JPH0131155B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to ultrasound imaging systems, and more particularly to an ultrasound imaging method using corrected wavefront reprojection.

In recent years, a method called aperture synthesis has come to be applied to medical ultrasound. The basis of this method is not only the simultaneous reception of echo waveforms, A/D conversion, and temporary memory, but also the acoustic geometry of the reprojection surface during reconstruction (whether computer-aided or not). It is essential that the In conventional imaging methods of this type, the set sound speed during reconstruction (by back projection) is controlled with a single parameter to maximize the focus and visibility of the image reflected on the screen. However, this method has the disadvantage that it is insufficient to obtain an image that is suitable for practical use.

The problem is what the average value of the sound speed was in the space being imaged, and how it was distributed locally.Therefore, if we controlled the wavefront so that it matched on one side, it would not match on the other. It can be broadly divided into two points: whether a phenomenon exists or not, and the extent to which it has an adverse effect.

However, there is currently no good solution to the latter point. Therefore, it is necessary to completely correct at least the former point, and while including the latter, it is necessary to make sure that the part you want to observe in particular has the best focus, and if possible, it should be done automatically. It is hoped that this will be done.

The present invention has been made in view of these points, and an object thereof is to provide an aperture-type ultrasonic imaging method that automatically focuses on a portion to be observed.

In order to achieve this object, the present invention first determines the overall sound velocity setting value so that the sharpness of the image within the small area D of interest is maximized, and then first obtains an image within the small area D using a partial aperture. A sound velocity correction value related to the sub-aperture is determined so as to have the maximum correlation value with the image in the small area, and the sub-aperture of interest is moved and resized so as to cover the entire aperture. The present invention is characterized in that the sound speed correction value corresponding to each part of the entire aperture is determined.

Hereinafter, the present invention will be explained in detail with reference to the drawings.

First, the principle of the present invention will be explained with reference to FIG.

(1) First, focus on the one-dimensional image along the line A-A' that crosses the point of interest TA(D) (Figure 1).
Although the line A-A' is not necessarily parallel to the aperture plane of the array transducer TD,
A direction parallel to the sound ray is not preferred because the detection effect is poor.

(2) Determine the amount of correction for the sound speed setting value so that the focus of the image due to the entire aperture on line A-A' becomes sharpest.

The following method is used to determine whether or not the focus has become sharpest.

(b) Detect and determine whether the peak value at the location where the brightness peaks has reached its maximum.

(b) Or, find and determine the point where the ratio of high to mid spatial frequencies is maximum.

(3) Next, line A in the image that has converged
Using the one-dimensional image along A' as a template, for each of the partial apertures C _A , C _B ,..., C _D ,
The same part in the image it gives (A-A')
The sound velocity in the fan-shaped section from the small region D to each partial aperture is corrected so that the correlation value with the one-dimensional image of is maximized.

In calculation processing, this correction is to correct the time axis scale factor of the signal from each element of the partial aperture (stored in a memory not shown), and the distance of the small area D to the aperture is If it is not too large compared to , it is equivalent to searching where the echoes before and after the point of interest TA(D) are on the time axis. Instead of changing the scale factor, it is almost equivalent to simply shift the time axis.

(4) Next, recreate the image for the entire aperture C with the sound speed or time axis shift corrected for C _A , C _B , ..., C _D .

It should be noted that in this case, the time axis shift also corrects the element position error.

By the above procedure, a sharply focused image can be obtained, but in the human body, the almost uniform sound speed C and density P are only slightly perturbed by a ^few percent at worst. If the aperture is made up of elements, by dividing it into subgroups of 2 ^n/2 elements and making partial corrections 2 ^n/2 times, it is possible to obtain a sharply focused image that is practically usable.

Even if the same process was performed using 2 ^n/2 or more finely divided elements, the image formed by each group and the image obtained by contributing the overall aperture would be too different in degree of blur ( That is, the difference in resolution obtained is too large), making it difficult to effectively discuss the correlation of line images regarding line A-A'.

Note that the above-described method for determining whether or not the image is in focus may be determined by focusing on the sharpness of the peak at the origin of autocorrelation as one method of focusing on the frequency spectrum of the image. In other words, focus on the peak waveform at the origin of the autocorrelation waveform as shown in Figure 2, and perform focusing operations.If it is wide as shown in Figure 3, it is out of focus, and there is a certain limit. If it becomes narrower than the value W, it may be determined that the focus is in focus.

Next, the ultrasonic imaging method of the present invention will be explained. FIG. 4 is a block diagram showing a receiving system of an ultrasonic imaging apparatus for implementing the ultrasonic imaging method of the present invention (the transmitting system is omitted). In Figure 4, 1 is the target object, 2 ₁ ...
2 _o is a transducer array that receives echo signals from the target object 1, 3 ₁ ... 3 _o is a first-stage amplifier array that amplifies echo signals from individual transducers, and 4 is the output of the first-stage amplifier array 3 ₁ ..., 3 _o. 5 is an auxiliary variable delay line array, and 6 is a beamformer. 7 is a signal processing unit that logarithmically amplifies the signal from the beamformer 6, performs analog-to-digital conversion, etc., and outputs the signal; 8 is a large number of original image memories that store original image data for focusing; 9 is an image processing unit;
10 is an input/output port, 11 is a display image memory for storing image data, and 12 is a comparison/calculation unit.
A display section consisting of a CRT, its peripheral circuits, etc., 13 a controller for controlling scanning operations, 14 a controller for controlling focusing, and 15 a keyboard. During the scan operation, the controller 13 controls the beam former 6, image processing/comparison/
Control signals are directly exchanged with the arithmetic unit 9, the display image memory 11, etc. (signals are exchanged with the solid signal line), and the switch array 4 and the auxiliary variable delay line array 5 are used as necessary. , the signal processing unit 7, the original image memory 8, etc. (signals are exchanged via dotted signal lines). During the focusing operation, the controller 14 controls the switch array 4, the auxiliary variable delay line array 5, the beamformer 6, the original image memory 8, the image processing/comparison/calculation unit 9, etc. under the control of the controller 13. Between the
Control signals are directly sent and received (signals are sent and received through the solid signal line).

In the above configuration, scanning is performed according to a control signal from the controller 13 (the scanning itself is performed by a known method). Image data obtained by scanning is captured into a display image memory 11 and then sent to a display section 12. This causes the image to be displayed on the CRT.
On the other hand, in focusing mode, controller 1
According to the control signal from switch array 4,
The auxiliary variable delay line 5, the beam former 6, the original image memory 8, etc. operate to execute the procedure based on the above-mentioned principle explanation. That is, it is assumed that at the time of starting focusing, one image obtained by scanning the target object 1 is stored in the original image memory 8 as image data. In addition, instructions regarding which position to focus on are also transmitted from the controller 13 to the controller 1.
Assume that it is given for 4. here,
The set sound speed at the time of scanning, that is, the sound speed value used to calculate the delay map necessary for the beamformer 6, is 1540 m/sec (a commonly used sound speed) or the sound speed of water at body temperature. In perturbing this set sound speed value,
The control data for either or both of the beamformer 6 and the delay distribution of the auxiliary variable delay line array 5 will be recalculated according to the new sound speed value. Then, through this trial, we will actually try to see if we can obtain better focus. To do this, as mentioned in the principle explanation above, we need data (one-dimensional data, i.e., line A-A ′ data) and compare it with the same data obtained from the previous image to determine whether it has become sharper or not. This is executed in the image processing/comparison/calculation section 9 by referring to the contents of the original image memory 8. Based on the above judgment result, the controller 14 determines the amount of Celsius to be tried or decides to terminate the trial. The judgment method is performed according to the principle explanation explained earlier, but in reality, the judgment from either viewpoint (a) or (b) is performed while checking whether the answer is roughly the same. It is preferable. To explain more specifically, in method (a), an isolated, small, and more or less stronger target is found in the region of interest, and if the target is focused, many elements are Since the signals are in phase, the amplitude of the synthesis result becomes large. Therefore, this task is a procedure of looking for the maximum point regarding the set sound speed while paying attention to the brightness level of the point on the image. On the other hand, in the method (b), it is not necessary to use isolated points as mentioned above;
Spatial filtering of the dimensional image, ie convolution with a kernel, is required. Also, in the method (b), there is a problem that if the focus is too much, the specularity of the image itself will interfere and make it difficult to judge, so after making rough adjustments using the method (b), It is preferable to perform this for fine adjustment or confirmation of convergence. To define the characteristics of the filter, it is necessary to prepare two kernel functions, one for high-frequency extraction and one for middle-frequency extraction, but they are based on the orientation of the system (with full aperture) in the region of interest. One preferred embodiment is a kernel function for spatial frequencies in the vicinity of the resolution, which includes the resolution, and a kernel function for lower frequencies that does not include the resolution. This means that even if the azimuth resolution theoretically estimated from the aperture size and the wavelength used is directly used to determine the filter kernel function, a filter that is comparable in practical terms can be obtained. That,
There is no need to explain that it may be used after being modified empirically. In the method (b), compared to the results of these two filter processes, trials and convergence are performed so that the value obtained by dividing the energy of the high frequency component by that of the low frequency component becomes maximum.

Next, the sound velocity distribution regarding the partial apertures C _A , C _B to C _D will be corrected, and a specific example of the method will be described. In this case, it is assumed that the method described above is able to focus within the range of the assumption that there is no distribution of sound speed in the medium without any major errors, and the resulting image is used as a reference. It is assumed that the image is stored somewhere in the original image memory 11. Therefore, the controller 14 operates the switch array 4 so that each of the partial openings C _A , C _B to C _D is selectively adopted, and images are taken for each of them.
The images are stored in each part of the original image memory 8. It is not necessary to accommodate the entire screen, and only the small area D to be focused is sufficient. For the first time, the settings of the preceding state are used without any correction or correction. That is, each partial aperture is still phase-synthesized using beamforming data on the premise that there is no (uniform) sound velocity distribution in the medium. The image formed by each partial aperture obtained in this way has a much lower spatial resolution than the image obtained using the entire comparison aperture due to the small number of elements, but it still has a good correlation with the comparison image. exhibits. Therefore, data on the line A-A' is extracted for both and compared, that is, correlation is determined.
Once the above correlation values for the images of each sub-aperture are obtained, by slightly changing the set sound speed for each sub-aperture, determine through trial and error whether the correlation value can be further improved or not, and proceed in the direction of further improvement. . The perturbation of the sound speed setting for this task is the beamformer 6
Since it is not possible to use the auxiliary variable delay line 5. However, instead of repeating image acquisition by correcting the set sound speed one by one, if the size of the region of interest D is sufficiently small compared to the distance from the aperture, as a result of correcting the set sound speed,
It can be simplified to say that what changes is roughly only the relative position in the distance direction, so in the end, the difference in the distance direction of the acquired results for each subaperture of the image of the small area D, or the time of the echo. This is equivalent to finding a deviation on the axis. Therefore, changing and correcting the set sound speed is equivalent to simply shifting the time axis by changing and correcting the common bias of the delay amount throughout each partial aperture. Therefore, in this case, by changing the position of cross section A-A' in the image of the small area of interest D obtained by each partial aperture in the sound ray direction,
By determining the correlation with the A-A' cross-section of the full-aperture image for reference, A-A' is fixed, and the set sound speed, that is, the common bias term of the delay amount described above, is corrected and the echo is calculated. Shifting the time axis of the image can be used as a substitute for repeating image acquisition, and decisions can be made more quickly. Of course, it is obvious that the latter method may be used. Once the delay correction value for each partial aperture is determined in this way, it is set and the image with the full aperture is taken again. Below, if this one step is insufficient, use this image (obtained in this way) as a reference again to calculate the correction terms for C _A , C _B to C _D , or repeat this as many times as necessary. repeat.

Next, we will return to the beginning and explain another method of adjusting the set value of the average sound speed using the full aperture. This is a modified version of the above-mentioned high-range and mid-range comparison method, but there is an autocorrelation method as an equivalent method that can replace the filter method using convolution. In other words, the sharpness of the peak at the origin of the normalized autocorrelation represents the relative amount of high-frequency components of the original data, so instead of comparing the power of the filter output, the sharpness of the peak at the origin of this autocorrelation is can be used. Compared to convolution calculation using a simple kernel, autocorrelation requires a large amount of calculations and is difficult, but it is easier when a dedicated processing unit (LSI chip correlator) is available.
This may be preferable in terms of processing speed.

As explained above, according to the present invention, the sound velocity correction values are respectively determined so that the same line image has the maximum correlation value with the image by full aperture, and the sound velocity correction values are reprojected using these sound velocity correction values. Since the image is reconstructed for the entire aperture, it is possible to realize an ultrasonic imaging method that automatically obtains a sharply focused image.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram showing the relationship between the aperture and the region of interest, Figs. 2 and 3 are waveform diagrams showing an example of the autocorrelation waveform of the frequency spectrum of an image, and Fig. 4 is an explanatory diagram showing the relationship between the aperture and the region of interest.
The figure is a configuration diagram showing a wave receiving system of an ultrasonic imaging apparatus for implementing the ultrasonic imaging method of the present invention. TD……Array transducer, TD(D)……
Point of interest, D...Small area, C...Full aperture, C _A , C _B ,
..., C _D ...partial aperture.

Claims (1)

[Claims] 1. After first determining the overall sound speed setting value so that the sharpness of the image in the small region D of interest is maximized, the image in the small region D by the partial aperture is compared with the image in the small region obtained initially. A sound velocity correction value related to the partial aperture is determined so as to have the maximum correlation value, and the partial aperture of interest is moved and changed in size so as to cover the entire aperture, and corresponds to each part of the entire aperture. 1. An ultrasound imaging method using wavefront reprojection corrected for sound speed, characterized in that a sound speed correction value is determined.