at time t _{7. The} maximum value selection circuit 6 continuously outputs D _O (t ₆ ) after time t _7. As a result, D _O (t ₆ ) is continuously output as the intermediate signal data D _P (t ₈ ) and D _P (t ₉ ).

Times t _{10 to} t ₁₂ shown in FIGS.
Since the code data string is "0000", the inflection point detection circuit 18 determines that there is no inflection point in the window. The selection circuit 14 selects the latest original signal data based on the control signal from the inflection point detection circuit 18,
Output. Therefore, the intermediate signal data D _P (t ₁₀ ), D _P (t ₁₁ ),
D _O (t ₁₀ ), D _O (t ₁₁ ), and D _O (t ₁₂ ) are sequentially output as D _P (t ₁₂ ).

[0045] At time t _13, the window original signal data at time t ₉ ~t ₁₃ is cut out (FIG. 2
(J)). In this case, a value larger than the immediately preceding D _O (t ₁₂ ) is input as the latest original signal data D _O (t ₁₃ ). Therefore, the output of the subtractor 12 changes from “0” to “1”, and the code data string output to the inflection point detection circuit 18 becomes “0001”. Based on this code data string, the inflection point detection circuit 18 determines that a minimum point exists in the window. The inflection point detection circuit 18 selects the output from the minimum value selection circuit 8 so as to select the output.
Control. The minimum value selection circuit 8 outputs the minimum value D _O (t ₁₂ ) at times t _{9 to} t ₁₃ . Therefore, this selection circuit 14 outputs this D _O (t ₁₂ ) as intermediate signal data D _P (t ₁₃ ).

FIG. 2K shows an intermediate signal data sequence generated by the above processing. The intermediate signal data is held at the maximum value D _O (t ₆ ) of the original signal data over three data after time t ₇ when the maximum of the original signal data sequence is detected. Accordingly, D _O (t ₆ ) is thereby held over four data equal to the sampling section width including the time t ₆ . The intermediate signal data Similarly, the minimum of the original signal data sequence is detected at time t _{1 3,} it is held in the minimum value D _O of the original signal data (t _12).

FIG. 3 is a schematic diagram illustrating the process of generating a resampled data sequence from an intermediate signal data sequence. 2A to 2F, similarly to FIG. 2, the horizontal axis corresponds to time t, and the vertical axis corresponds to data value D. FIG. 7A shows an original signal data string, which is shown for comparison. FIG. 7B shows an intermediate signal data sequence generated by the above-described processing. When the intermediate signal data sequence is resampled for each section width 4, there are four sampling phases. FIGS. 8C to 8F show the four resampling results. In the case of (c) sampling D _P (t _i ) (i = 4n + 1), where n is an integer, the drawing (d) samples D _P (t _i ) (i = 4n). In this case, FIG. 11E shows D _P (t _i ) (i = 4n−
In the case of sampling 1), FIG. 11F shows D _P (t _i )
(I = 4n-2) is sampled.

As described above, when a plurality of inflection points do not exist within one window width, the intermediate signal data sequence is based on the waveform of the original signal data sequence, but in the vicinity of the extremum, The waveform is deformed such that the extremum continues for a number equal to the resampling interval. Therefore, no matter what phase the intermediate signal data sequence is sampled, the extreme value is always sampled. For example, in the phase shown in FIG. 3C, the resampled intermediate signal data D
_P (t ₉ ) has the maximum value D _O (t ₆ ) of the original signal data, and similarly in FIGS. 3D to 3 F, the resampled intermediate signal data D _P (t ₈ ) and D _P (t ₈ ), respectively. _P (t ₇ ) and D _P (t ₆ ) have the maximum value D _O (t ₆ ) of the original signal data.

Therefore, the sampling circuit 20 can adopt a simple circuit configuration for simply sampling every four data.

FIG. 4 shows a waveform of an original signal data sequence and an output waveform obtained by each sampling method in order to compare the resampling result of the present apparatus with the results of conventional point sampling and maximum value sampling. FIG. FIGS. 4A to 4C are the same as FIGS. 6A to 6C used in the description of the related art. FIG.
(D) shows the result of the inflection point detection sampling by this device. The output of the apparatus, it is possible to capture the maximum value at time t _9, to have no stringing tails such as maximum sampling will be appreciated from FIG.

The tailing will be discussed in more detail. The resampled data is denoted as D _R (t). Here, it is noted that t represents the timing at which the resampled data is output, and not the timing of occurrence in the original signal data. For maximum value sampling, for example,
D _R (t ₇ ) = D _O (t ₇ ), D _R (t ₁₁ ) = D _O (t ₉ ), D _R (t ₁₅ ) = D _O
(t ₁₁ ). Comparing the output timing of the sampled data with the generation timing of the data, the maximum value sampling does not deviate from the output timing at the front end of the waveform, while the output timing occurs at the rear end. It may be delayed by the same amount as the resampling interval than the timing. This is recognized as a phenomenon called tailing.

On the other hand, according to the sampling of the present apparatus, for example, D _R (t ₇ ) = D _O (t ₇ ), D _R (t ₁₁ ) = D _O (t ₉ ), D
_R (t ₁₅ ) = D _O (t ₁₅ ). As described above, in the sampling by the present apparatus, the output timing of the sampled data and the generation timing of the data are basically synchronized at both the front end and the rear end of the waveform. That is, there is no difference between the output timing and the generation timing at both positions. This means that the sampling of this device does not cause tailing.

FIG. 5 is a schematic diagram showing a state in which a sine wave is sampled at three different phases by the present apparatus.
This corresponds to FIGS. 7 and 8 relating to the conventional sampling, and FIGS. 5A to 5C respectively show that the resampling timing is the sine wave phase θ = 90 °, 2
70 °, θ = 0 °, 180 °, θ
= 45 ° and 225 °. As shown in this figure, according to the present apparatus, the maximum value and the minimum value are correctly captured in any resampling phase, and the amplitude of the waveform after resampling is equal to the waveform of the original signal.
Further, the fluctuation of the data value having the same cycle as the original signal is always reproduced irrespective of the sampling phase, and the original signal having a frequency as close as possible to the Nyquist frequency can be suitably sampled.

[Second Embodiment] In the above-described apparatus, the inflection point detecting circuit 18 is configured to detect both the inflection point at which the original signal data changes from increase to decrease and the inflection point at which the original signal data changes from decrease to increase. Correspondingly, in the intermediate signal data sequence, both the maximum value and the minimum value of the original signal data sequence are held over the sampling section width.

On the other hand, the inflection point detection circuit 18 may be configured to detect only one of the inflection points. For example, when the inflection point detection circuit 18 is configured to detect only the inflection point where the original signal data changes from increasing to decreasing, only the maximum value is held in the intermediate signal data sequence. The maximum value is always captured in the result of resampling this intermediate signal data sequence. On the other hand, in this configuration, there is no guarantee that the minimum value is always captured. That is, the effect of raising the resampling value as a whole is generated. This means that the amplitude is underestimated, but when applied to an ultrasonic image, the contrast and frequency characteristics are maintained at an appropriate level, there is no sampling omission of the maximum value, and the brightness is raised to a brighter direction. The resulting image is obtained.

Here, the ultrasonic image expresses the shape of a structure by a group of small points having different luminances. In the image, within a region of relatively uniform brightness,
When there are a bright point and a dark point, the dark point is visually more prominent than the bright point, even if the luminance difference between them is almost the same. Such a conspicuous dark spot appears as noise on the image observer's eyes, and the S / N
Gives the impression that the ratio is reduced.

According to the present apparatus, since a decrease in the resampling value corresponding to such a point of low luminance is suppressed, noise is made inconspicuous, and the image quality of an ultrasonic image is improved.

On the other hand, in a configuration in which only the inflection point from the decrease of the original signal data sequence to the increase is detected, the minimum value is always sampled, but the maximum value is not guaranteed to be sampled. Such sampling may also be suitable depending on the application.

[Other Embodiments] In the above configuration, the latest original signal data is sampled unless there is only one extremum in the resampling section. That is, the trailing end of the resampling section is sampled in principle, and only when there is only one extremum in the section, the sampling position is changed to the position of the extremum in the section. However, the sampling position in principle need not be limited to the rear end of the section, but may be any predetermined position in the section.

In the maximum value sampling, when an extreme value exists, the sampling position is not only a point in the middle of the section, but is sampled at the end of the section in the case of monotonic increase and in the case of monotone decrease. It will be sampled at the front end. Such a variation in the sampling position causes tailing. Therefore, tailing is prevented by keeping the sampling interval unchanged. The present invention realizes sampling at a constant interval by sampling at a predetermined position in a sampling section in principle,
Thereby, tailing can be prevented. That is, in order to prevent tailing, it is sufficient that the sampling position is a fixed position in the section in principle, and it does not matter whether or not the sampling position is an end.

In the above-described configuration, a window that moves on the original signal data sequence is introduced. In this configuration, an intermediate signal data sequence having the same rate as the original signal data sequence was once generated, and the data sequence was resampled. However, a configuration that does not use a window that moves one data at a time is also possible. For example, a configuration is also possible in which the original signal data sequence is divided into predetermined sampling sections, and one sampling value is directly selected from each section. In this configuration, when it is detected that an extreme value exists in the section, the extreme value is sampled, and when no extreme value exists, the original signal data at a fixed position in the section is sampled.

[0062]

According to the signal sampling apparatus of the present invention, it is detected whether or not the original signal has an extreme value within the sampling interval, and if it exists, the value based on the extreme value is sampled. . As a result, sampling is performed in which the extreme value of the original signal is reflected, and an effect that the contrast does not decrease is obtained. In addition, since the maximum and minimum values of the original signal near the Nyquist frequency are captured, a sampling data sequence that fluctuates at the same frequency as the original signal is obtained, and the cutoff frequency can be raised to the Nyquist frequency. The effect is also obtained.

According to the signal sampling apparatus of the present invention, in a sampling section having no extreme value, the value of the original signal at a predetermined position in the section is sampled, thereby preventing a phenomenon such as tailing. The effect of preventing a decrease in resolution can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a main circuit configuration of the ultrasonic diagnostic apparatus according to an embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating a process of generating an intermediate signal data sequence from an original signal data sequence.

FIG. 3 is a schematic diagram illustrating a process of generating a resampled data sequence from an intermediate signal data sequence.

FIG. 4 is a schematic diagram showing a waveform of an original signal data sequence, a conventional sampling waveform, and a sampling waveform of the present apparatus.

FIG. 5 is a schematic diagram showing a state where a sine wave is sampled at three different phases by the present apparatus.

FIG. 6 is a schematic diagram showing an example of sampling by a conventional method.

FIG. 7 is a schematic diagram showing a state where a sine wave is point-sampled at three different phases.

FIG. 8 is a schematic diagram showing a state in which a sine wave is sampled at a maximum value at three different phases.

[Explanation of symbols]

2 A / D converter, 4 signal sampling processor, 6
Maximum value selection circuit, 8 minimum value selection circuit, 10 data delay circuit, 12 subtractor, 14 selection circuit, 16 shift register, 18 inflection point detection circuit, 20 sampling circuit.

Claims (11)

[Claims] 1. A signal sampling device for dividing an original signal into a plurality of sections and obtaining a sampling value for each section, comprising: an extreme value detecting means for detecting an extreme value of the original signal; A sampling value determination unit that determines a value based on the extreme value as the sampling value for the section. 2. The signal sampling device according to claim 1, wherein the sampling value determining unit sets a value of the original signal at a predetermined position in the section as the sampling value for the section having no extreme value. A signal sampling device. 3. The signal sampling apparatus according to claim 1, wherein the sampling value determination unit sets, as the sampling value, a value of the original signal at a predetermined position in the section with respect to the section having a plurality of the extreme values. A signal sampling device. 4. The signal sampling device according to claim 2, wherein the predetermined position is one end of the section. 5. The signal sampling device according to claim 2, wherein the predetermined position is a center of the section. 6. The signal sampling device according to claim 1, wherein the sampling value determination unit determines the extreme value as the sampling value for the section having a single extreme value. A signal sampling device. 7. The signal sampling device according to claim 1, wherein the extremum is a maximal value. 8. A signal sampling apparatus which divides an original signal into a plurality of sections having a constant width and obtains a sampling value for each section, an extreme value detecting means for detecting an extreme value of the original signal; Signal conversion means for generating an intermediate signal by deforming the original signal based on the following, and a sampling value determination means for sampling the intermediate signal at intervals of the constant width and determining a sampling value for each section. A signal sampling device comprising: 9. The signal sampling apparatus according to claim 8, wherein the intermediate signal has the constant width and the signal value is maintained at the extreme value in a range including the position of the extreme value. Characteristic signal sampling device. 10. The signal sampling device according to claim 8, further comprising window setting means for moving a window having a width corresponding to the section by the original signal, wherein the extreme value detecting means sets the window at each window position. Detecting an extremum of the original signal included in a window, the signal converting unit sets the extremum as a representative value at the window position when the extremum is detected, and when the extremum is not detected. Wherein the original signal value at a predetermined position in the window is set as a representative value, and the intermediate signal is constituted by the representative value. 11. The signal sampling apparatus according to claim 10, wherein the extreme value detecting means detects a change rate of the original signal in the window, and a polarity of the change rate changes. An extreme value position determining means for determining a position of the extreme value based on a position.