JP3696616B2 - Ultrasonic diagnostic equipment - Google Patents

Description

  The present invention relates to an ultrasonic diagnostic apparatus that obtains a tomographic image of a subject using ultrasonic waves, and more particularly to an ultrasonic diagnostic apparatus that incorporates a technique for extracting the outline of a tissue in a subject.

  Conventionally, an ultrasonic wave is transmitted into a subject, and an ultrasonic wave reflected and returned by a tissue in the subject is received to obtain a received signal, and a tomographic image in the subject is displayed based on the received signal. In recent years, in order to make the diagnosis of the disease easier, the outline of the tissue appearing on the tomographic image, for example, the contour of the left ventricle of the heart, is used. It is desired to display the extracted area or the area of the region surrounded by the outline and the volume of the tissue estimated from the outline.

  In order to extract the outline, simply, for example, the light pen 2 is traced on the tomographic image displayed on the display 1 along the outline of the desired tissue, as shown in FIG. The contour can be extracted from the locus.

Also, as shown in FIG. 46, for example, an area 3 is designated so as to surround a desired tissue using a trackball or a mouse, and the image data in the designated area is binarized with an appropriate threshold value. A method for extracting a contour has been proposed (see Non-Patent Document 1).
"Measurement of left ventricular cross-sectional area and left ventricular volume by the acoustic quantification method; influence of region of interest setting site" Toshihito Yagi et al. 63rd Annual Meeting of the Japanese Society of Ultrasonic Medicine 517-518 pages 1993 November

  However, in the former method in which the person traces the outline, it is necessary to carefully trace the locus by the light pen or the like so as not to deviate from the outline. In addition, if the tissue for which the contour is to be obtained is the left ventricle of the heart, it may be necessary to repeat such operations for all the tomographic images for one beat of the heart movement. Nevertheless, there is a problem that it is easy to obtain inaccurate data lacking objectivity and is far from practical.

  Also, in the latter method of extracting a contour by specifying an approximate region 3 and binarizing the image data in the region, the heart continuously beats and repeatedly expands and contracts. For example, as shown in FIG. 46A, not only the left ventricle 4 but also the left atrium or the right ventricle is included in the region 3 as shown in FIG. If a part of the left ventricle 4 deviates from the region 3 as shown in FIG. 3), an incorrect area or volume is obtained as the area or volume of the left ventricle 4 in any case. To avoid this, it is necessary to carefully handle the left ventricle 4 and surround the entire left ventricle 4 by hand, and the problem of time-consuming work is not solved, even if the outline itself is not traced. . Further, in the method of extracting the outline by binarizing the image data in the approximate region 3, because of the influence of noise or the like, and because there is a difference in luminance (size of image data) depending on the location, A contour line lacking connection as a two-dimensional image, that is, a contour line that does not form a closed curved surface may be extracted.

  In view of the above circumstances, an object of the present invention is to provide an ultrasonic diagnostic apparatus having a function capable of objectively extracting a contour of a tissue without a manual operation or with a simple manual operation.

  It is another object of the present invention to provide an ultrasonic diagnostic apparatus having a function capable of extracting contours for a large number of frames without manual operation or simply inputting contours for a very small number of frames. One.

The first ultrasonic diagnostic apparatus of the present invention is an ultrasonic diagnostic apparatus that obtains image data corresponding to each point in a tomographic plane extending in a subject by receiving ultrasonic waves reflected in the subject.
(1) Gradient calculation means for obtaining a gradient of image data for a plurality of points in the tomographic plane (2) Scalar amount calculation means for obtaining a scalar quantity corresponding to the gradient for a plurality of points in the tomographic plane ( 3) Maximum point calculation means for obtaining a plurality of local maximum points in the tomographic plane where the scalar quantity is maximum (4) Contour extraction means for obtaining the contour of a predetermined tissue in the subject based on the plurality of local maximum points It is provided with.

Here, in the first ultrasonic diagnostic apparatus of the present invention, the local maximum point calculation means of (3) is:
(5) Preferably, the maximum point is obtained on the designated line based on a change in the scalar quantity in the direction along each of the designated lines extending in the tomographic plane,
Furthermore, it is preferable that the local maximum point calculation means uses a plurality of designated lines extending radially in the tomographic plane starting from a predetermined center point in the tomographic plane as the plurality of designated lines.

Further, when the local maximum point calculation means is for obtaining the local maximum point along the designated line extending in the tomographic plane described in (5) above,
(6) Means for detecting the change point at which the sign of the difference changes by the difference between the scalar quantities of a plurality of points adjacent to each other along the specified line, and the change point Means for correcting the scalar quantity of the change point based on the scalar quantity of the neighboring point, and the change point based on the sign of the difference of the change point when using the scalar quantity corrected for the change point It is preferable to include means for determining whether or not to extract as a maximum point.

In the first ultrasonic diagnostic apparatus of the present invention, as the gradient calculation means of (1), for example,
(7) A means for differentiating in two different directions within the tomographic plane can be employed.
(8) Further, this gradient calculation means is not less than 0.5 times the width of the scooping one of the widths in the differential direction of each of the two tissues in contact with the contour to be extracted appearing on the fault plane. It is preferable that a two-dimensional differential operation is performed using a differential filter having a width less than double.

Furthermore, in the first ultrasonic diagnostic apparatus of the present invention,
(9) The scalar quantity calculation means of (2) may be configured such that, as the scalar quantity, the absolute value of the gradient, a function having the absolute value as a variable, and each of two different directions on the gradient plane of the gradient. It is preferable to obtain at least one selected from the group consisting of a sum of absolute values of components and a function having the sum as a variable.

Further, in the first ultrasonic diagnostic apparatus of the present invention, when the contour is extracted by the contour extracting means of (4), the contour extracting means
(4-1) A feature amount calculating means for obtaining a feature amount indicating the probability that the maximum point is a point on the contour is provided, and a point on the contour is extracted from the maximum point based on the feature amount, It is preferable that the contour is obtained by connecting the points on the contour.

The feature quantity is based on the change in the scalar quantity in the direction along each of a plurality of designated lines extending in the tomographic plane when the local maximum point calculation means in (3) satisfies the requirement in (5) above. When the maximum point is obtained on the specified line,
(A) the scalar quantity of the local maximum point or a function having the scalar quantity as a variable (b) an average value of image data corresponding to a predetermined area including a point corresponding to the local maximum point, or an average thereof (C) a sign of the scalar product of the vector along the specified line and the gradient of the maximum point on the specified line, or a numerical value (d) representing the sign, The angle between the gradients of the first and second maximum points respectively located on the designated line, or a function having the angle as a variable (e) The first designation of the first and second designated lines different from each other The distance between the intersection of the second specified line and the straight line extending in the direction perpendicular to the gradient direction of the maximum point located on the line and the maximum point located on the second specified line, or That distance It is preferable to employ a feature quantity including at least one or more combinations selected from the group consisting of a function whose variable.

In the first ultrasonic diagnostic apparatus of the present invention, when the contour extraction unit of (4) includes the feature amount calculation unit of (4-1), the contour extraction unit further includes:
(4-2A) Candidate point extracting means for obtaining a candidate contour point having a probability existing on the contour from the maximum points based on the feature amount (4-3A) The contour obtained by the candidate point extracting means Contour point extraction means for adopting the contour candidate point as a point on the contour when the distance between the candidate point and a point on the contour adjacent to the contour candidate point is within a predetermined distance It is preferable.

Alternatively, when the contour extracting unit (4) includes the feature amount calculating unit (4-1), the contour extracting unit further includes:
(4-2B) Based on the above feature amount obtained for a predetermined frame among a plurality of frames representing the same tomographic plane at different times, it exists on the contour from among the local maximum points in the predetermined frame. Candidate point extracting means (4-3B) for obtaining a contour candidate point having a probability of the contour candidate point in the predetermined frame obtained by the candidate point extracting means, and among the plurality of frames corresponding to the contour candidate point Contour point extraction that adopts the contour candidate point as a point on the contour in the predetermined frame when the distance between the points on the contour in a frame different from the predetermined frame is within a predetermined distance It is also a preferable aspect to provide means.

Furthermore, when obtaining the contour of the left ventricle using the first ultrasonic diagnostic apparatus of the present invention as the subject of the human heart,
(10) provided with a valve root detecting means for detecting a boundary point between each of the two valves existing between the left ventricle and the left atrium and the septum of the left ventricle appearing on the tomographic plane;
The (4) contour extracting means includes the feature amount calculating means of (4-1), and on the basis of the feature amount, the contour is extracted on an area on the tomographic plane excluding the area connecting the boundary points. It is also a preferred embodiment that the points are extracted.

  Specifically, the above-mentioned (10) valve root detecting means includes, for example, the above (4) contour extracting means, the above (4-1) feature amount calculating means, and the above (4-2A) candidate point extracting. And the above-mentioned (4-3A) contour point extracting means, the above-mentioned (10) valve root detecting means is obtained by the above-mentioned (4-2A) candidate point extracting means. A distance between one contour candidate point and a point on the contour adjacent to the first contour candidate point is a distance within a predetermined distance, and the first contour candidate point and the first contour candidate The distance from the second contour candidate point adjacent to the point is a distance that is a predetermined distance or more, and the first contour candidate point is extracted as the boundary point as at least one criterion. Can do.

  Here, the “predetermined distance” referred to in the valve root detecting means may be a distance different from the “predetermined distance” referred to in the contour extracting means (4-3A).

  Alternatively, the contour extracting unit (4) includes the feature amount calculating unit (4-1), the candidate point extracting unit (4-2B), and the contour point extracting unit (4-3B). The root detection means of (10) above corresponds to the contour candidate points in the predetermined frame obtained by the candidate point extraction means of (4-3A) and the contour candidate points. Among the plurality of frames, the distance between a point on the contour in a frame different from the predetermined frame is a distance that is a predetermined distance or more, and at least one of the determination criteria, the contour candidate point is The boundary point in a predetermined frame can be extracted.

  The “predetermined distance” referred to in the valve root detecting means may be a distance different from the “predetermined distance” referred to in the contour point extracting means (4-3B).

  Further, the first ultrasonic diagnostic apparatus of the present invention includes a contour line connecting the points on the contour with a straight line or a curve, and a screen in which the inside and the outside of the contour are distinguished by color, brightness, or pattern. It is preferable to include a contour display means for displaying at least one of them.

  Further, the first ultrasonic diagnostic apparatus of the present invention displays an area calculating means for calculating an area inside the contour extracted by the contour extracting means of (4), and an area calculated by the area calculating means. It is preferable to have a configuration including an area display means. In this case, it is preferable that the area display means displays the area at least one of a number representing the area and a graph showing the time change of the area.

  The first ultrasonic diagnostic apparatus of the present invention includes a volume calculation means for obtaining the volume inside the contour extracted by the contour extraction means of (4), and a volume display means obtained by the volume calculation means. It is also a preferable aspect to have a configuration including In this case, like the area display means, the volume display means preferably displays the volume at least one of a number representing the volume and a graph showing the time change of the volume. .

The second ultrasonic diagnostic apparatus of the present invention is an ultrasonic diagnostic apparatus that obtains image data corresponding to each point in a tomographic plane extending in a subject by receiving ultrasonic waves reflected in the subject.
(12) Contour calculation means for obtaining the contour of a predetermined tissue in the subject based on the image data (13) Gradient calculation means for obtaining the gradient of the image data for a plurality of points in the tomographic plane (14) Based on the direction of the gradient, the calculation effective area detection means for determining the calculation effective area containing the predetermined tissue in the tomographic plane,
The contour calculating means of (12) is characterized in that the contour of the predetermined tissue is obtained inside the calculation effective area.

  Here, in the second ultrasonic diagnostic apparatus of the present invention, the calculation effective area detecting means of (14) specifically originates from, for example, a predetermined center point within the tomographic plane, and within the tomographic plane. The image data corresponding to each point in the tomographic plane is based on the angle formed by the direction of the designated line extending through the point where the gradient is obtained and the direction of the gradient at the point where the gradient is obtained. The binarized image data can be configured to detect the calculation effective region based on the distribution in the tomographic plane of the binarized image data.

Note that the second ultrasonic diagnostic apparatus of the present invention can be configured to incorporate the first ultrasonic diagnostic apparatus of the present invention as it is. As a basic aspect in that case, the contour calculation means of the above (12) in the second ultrasonic diagnostic apparatus of the present invention,
(12-1) Scalar amount calculation means for obtaining a scalar amount corresponding to the gradient with respect to a plurality of points in the tomographic plane (12-2) A plurality of points in the calculation effective region in which the scalar amount is maximized Maximum point calculation means for obtaining a maximum point (12-3) A contour point extraction means for obtaining a contour of the predetermined tissue based on the plurality of maximum points.

  The contour extracting means (12-3) is characterized by the probability that the maximum point is a point on the contour, like the contour extracting means (4) in the first ultrasonic diagnostic apparatus of the present invention. It is preferable that a feature amount calculating means for obtaining a quantity is provided, a point on the outline is extracted from the maximum points based on the feature quantity, and the outline is obtained by connecting the points on the outline.

As the feature amount, as in the case of the first ultrasonic diagnostic apparatus of the present invention, the maximum point calculation means of (12-2) follows a direction along each of a plurality of designated lines extending in the tomographic plane. Based on the change in the amount of scalar, the maximum point is determined on the specified line.
(A) the scalar quantity of the local maximum point or a function having the scalar quantity as a variable (b) an average value of image data corresponding to a predetermined area including a point corresponding to the local maximum point, or an average thereof (C) a sign of the scalar product of the vector along the specified line and the gradient of the maximum point on the specified line, or a numerical value (d) representing the sign, The angle between the gradients of the first and second maximum points respectively located on the designated line, or a function having the angle as a variable (e) The first designation of the first and second designated lines different from each other The distance between the intersection of the second specified line and the straight line extending in the direction perpendicular to the gradient direction of the maximum point located on the line and the maximum point located on the second specified line, or That distance It is preferable to employ a feature quantity including at least one or more combinations selected from the group consisting of a function whose variable.

  In addition, the second ultrasonic diagnostic apparatus of the present invention can be configured apart from the first ultrasonic diagnostic apparatus of the present invention. In that case, as the contour calculation means of (12) above, for example, a contour calculation means for extracting a contour by binarizing the image data described with reference to FIG. 46 can be employed.

In addition, the second ultrasonic diagnostic apparatus of the present invention is
(15) It is preferable to include display means for displaying the contour.

In the second ultrasonic diagnostic apparatus of the present invention,
(16) It is also preferable to include a guidance amount calculation means for obtaining a guidance amount calculated based on the contour obtained by the contour calculation means.

  In this case, it is preferable that the guidance amount calculation means of (16) obtains at least one of the area inside the contour, the center of gravity position of the contour, and the volume inside the contour as the guidance amount.

Furthermore, in the second ultrasonic diagnostic apparatus of the present invention, when the guidance amount calculation means of (16) is provided,
(17) It is preferable to include display means for displaying at least one of the induced amount and the amount of change of the induced amount over a plurality of frames.

A third ultrasonic diagnostic apparatus of the present invention is an ultrasonic diagnostic apparatus that obtains image data corresponding to each point in a tomographic plane extending in a subject by receiving ultrasonic waves reflected in the subject.
(18) Motion calculation means for calculating the motion of a plurality of points in the tomographic plane based on the image data of a plurality of frames (19) Contour calculation means for calculating the contour of a predetermined tissue in the subject based on the image data (20) It is characterized by comprising contour deformation means for obtaining a contour deformed from the contour obtained by the contour calculation means based on the motion detected by the motion calculation means.

  Here, in the third ultrasonic diagnostic apparatus of the present invention, the motion producing means of (18) uses the optical flow method or the cross-correlation method as the motion, and moves in the tomographic plane. It is preferable that one of the two-dimensional motion vector and the magnitude of motion in a predetermined direction within the tomographic plane be calculated. In this case, as the predetermined direction, directions of a plurality of designated lines extending radially within the tomographic plane from a predetermined center point inside the contour can be selected.

  Further, the motion calculating means of (18) calculates a motion of each of a plurality of points for each of a plurality of regions set in the tomographic plane, and is highly likely to represent an accurate motion for each of the plurality of regions. It is preferable that a representative value of the extracted motion is obtained for each of the plurality of regions.

Further, the third ultrasonic diagnostic apparatus of the present invention described above,
(21) Preprocessing means for performing one of smoothing processing and binarization processing on the image data is provided,
It is also a preferred aspect that the motion calculation means (18) calculates motion based on the image data subjected to the one process by the preprocessing means.

The third ultrasonic diagnostic apparatus of the present invention has a configuration in which the first ultrasonic diagnostic apparatus of the present invention is incorporated as it is, or the first ultrasonic diagnostic apparatus of the present invention and the second ultrasonic diagnostic of the present invention. It is possible to adopt a configuration in which both the apparatus and the apparatus are incorporated as they are. As a basic mode in that case, the contour calculation means of the above (19) in the third ultrasonic diagnostic apparatus of the present invention,
(19-1) Gradient calculation means for obtaining a gradient of image data for a plurality of points in a tomographic plane (19-2) A scalar for obtaining a scalar quantity corresponding to the gradient for a plurality of points in a tomographic plane Quantity calculation means (19-3) Maximum point calculation means for obtaining a plurality of local maximum points in the tomographic plane where the scalar quantity is maximum (19-4) Based on the plurality of local maximum points, a predetermined value in the subject An outline extracting means for obtaining the outline of the tissue is provided.

Further, the third ultrasonic diagnostic apparatus of the present invention can be configured to be separated from the first ultrasonic diagnostic apparatus of the present invention and the second ultrasonic diagnostic apparatus of the present invention. In this case, for example, the third ultrasonic diagnostic apparatus of the present invention is
(22) a contour point designating operator for designating a contour point located on a contour of a predetermined tissue;
The contour calculation means (19) may be configured to obtain a contour based on the contour point designated by the contour point designating operator.

The third ultrasonic diagnostic apparatus of the present invention is
(23) A synchronization signal acquisition means for obtaining a heartbeat synchronization signal synchronized with the heartbeat of the subject,
It is also a preferable aspect that the contour calculating means (19) obtains a contour for a frame designated based on the heartbeat synchronization signal.

Furthermore, the third ultrasonic diagnostic apparatus of the present invention is
(24) Storage means for storing a plurality of frames of image data in an overwritable manner (25) A freeze means for prohibiting overwriting to the storage means,
At least one predetermined frame of the plurality of frames stored in the storage means in the above (24) in the freeze state in which the contour calculation means in the above (19) is prohibited from being overwritten by the freeze means in the above (25). Is to seek the contour,
The motion detection means of (17) calculates motion for other frames excluding the predetermined frame among the plurality of frames stored in the storage means of (23) in the frozen state,
It is also a preferable aspect that the contour deforming means of the above (19) obtains a contour deformed from the contour obtained by the contour calculating means for the other frame.

Further, as in the second ultrasonic diagnostic apparatus of the present invention, it is also preferable that the third ultrasonic diagnostic apparatus of the present invention includes (25) display means for displaying the contour.

In the third ultrasonic diagnostic apparatus of the present invention,
(26) It is also preferable to include a guidance amount calculation means for obtaining a guidance amount calculated based on at least one of the contour obtained by the contour calculation means and the contour deformed by the contour deformation means.

  In this case, it is preferable that the guidance amount calculation unit of (26) obtains at least one of the area inside the contour, the center of gravity position of the contour, and the volume inside the contour as the guidance amount.

Furthermore, in the third ultrasonic diagnostic apparatus of the present invention, when the guidance amount calculating means of (26) is provided,
(27) It is preferable to include display means for displaying at least one of the induced amount and the amount of change of the induced amount over a plurality of frames.

  Since the first ultrasonic diagnostic apparatus of the present invention has the above-described configurations (1) to (4), it designates one point or one line segment on the tomographic image display screen without manual work or for example. It is possible to objectively extract the outline of the tissue only with an extremely simple manual operation.

  In addition, in the first ultrasonic diagnostic apparatus of the present invention, when the concept of the designated line described in (4) above is adopted, in obtaining the maximum point of the scalar quantity, the scalars at each point adjacent to the designated line are obtained. It is only necessary to calculate the difference between the quantities, and the maximum point can be easily obtained. In addition, if a specified line extending radially from a predetermined center point is adopted as these specified lines, a point on the contour will be searched radially from the center point, so in which direction ( The same algorithm can be used for any specified line.

  Further, when the local maximum calculation means that satisfies the requirement (6) is provided, the possibility that the local maximum point is erroneously detected due to noise or the like included in the image data is reduced.

  Further, the gradient calculation means (1) is constituted by means for performing two-dimensional differentiation as described in (7) above. In this case, a tomographic image can be obtained by a generally known differential filter. For example, it is sufficient to differentiate in the vertical direction and the horizontal direction, so that this gradient calculation means is configured relatively easily.

  In that case, as described in the above (8), it is preferable to employ a differential filter having a width not less than 0.5 times and not more than twice the width of the scooping tissue.

  When the width of the differential filter is less than 0.5 times, even if it is a point on the contour to be extracted, a small value is obtained as a differential value (a component in the differential direction of the gradient) and the contour cannot be extracted well. On the other hand, when the width of the differential filter exceeds twice, the difference between the differential value of the point on the contour and the differential value of the surrounding points is small, and a smooth differential value is obtained as a whole, and the contour is blurred. Because it will end up.

  In the first ultrasonic diagnostic apparatus of the present invention, the scalar quantity calculating means (2) is an amount corresponding to the magnitude among the direction and magnitude (gradient absolute value) constituting the gradient. Any one may be used, and for example, any one of the quantities listed in the above (9) can be selected.

In addition, in the first ultrasonic diagnostic apparatus of the present invention, when the contour is obtained by the contour extraction means of (4), as described above, the feature amount representing the probability that the maximum point is a point on the contour is used. It is preferable to provide a feature amount calculation means to be obtained and to obtain a contour based on the feature amount. In that case, any one of the above (a) to (e) or a plurality of them It is preferable to include a combination of the above in the feature amount. The operation when each of the above (a) to (e) is used as a feature amount is as follows.
(A) Scalar amount at local maximum point As described above, this scalar amount is an amount corresponding to the gradient of the local maximum point, and represents the degree of change in image data at the local maximum point. The image data greatly changes in the outline of the tissue. Therefore, this scalar quantity can be adopted as the feature quantity.
(B) Average value of the image data at the point corresponding to the maximum point The average value of the image data of the two tissues sandwiching the contour is usually greatly different. Therefore, the value of the image data itself can be adopted as the feature amount. However, the corresponding point here is a point separated from the maximum point in the gradient direction of the maximum point by about ½ of the width of the tissue (for example, myocardium) to which the corresponding point belongs. The reason for taking the average is that the value of the image data may fluctuate greatly due to the influence of noise or the like when there is only one point.
(C) Sign of scalar product between gradients As described above, the average values of the image data of the two tissues sandwiching the contour are usually greatly different. Therefore, in the contour, the vector direction of the gradient is always directed from one tissue to the other. Therefore, the above code can be adopted as a feature amount.
(D) Angle The contour is continuous. Therefore, for example, the gradients of the contour points on the adjacent designated lines are directed in substantially the same direction. That is, the angle between these gradients is considered to be small. Therefore, this angle can be adopted as the feature amount.
(E) Distance As described above, the contour is continuous. Therefore, the probability that the one where the distance mentioned above is short is a point on an outline will be large. Therefore, this distance can be adopted as a feature amount.

  In the first ultrasonic diagnostic apparatus of the present invention described above, the contour extracting unit in (4) above includes the candidate point extracting unit in (4-2A) and the feature amount calculating unit in (4-1). In the case where the contour point extracting means (4-3A) is provided, or in addition to the feature amount calculating means (4-1), the candidate point extracting means (4-2B) and the (4-3B) When the contour point extracting means is provided, it is possible to prevent the adjacent contour points from being erroneously detected at positions far apart from each other, and to perform contour extraction with higher accuracy.

  When the first ultrasonic diagnostic apparatus of the present invention is used to obtain the contour of the left ventricle using the human heart as the subject, the valve root detecting means (10) is provided, and two valves on the tomography are provided. If a point on the contour is extracted on the region excluding the region connecting the roots of the two points, that is, if a point considered to be a point on the contour is extracted within the region connecting the two valve roots If the point is rejected without being regarded as a point on the contour, it is possible to prevent an erroneous contour from being extracted due to the movement of the valve, and to perform contour extraction with higher reliability.

  As the valve root detecting means of (10) above, adjacent contour candidate points are predetermined as in (4-2A) and (4-3A) or (4-2B) and (4-3B) above. It is possible to employ a means for extracting a boundary point located at the root of the valve with at least one of the determination criteria as to whether or not the distance is exceeded.

  Further, the first ultrasonic diagnostic apparatus of the present invention includes the contour display means, and displays an image that can identify the contour line indicating the contour obtained as described above and the inside and outside of the contour, By obtaining the area and volume inside the contour and displaying the area and volume, information effective for observation and diagnosis can be provided in an easily understandable form. When displaying the area or volume, the area or volume may be displayed as a number, but when a graph showing the temporal change of the area or volume is displayed, the expansion and contraction of the tissue due to the heartbeat etc. can be seen at a glance. It is clear and more effective.

  Since the second ultrasonic diagnostic apparatus of the present invention has the above-described configurations (12) to (14) and obtains the contour inside the calculation effective region, for example, in the method described with reference to FIG. A problem when a person designates the general area 3 in which a necessary area including an unnecessary area is removed due to expansion or contraction of the heart, for example. Is solved, and the contour is accurately obtained.

  Further, as described above, the second ultrasonic diagnostic apparatus of the present invention obtains a contour in the calculation effective region as described above. Therefore, the second ultrasonic diagnostic apparatus is used with a scalar quantity corresponding to the gradient. When applied to the first ultrasonic diagnostic apparatus of the present invention for detecting a contour based on a local maximum point, even if a point considered to be a contour point is detected at a position outside the calculation effective region, this is regarded as a contour point. The outline extraction with higher accuracy is possible by rejecting the calculation and obtaining the outline point only within the calculation effective area.

  In the second ultrasonic diagnostic apparatus of the present invention, the calculation effective area detecting means of (14) starts from a predetermined center point in the tomographic plane and passes through the point in the tomographic plane where the gradient is obtained. Binarization when the image data corresponding to each point in the tomographic plane is binarized based on the angle between the direction of the specified line extending and the gradient direction at the point where the gradient was obtained Based on the distribution in the tomographic plane of the image data, a means for detecting the calculation effective area can be employed. In this case, particularly when extracting the contour of the left ventricle, the left ventricle is reliably captured. The calculation effective area is determined. Therefore, the contour of the left ventricle is extracted with high accuracy.

  In addition, since the third ultrasonic diagnostic apparatus of the present invention has the above-described configurations (18) to (20), it is obtained for a certain frame without performing an operation for obtaining the contour from the beginning for each frame. Each contour for a plurality of series of frames is obtained by deforming the contour according to the movement of a plurality of points in the tomographic plane. In deforming the contour, the contour obtained by the contour calculating means may be directly deformed, or the contour obtained by deforming the contour may be further deformed.

  The movement of each point in the tomographic plane can be obtained by employing, for example, an optical flow method or a mutual interlayer method.

  The optical flow method and the cross correlation method will be described below.

[Optical flow method]
The optical flow method is a method for calculating an apparent velocity field of a moving image. Details are detailed in Koga, edited by Miike, “Moving Image Processing Using a Personal Computer” (Morita Kita Publishing). Now, the spatial coordinates on the tomographic plane are represented by (x, y), the time is represented by t, and the luminance on the image is represented by P (
When expressed as x, y, t), movement without a change in luminance is expressed by the following equation.

P (x, y, t) = P (x + δx, y + δy, t + δt) (1)
Here, if the right side is Taylor-expanded and the second and subsequent orders are ignored, the following expression is established.

  In the above equation, Px and Py are spatial gradients in the x and y directions, Pt is a temporal gradient, and u and v are optical flow velocities to be calculated. To solve the equation (2), one or more constraint conditions are further required. The accuracy of the calculated speed (u, v) is determined depending on how the constraint is given. However, since the method of giving the constraint itself is not directly related to the basis of the present invention, a typical method is described below. Show.

In order for the image to be spatially smooth, i.e. to minimize spatial variation,
(∂u / ∂x) ² + (∂u / ∂y) ² + (∂v / ∂x) ² + (∂v / ∂y) ²
→ min. ...... (3)
The optical flow is determined so that the error function of the following equation is minimized.

E = ∬ [{Px · u + Py · v + Pt} ²
+ Α ² {ux ² + uy ² + vx ² + vy ² }] dxdy (4)
However, ux = ∂u / ∂x, ui = ∂u / ∂y, vx = ∂v / ∂x,
vy = ∂v / ∂y.

  By using the expression (4) as a constraint condition, the calculated speed (u, v) is spatially sufficiently smooth and satisfies the expression (2). α gives a relative weight between the equations (2) and (3).

  The following partial differential simultaneous equations are obtained by the variational method.

  If the above equation is solved, the optical flow velocity (u, v) can be obtained.

  Further, as another constraint condition, a condition that the optical flow is substantially constant in the local region S is given, and minimization of the following expression is considered in the spatial coordinates (i, j) included in the region S.

E = ΣΣ [Px (i, j, k) u + Py (i, j, k) v
+ Pt (i, j, k)] ² (6)
In this case, the speed component is calculated as follows by the conditions of ∂E / ∂u = 0 and ∂E / ∂v = 0.

  Although the two-dimensional optical flow velocity is calculated here, the optical flow velocity may be calculated only for a predetermined one-dimensional direction in the tomographic plane.

[Correlation method]
Regarding the cross-correlation method, when the cross-correlation coefficient is r (m, n), for example,

Or the following values may be calculated for the binarized image data Q (i, j, t).

r (m, n) = ΣΣQ (i, j, t) Q (i + m, j + n, t + Δt)
...... (9)
If the maximum value is searched for in the cross-correlation function r (m, n), m and n at that time correspond to the movement. Further, when obtaining a motion only in a predetermined direction, a one-dimensional motion can be calculated by making the two-dimensional correlation function one-dimensional.

  In addition, in both the optical flow method and the cross-correlation method, when obtaining a motion only in a predetermined direction, adopting each direction of a plurality of specified lines extending radially within the tomographic plane starting from a predetermined center point inside the contour. The movement of the contour can be known with high accuracy by a one-dimensional calculation. In this case, the “predetermined center point” may be input by the operator, or the center of gravity inside the contour of the frame whose contour has already been determined is obtained, and the center of gravity is adopted as the predetermined center point. May be.

  Further, the motion calculating means of (18) calculates a motion of each of a plurality of points for each of a plurality of regions set in the tomographic plane, and is highly likely to represent an accurate motion for each of the plurality of regions. In the case where the representative value of the extracted motion is obtained for each of the plurality of regions, the accurate motion of each region in the tomographic plane is obtained, and the contour can be accurately deformed. .

  Further, in the third ultrasonic diagnostic apparatus of the present invention, when the preprocessing means of (21) is provided, the movement calculation means of (18) obtains a movement with more accurate or simpler calculation. It is done. In this case, in particular, when the motion calculation means is for obtaining a motion by the optical flow method, a more accurate motion is obtained by adopting a preprocessing means for performing a smoothing process as the preprocessing means. In addition, when the motion calculation means obtains a motion by the cross-correlation method, the calculation is simplified as shown in the above equation (9) by adopting a pre-processing means for performing binarization processing. , Movement is required more quickly.

  Further, in the third ultrasonic diagnostic apparatus of the present invention, the contour calculation means of (19) is the gradient calculation means, scalar quantity calculation means, maximum of the above (19-1) to (19-4). When the point calculation means and the contour extraction means are provided, each of these means (19-1) to (19-4) is the above (1) to (4) of the first ultrasonic diagnostic apparatus of the present invention. Thus, the third ultrasonic diagnostic apparatus of the present invention can be configured to incorporate the first ultrasonic diagnostic apparatus of the present invention as it is. However, in the third ultrasonic diagnostic apparatus of the present invention, when there are a plurality of frames, the outline is obtained for each of the frames in the same manner as in the first ultrasonic diagnostic apparatus of the present invention. Instead, the contour is obtained for a certain frame in the same manner as the first ultrasonic diagnostic apparatus, and the other frames are calculated so as to obtain the contour for the frame by deforming the contour.

  In the case of obtaining the contour using the same method as that in the first ultrasonic diagnostic apparatus of the present invention, the region for calculating the contour is calculated in the same manner as in the second ultrasonic diagnostic apparatus of the present invention. You may limit within an effective area | region. In this case, both the first and second ultrasonic diagnostic apparatuses of the present invention are incorporated into the third ultrasonic diagnostic apparatus of the present invention.

  Further, in the third ultrasonic diagnostic apparatus of the present invention, the configuration is separated from the configurations of the first ultrasonic diagnostic apparatus and the second ultrasonic diagnostic apparatus, for example, the contour of (22) above. It is good also as a structure provided with the operator for point specification. At this time, the contour calculation means of (19) may obtain the contour by smoothly connecting the contour points input by the operation of the contour point designating operator, for example, with a spline function. When there is a large number of contours, the contour computing means may obtain the contours by, for example, simply connecting the contour points with a broken line without any substantial computation, or manually or semi-manually. You may comprise so that the input outline may be corrected in consideration of the outline calculated | required by the outline calculating means which consists of said (19-1)-(19-4), for example. The contour calculation means referred to in the present invention represents a concept including all these aspects.

  Further, the third ultrasonic diagnostic apparatus of the present invention comprises the synchronization signal acquisition means of (23) above, and the contour calculation means of (19) obtains an outline for a frame designated based on the heartbeat synchronization signal. If this is the case, specify the frame that can accurately determine the outline synchronized with the heartbeat, obtain the outline for that frame, and for other frames synchronized with the heartbeat, modify the outline of the frame for which the outline was obtained. Since the contour of the frame can be obtained and an accurate contour is obtained as an initial value, a more accurate contour is obtained for all frames.

  As the heartbeat synchronization signal, for example, an electrocardiogram waveform signal, a Doppler waveform signal, or the like can be employed.

  Further, in the third ultrasonic diagnostic apparatus of the present invention, the storage means and the freeze means of (24) and (25) are provided, the contour is obtained for a certain frame among the plurality of frames in the freeze state, and the other frames As for the frame contour, the frame contour may be obtained by deforming the contour. At this time, when the contour is not obtained correctly, it is possible to perform fine work such as manually changing the contour, or fine observation such as slowly reviewing the contour of each frame.

  In the second ultrasonic diagnostic apparatus of the present invention to the third ultrasonic diagnostic apparatus of the present invention, the outline is displayed by the display means (15) and (25), for example, as in the first ultrasonic diagnostic apparatus. Alternatively, the guidance amounts (16) and (26), for example, the area inside the contour, the position of the center of gravity of the contour, the volume inside the contour, etc., or the amount of change of these guidance amounts over a plurality of frames are the above (17). , (27).

  The method for calculating the volume inside the contour is not limited to a specific method. For example, the volume of a sphere or spheroid when the contour is approximated by a circle or an ellipse and the circle or ellipse is rotated. Alternatively, the outline may be approximated by a stack of a plurality of rectangular parallelepipeds, and the total volume of each disk when each cube is rotated may be obtained.

  As described above, according to the present invention, the outline of a tissue can be objectively extracted without a manual operation by an operator or only with a simple manual operation.

  In addition, according to the second ultrasonic diagnostic apparatus of the present invention, it is possible to extract the tissue contour more accurately by determining the calculation effective region.

  Further, according to the third ultrasonic diagnostic apparatus of the present invention, the contour obtained for a certain frame is deformed according to the movement of a plurality of points in the tomographic plane, thereby calculating the contour for each frame first. The contours for a series of frames are obtained without performing the above.

  Examples of the present invention will be described below.

  First, an embodiment of the first ultrasonic diagnostic apparatus of the present invention will be described. Since the basic configuration of the ultrasonic diagnostic apparatus is widely known, the illustration and description of the basic configuration of the ultrasonic diagnostic apparatus are omitted here.

  FIG. 1 is a diagram showing a flow of extracting and displaying an outline in an embodiment of the first ultrasonic diagnostic apparatus of the present invention, and FIG. 2 is a plurality of radially extending from a predetermined center point (X0, Y0). It is a figure which shows (16 in this example) designation | designated line.

  First, image data representing a tomographic image in the subject is obtained by transmitting the ultrasonic wave into the subject and receiving the ultrasonic wave reflected and returned in the subject. Hereinafter, the image data may be referred to as an image or a tomographic image without particularly distinguishing the image data from the image.

  Next, the tomographic image in the image memory of the ultrasonic diagnostic apparatus is read out in both the horizontal and vertical directions of the screen and applied with a differential filter. These are the x and y components of the gradient. This is referred to as a gradient calculation means.

  When obtaining the gradient, the width of the differential filter is preferably not less than half and not more than twice the thickness of the myocardium. This is because if the width is less than half of the myocardium, the weak boundary cannot be emphasized, and if the gradient size is displayed, the image will be cut off. The change in the size of the gent becomes gentle and the peak becomes flat. As a result, it is difficult to specify the boundary in any case.

  Next, the scalar quantity calculating means obtains the scalar quantity (gradient magnitude). The magnitude of the gradient is the sum of the square of the x component and the square of the y component, that is, the absolute value of the gradient, or the square root of the sum, that is, the absolute value of the gradient, or the x component and y. The sum of the absolute values of the components can be considered.

  Next, as shown in FIG. 2, the center point (X0, Y0) input manually or the center point (X0, Y0) designated in advance is scanned radially along each designated line. Then, the maximum point of the gradient size is detected by taking the difference of the gradient size on each specified line (maximum point calculation means). The point where the sign of the difference changes from plus to minus is the maximum point. At that time, if the size of the gradient changes gently, multiple maxima are extracted from one mountain. Therefore, when the sign of the difference changes, a representative local maximum point can be selected by replacing it with the average value of the data before and after that and checking the sign again. Details of this will be described later.

  Since many local maximum points extracted by the above method are usually found on a single line, it is necessary to select an optimum point on the contour. Therefore, in the contour extraction means, several feature values are calculated for each local maximum point, and the total feature amount is calculated for each local maximum point by combining those feature amounts, and the total feature amount is calculated. Based on the above, a point on the contour (contour point) is extracted from the local maximum points (contour extraction means). An example of the feature amount will be described later.

  When the contour points are obtained as described above, the contour points are connected by straight lines or curves to display the contour (contour display means).

  Some means for obtaining the area from the contour points, straight lines, curves, contours and the like are also conceivable. For example, there is a method of finding the product by summing up the area of a triangle connecting two adjacent contour points and the center point.

  Details of each of the above means will be described below.

  3A and 3B are diagrams showing examples of gradient computing means, and FIG. 4 is a diagram showing a differential filter.

  FIG. 4A is an operator of a one-dimensional differential filter in the horizontal direction (this is assumed to be the X direction), and FIG. 4B is a graph of the differential filter. By applying such a differential operator having a certain inclination over the width d to the tomographic image, an image obtained by differentiating the tomographic image in the lateral direction (X direction) can be obtained.

  In the gradient calculation means shown in FIG. 3 (A), the horizontal differential filter having the shape shown in FIGS. 4 (A) and 4 (B), and the vertical differential filter having a shape in which the horizontal differential filter is rotated 90 degrees to be vertically oriented. Thus, the entire tomographic image is differentiated vertically and horizontally. Of the images obtained by each differentiation, an image differentiated in the horizontal direction (X direction) is moving averaged in the vertical direction (Y direction), and an image differentiated in the vertical direction (Y direction) is horizontal (X Moving average).

  This is because the differential value largely fluctuates due to noise or the like included in the tomographic image only by differentiating, and thus the fluctuation is suppressed by performing a moving average. By such calculation, the X component Gx and the Y component Gy of the gradient of each point of the tomographic image are obtained.

  The gradient calculation means shown in FIG. 3 (B) performs the moving average first and then performs the differential calculation. Thus, the order of the moving average and the differential calculation may be reversed.

  4C and 4D are respectively a differential operator that differentiates in the X direction and performs an averaging operation in the Y direction, and a differential operator that differentiates in the Y direction and performs an averaging operation in the X direction. FIG.

  Instead of performing the differential operation and the moving average separately, such a differential averaging operator may be applied to the tomographic image.

  FIG. 5 is a schematic diagram showing the values of the image data of each point (X, Y) on the tomographic image as a graph. FIG. 5A shows image data before differential calculation, and FIGS. 5B to 5D show image data after differential calculation.

  Consider a case in which image data whose value has changed in the X direction is differentiated in the X direction as shown in FIG. At this time, if the width d of the differential filter shown in FIG. 4B is too narrow, it is small in that the image data of the contour of the left ventricle and the myocardium changes gently as shown in FIG. 5B. Only differential data of values can be obtained. On the other hand, if the width d is too wide, the differential peak is not clear as shown in FIG. In the experiment by the present inventor, the width d of the differential filter is selected to be not less than 0.5 times and not more than twice the width of the narrow myocardium of the left ventricle and the myocardium that touch the contour to be obtained here. Thus, it was found that appropriate differential data as shown in FIG.

  In this embodiment, as an example, an image of 640 × 480 dots and 1 bit (256 gradations) is captured as a tomographic image, and the width of the differential data (number of dots 2n + 1) shown in FIG. Almost the same 21 dots are selected. As shown in FIGS. 4C and 4D, in this embodiment, 2n + 1 = 21 (dots) is selected as the moving average width.

  FIGS. 6A to 6C are diagrams showing examples of the scalar quantity calculation means.

  In FIG. 6A, the gradient X component Gx and Y component Gy are squared to obtain Gx2 and Gy2, and these are added to further calculate the square root to obtain the absolute value of the gradient (Gx2 + Gy2) 1 / 2 is calculated.

  In FIG. 6B, the square root calculation is omitted as compared with FIG. 6A, and the square of the gradient, that is, Gx2 + Gy2 is calculated.

  In FIG. 6C, the absolute values | Gx | and | Gy | of the gradient X component Gx and Y component Gy are calculated and added together to calculate | Gx | + | Gy |. .

  As shown in each example here, as the scalar quantity referred to in the present invention, various computation quantities corresponding to the magnitude of the gradient can be adopted.

  FIG. 7 is a flowchart showing data processing in the local maximum point calculation means, and FIG. 8 is an explanatory diagram thereof.

  In this local maximum point calculation means, first, referring to the magnitude of the gradient calculated by the scalar quantity calculation means, a difference is taken radially along each specified line shown in FIG. Thus, the position where the gradient magnitude becomes maximum, that is, the position where the sign of the difference in gradient magnitude changes from positive to negative is detected, and thus the maximum point considered to be on the left ventricular contour is detected. Candidate points are extracted. However, when the gradient varies, there is a problem that points that are not clearly points on the contour are selected as candidate points. Therefore, when the difference value of the gradient magnitude changes from positive to negative, for the candidate points extracted by the change, the average value of the gradient magnitude difference values of the points before and after the candidate point is calculated. After replacing with a new difference value, positive / negative determination is performed again, and only when a change in sign is detected again at that time, it is extracted as a candidate point.

For example, in FIG. 8, A _n is the gradient magnitude, and S _n = A _{n + 1} −A _n is the gradient magnitude difference. A ₅ is extracted once as a candidate point for the maximum point because the signs of S ₄ and S ₅ are different, but if S ₅ ′ = (S ₄ + S ₆ ) / 2 is substituted, S ₄ and S ₅ ′ are Since it has the same sign, this candidate point is not extracted as a maximum point. On the other hand, A ₈ have different sign of S ₇ and S _8, S ₈ = as _{(S 7 + S 9) /} 2, S 7 and S ₈ 'is extracted as the maximum point for the opposite sign.

  FIG. 9 is a diagram showing various feature amounts obtained by the contour extracting means.

  When extracting the contour of the left ventricle in the contour extraction means, “A: size”, “B: average value”, “C: inside / outside determination” for each of the maximum points extracted as described above. , “D: angle difference” and “E: distance” are obtained, and these feature quantities are multiplied by each other to calculate each total feature quantity for each local maximum point.

The feature quantity “A: size” refers to the magnitude of the gradient. Here, for example, the square of the absolute value of the gradient G _x ² + G _y calculated according to the operator shown in FIG. 6B. ² is adopted.

  FIG. 10 is a calculation block diagram for calculating the feature quantity “B: average value”.

  Here, for both the X direction and the Y direction of the tomographic image, the image data is subjected to moving average for 21 dots which is the same as the width d of the differential filter (see FIG. 4B), and the average value B is obtained. The order of the moving averages may be performed first in the X direction as shown in FIG. 10A, or may be performed first in the Y direction as shown in FIG.

  That is, for each pixel, as shown in the following equation (10), the surrounding image data for ± 10 dots in the X direction and ± 10 dots in the Y direction are added.

  Mathematically, the average value needs to be divided by the number of pixels in the above equation. However, since the number of pixels to be calculated is fixed, the added value is equivalent to the average value in data processing. In order to reduce the amount of calculation, division by the number of pixels is omitted.

  FIG. 11 is an explanatory diagram of the calculation of the feature amount “C: inside / outside determination”.

Here, the gradient vector (G _x , G _y ) of the maximal point and the direction vector (X-X0, Y-) from the center point (X0, Y0) (see FIG. 2) to the position (X, Y) of the maximal point. Whether the gradient vector is inward or outward as viewed from the center point (X0, Y0) is determined based on whether the inner product (scalar product) is positive or negative.

That is,
(Inner product) = (X−X0) · G _x + (Y−Y0) · G _y (11)
To calculate
If (inner product)> 0, it is outward as shown in FIG.
If (inner product) ≦ 0, it is inward as shown in FIG.

  Eventually, all the individual feature quantities are multiplied in order to obtain a total feature quantity, so that a point having an inward gradient vector is rejected. This is because only the inner wall of the myocardium on the left ventricle side is extracted as a contour, and the outer wall of the myocardium is not extracted as a contour.

  FIG. 12 is a calculation block diagram for obtaining the feature amount C.

  The difference (X−X0, Y−Y0) between the coordinates (X, Y) of the local maximum point and the coordinates (X0, Y0) of the center is obtained, and the gradient of the X component X−X0, Y component Y−Y0 The X component Gx and the Y component Gy are multiplied and added to each other, and their signs are determined. If the sign is positive, the gradient is outward C = 1, and if the sign is negative, the gradient is inward. C = 0 is output.

  FIG. 13 is an explanatory diagram of the calculation of the feature amount “D: angle difference”.

Here, it is assumed that the calculation is performed sequentially for a plurality of radial designated lines shown in FIG. 2, and as shown in FIG. 13A, the gradient vector G1 = (G _x1,. An angle dθ formed by G _y1 ) and the gradient vector G2 = (G _x2 , G _y2 ) of each local maximum point # 1 to # 5 on the specified line is calculated. Since the directions of two adjacent ridgelines on the contour line are considered to be substantially the same, here, the feature quantity D, which increases in value as the angle difference dθ decreases, is output.

Here, as shown in FIG. 13B, if the gradient angles are θ1 and θ2, respectively, the gradient vectors are
G1 = | G1 | exp (jθ1)
G2 = | G2 | exp (jθ2)
It becomes. However, j represents an imaginary unit. Here, when the complex conjugate of G1 is multiplied by G2,
G1 ^* · G2 = | G1 | exp (−jθ1) × | G2 | exp (jθ2)
= | G1 || G2 | exp (j (θ2-θ1)
Therefore

It becomes. However, * represents a complex conjugate.

On the other hand, the gradient vector is
G1 = Gx1 + jGy1
G2 = Gx2 + jGy2
Because
G1 ^* .G2 = (Gx1-jGy1) (Gx2 + jGy2)
= Gx1, Gx2 + Gy1, Gy2
+ J (Gxl · Gy2−Gx2 · Gy1)
The real component and imaginary component of G1 ^* · G2 are
Real (G1 ^* · G2) = Gx1 · Gx2 + Gy1 · Gy2
Imag (G1 ^* · G2) = Gx1 · Gy2−Gx2 · Gy1
It turns out that it is. From this, the desired angle dθ is

It becomes.

  As shown in FIG. 13C, the feature quantity D is defined by the following equation so that the maximum value π is taken when the angle difference dθ is 0, and it approaches 0 as the angle difference dθ approaches π or −π.

D = π− | dθ | (14)
FIG. 14 is a calculation block diagram for obtaining the feature amount D.

X = Gx1 * Gx2 + Gy1 * Gy2
Y = Gx1 * Gy2-Gx2 * Gy2
Is calculated,
dθ = atan (Y / X)
Thus, the angle dθ according to the equation (13) is obtained.

  FIG. 15 is an explanatory diagram of the calculation of the feature amount “E: distance”.

  As shown in FIG. 15 (A), the intersection of the current designated line with the line extending in the direction perpendicular to the gradient of the contour point from the position (X1, Y1) of the contour point on the previous designated line ( Xr, Yr), the distance dr to the position of each local maximum point on the specified line is obtained, and the feature quantity E, the value of which increases as the distance dr becomes shorter, is output. For example, in FIG. 15A, since the distance dr of the local maximum point # 2 is the shortest, the feature amount E of the local maximum point # 2 takes the largest value.

  First, a reference length, that is, a distance r between the center point (X0, Y0) and the intersection point (Xr, Yr) is obtained.

  In FIG. 15B, (X0, Y0) is the center point of the radial line segment, (X1, Y1) is the contour point of the previous designated line, (Gx, Gy) is the gradient vector of the contour point, (Xr , Yr) is the intersection of the current designated line and the straight line that passes through the contour point on the previous designated line and is orthogonal to the gradient of the contour point, and θ is the current designated line with respect to the X axis. Is the angle. When the ratio R of the X component and the Y component of the vector of the thick line from (Xr, Yr) to (X1, Y1) is expressed by an equation:

It is.

By transforming this equation,
−Gx (X1−X0−rcos θ) = Gy (Y1−Y0−rsin θ)
r (Gxcos θ−Gysin θ) = Gx (X1−X0)
+ Gy (Y1-Y0)
Therefore,

It becomes. The coordinates (Xr, Yr) of the reference intersection are
Xr = X0 + rcos θ,
Yr = Y0 + rsinθ (16)
Can be expressed as

When the distance dr from this coordinate is determined for each local maximum point (X, Y),
dr = {(X−Xr) ² + (Y−Yr) ² } ^1/2 (17)
It becomes. In this case, the candidate point # 2 is considered to be the closest point.

  The feature amount E is an exponential function such that the output value is halved as shown in FIG. 15C every time the distance becomes longer by a predetermined number of dots dr0 (for example, dr0 = 20 (dots)). That is, the feature quantity E is

FIG. 16 is a calculation block diagram for obtaining the distance r and the coordinates (Xr, Yr) of the intersection, and FIG. 17 is a calculation block diagram for obtaining the distance dr.

As shown in FIG. 16, here, the coordinates of the center point (X0, Y0), the coordinates (X1, Y1) of the contour point on the previous designated line and the gradient (Gx, Gy) of the contour point, and the current time The angle θ of the specified line is input,
A = Gx (X1-X0) + Gy (Y1-Y0)
B = Gx cos θ−Gysin θ
Is calculated,
r = A / B
Is calculated, and the distance r according to the equation (15) is obtained.

Furthermore, this distance r is multiplied by cos θ and sin θ, respectively, and X0 and Y0 are added, respectively, as shown in equation (16).
X _r = X0 + _r cos θ
Y _r = Y0 + rsinθ
Is required.

Next, as shown in FIG. 17, the coordinates (Xr, Yr) of the intersection point obtained as described above and the coordinates (X, Y) of the local maximum point on the specified line are input, and the equation (17) is entered. Show,
dr = {(X−Xr) ² + (Y−Yr) ² } ^1/2
Is required. This distance dr is further converted in accordance with equation (18) to obtain the feature quantity E.

  The feature amounts A to E obtained as described above are multiplied by all of them as shown in FIG. 9, and a comprehensive feature amount N is calculated for each local maximum point. In calculating the comprehensive feature quantity N, any one of the above-described feature quantities A to E may be omitted, or other feature quantities may be taken into account.

  FIG. 18 is a circuit block diagram for extracting one contour point for each designated line based on the total feature amount N calculated as described above.

  In this figure, n indicates the number assigned to the maximum point for each designated line.

  For one specified line, the coordinates (Xn, Yn) of the local maximum point and the feature quantity Nn of the local maximum point are sequentially input. The feature amount Nn is input to both the comparator 11 and the selector 12. The values stored in the register 13 are also input to the comparator 11 and the selector 12, and the comparator 11 compares them. The comparison result is input as a control signal to the selector 12, and a larger value of the two values compared by the comparator 11 passes from the selector 12 and is stored in the register 13.

  Further, the coordinates (Xn, Yn) of the local maximum point are input to the selector 14, and the coordinate value stored in the register 15 is also input to the selector 14, and the contents of the register 15 are changed according to the comparison result in the comparator 11 this time. The coordinates (Xn, Yn) are updated or the previous contents of the register 15 are held.

  When the coordinates (Xn, Yn) of the local maximum point on one designated line and the feature value Nn of the local maximum point are sequentially input in this way, the register 15 has the maximum of the local maximum points on the designated line. The coordinates (X, Y) of the maximum point having the feature amount are stored, and the maximum point having the maximum feature amount is extracted as a contour point.

  After extracting the maximum point having the maximum feature amount on one designated line as described above, the extracted maximum point and the designated line adjacent to the one designated line are extracted in the same manner. When the distance to the already determined contour point is calculated and the distance is a large distance that is more than a predetermined distance, that is, the extracted local maximum point is adjacent to it even though the contour line is continuous. When the distance is beyond the distance that can be considered to be continuous with the contour point on the designated line, the extracted maximum point is stopped as a contour point, and the feature on the one designated line The maximum point next to the extracted maximum point is extracted again, the same determination as above is performed, and this is repeated as necessary, so that the maximum point within a predetermined distance is used as the contour point. Like to extract Good. When comprised in this way, an outline can be extracted with higher precision.

  When finding the contour point on the first specified line, if there is a local maximum point on the first specified line that has a large feature value compared to the other local maximum points, The maximum point may be determined as the contour point, and then the contour point on the adjacent designated line may be determined sequentially according to the above method. Alternatively, for the first one designated line, the maximum point with the maximum feature amount is displayed on the display screen. And the operator determines whether or not the local maximum point exists on the contour. When it is determined that the local maximum point does not exist on the contour, a sequence for displaying the local maximum point having the next largest feature amount on the display screen is displayed. The contour point on the first specified line may be obtained repeatedly, or the contour point on the first specified line may be calculated by the operator without performing the above feature amount calculation. An outline point may be designated.

  Also, between a plurality of frames representing the same tomographic plane at different times, after a contour point is determined for a certain frame (previous frame), each local maximum is determined as described above for the next frame (current frame). Find the feature value for the point, find the local maximum point with the maximum feature value for each designated line, and on the same designated line when these two frames are overlapped, the contour points on the previous frame that have already been established And the maximum point having the maximum feature amount obtained for the current frame, and depending on whether or not the distance is within a predetermined distance, the maximum having the maximum feature amount, respectively. A sequence that either determines a point as a contour point, or rejects the maximum point with the largest feature value and finds the maximum point with the next largest feature value on the specified line. It may be returned Ri. Even in such a configuration, the contour line can be extracted with higher accuracy. As for the method for determining the contour point of the first frame, the method can be selected from various methods as described above.

  As a result of rejecting the maximal point having a large feature amount as described above, when an appropriate maximal point cannot be found within a predetermined distance, the contour point on the adjacent designated line is further connected with a straight line, The intersection of the straight line and the designated line where the local maximum point could not be found may be found, and the intersection may be defined as a contour point on the designated line, or it is determined that there is no contour point on the designated line. An outline based only on the outline point may be determined.

  One of the causes for reducing the accuracy of contour extraction is the presence of a valve between the left ventricle and the left atrium. As a method for preventing a decrease in the accuracy of contour extraction due to the presence of this valve, the above-described method for determining the magnitude of the distance can be used.

  FIG. 19 is a diagram illustrating one technique for preventing a decrease in contour extraction accuracy due to the presence of a valve. In each of FIGS. 19A to 19C, + indicates a center point.

  First, it is assumed that contour candidate points as shown in FIG. 19A are obtained on each designated line as described above, and it is determined whether or not the distance between adjacent candidate points is within a predetermined distance. As a result, as shown in FIG. 19B, candidate points at both ends corresponding to the roots of the two valves are obtained, and candidate points in the region sandwiched between the candidate points at both ends are rejected, and the region Only candidate points existing in the area excluding are valid. Therefore, as shown in FIG. 19C, in this embodiment, the candidate points at both ends are connected by a straight line to complete a closed curve as an outline.

  By doing so, it is possible to prevent a decrease in contour extraction accuracy due to the presence of the valve.

  FIG. 20 is a diagram illustrating another method for preventing a decrease in contour extraction accuracy due to the presence of a valve.

  Here, the distance between the previous frame candidate point (FIG. 20A) and the current frame candidate point (FIG. 20B) on the same designated line is examined, and the distance is within a predetermined distance. The candidate points that are more than a predetermined distance away are rejected (FIG. 20C). In this way, a correct contour is obtained as in FIG. 19 (FIG. 20D).

  FIG. 21 is a diagram showing the contour of the left ventricle obtained by connecting the contour points on each designated line obtained as described above.

  Schematically, the contour points may be connected by straight lines as shown in FIG. 21A, or more precisely, by using a spline function as shown in FIG. 21B.

  In the contour display means shown in FIG. 1, a contour line as shown in FIG. 21A or FIG. 21B is displayed superimposed on the tomographic image. Alternatively, instead of displaying the contour line, as shown in FIG. 21C, display may be performed in which the inside and outside of the contour line are divided by color, luminance, and the like.

  FIG. 22 is a diagram illustrating an example of how to obtain the area of the left ventricle that appears in the tomographic image.

  Since the area of the triangle can be obtained if the lengths a and b of the two sides and the angle between them are known as shown in FIG. 22 (A), the area of the triangle is added once as shown in FIG. 22 (B). The approximate area of the left ventricle is determined.

  The internal (left ventricle) area may be obtained by integrating the area enclosed by the contour line, but the left ventricular area can be obtained with sufficient accuracy and high speed even with the above calculation method. Can be sought.

  FIG. 23 is a diagram illustrating an example of how to determine the volume of the left ventricle.

  After obtaining the contour line as described above, the sum of the distances between the center line (XO, YO) and the contour point of the two designated lines extending in the opposite directions from the center point (XO, YO) is the maximum. The two specified lines are extracted, and the volume inside the rotating body when the contour line is rotated with the specified lines as the rotation axis is calculated. This is based on the finding that the left ventricle is almost elliptical as a whole, and the volume calculated in this way is regarded as the volume of the left ventricle. As a volume calculation method, for example, the Tichkolz method, the Pombo method, the Gibson method, and the like are known, and any of them may be adopted.

  FIG. 24 is a diagram showing an example of a display method of the area or volume obtained as described above.

  Since the tomographic images are sequentially obtained at a predetermined frame rate as time elapses, the area or volume is obtained for each of the tomographic images, and the area or volume is displayed as a function of time as shown in FIG. The degree can be understood at a glance and is more effective for observation and diagnosis. Of course, the area or volume may be displayed numerically.

  FIG. 25 is a diagram illustrating another example of the designation line.

  In the above-described embodiment, as shown in FIG. 2, a designation line extending from the center line (XO, YO) to the radiation is employed. However, in the present invention, it is not always necessary to employ such a radial designation line. Instead, for example, a designation line extending parallel to the X direction as shown in FIG. 25A or a designation line extending parallel to the Y direction as shown in FIG. 25B may be adopted.

  As shown in FIGS. 25A and 25B, when the designated line parallel to the X direction or the Y direction is adopted, the maximum point on the designated line is compared with the case where the radial designated line as shown in FIG. 2 is adopted. Calculation for obtaining coordinates is simplified.

  Further, as shown in FIG. 25C, the operator may designate a rough major axis on the tomographic image display screen, and a designated line extending in a direction orthogonal to the designated major axis may be employed.

  Furthermore, it is not always necessary to use the designated line in the first ultrasonic diagnostic apparatus of the present invention. For example, the approximate position of the contour is obtained by using a conventional method or another simplified method, and the approximate contour is followed. In this way, a large number of small areas are set on the tomographic image, maximum points and feature quantities are obtained in each of the small areas, and the maximum points where the feature quantities are maximized are extracted for each small area. The contour may be obtained by connecting.

  Next, an example of the second ultrasonic diagnostic apparatus of the present invention will be described.

  FIG. 26 is a diagram showing a flow for extracting and displaying an outline in an embodiment of the second ultrasonic diagnostic apparatus of the present invention.

  In this flow, the only difference from the flow in the embodiment of the first ultrasonic diagnostic apparatus shown in FIG. 1 is that the calculation effective area detection means is provided. The description will focus on the effective area detection means.

  In the calculation effective area detecting means, as described with reference to FIGS. 11 and 12, whether the gradient direction is inward facing toward the center point (X0, Y0) (FIG. 11A). Whether the pixel faces outward from the center point (X0, Y0) (FIG. 11B) is determined for each pixel over the entire tomographic plane. The calculation effective area detection means further creates a binary image based on the determination result, and determines the calculation effective area based on the binary image.

  FIG. 27 is an explanatory diagram showing how to determine the calculation effective area.

FIG. 27A shows a binary image created based on the gradient direction as described above, and a connection process by labeling according to the following algorithm is performed on the binary image. In this binary image, the x mark represents a pixel whose gradient direction is inward.
(1) When assigning a numerical value label to a certain pixel, among the pixels having the same value in the adjacent pixels (the same gradient direction), a label having the same numerical value as the pixel with the smallest numerical value is attached. Attached.
(2) If there is no pixel having the same value in the adjacent pixel, or if there is a pixel having the same value in the adjacent pixel but the adjacent pixel having the same value is not yet labeled, Add a label.
(3) When different numerical labels are attached to a plurality of adjacent pixels having the same value, the labels having the smaller numerical value are unified.

  FIG. 27B shows a labeled image in which each pixel is labeled in accordance with the above algorithms (1) to (3). In the labeling image created in this way, the outer periphery of the connecting portion (the portion labeled “1” in this example) surrounding the center point (+) and having the gradient direction inward is shown in FIG. As shown in C), the boundary is defined as the boundary of the computation effective area, and even if there is a part with the gradient direction outward in the area surrounded by the boundary, the boundary is surrounded by the boundary. The entire area is defined as the calculation effective area.

  FIG. 28 is an explanatory diagram showing how to obtain the contour in the contour computing means shown in FIG.

  When the contour point (maximum point having the maximum feature amount) is obtained by ignoring the calculation effective area on each designated line shown in FIG. 2, the calculation effective area is out as shown in FIG. When a contour point is obtained for a point, the contour point is rejected, and a new contour point is obtained within the calculation effective area as shown in FIG. 28 (B), and the contour as shown in FIG. 28 (C) is obtained. Extracted.

  According to the flow shown in FIG. 26, since the calculation effective area detecting means as described above is provided, a more accurate contour is obtained.

  FIG. 29 is a diagram showing another flow for extracting and displaying the contour in the second embodiment of the ultrasonic diagnostic apparatus of the present invention.

  In this flow, the difference from the flow shown in FIG. 26 is that the flow shown in FIG. 26 is provided with a binarizing means instead of the scalar quantity calculating means and the maximum point calculating means. . In this binarization means, image data (luminance) for each pixel is compared with a predetermined threshold value and binarized, and an outline is obtained based on the binary image.

  FIG. 30 is an explanatory diagram showing how to obtain a contour based on a binary image.

  As shown in FIG. 30 (A), the blood flow portion indicated by the binary image extends inside and outside the boundary line of the calculation effective region, and as shown in FIG. 30 (B), the boundary line in the blood flow portion. Only the inner part is extracted, and the contour is defined so as to follow the outer periphery of the blood flow part (FIG. 30C).

  In this way, by defining the calculation effective area, it is not necessary for the operator to input the approximate area 3 as described with reference to FIG. 46, and therefore, an erroneous extraction of the contour due to an input error is prevented.

  Next, an embodiment of the third ultrasonic diagnostic apparatus of the present invention will be described.

  FIG. 31 is a schematic block diagram of an embodiment of the second ultrasonic diagnostic apparatus of the present invention.

  The ultrasonic diagnostic apparatus includes an ultrasonic probe 101 that transmits and receives ultrasonic waves. Transmission pulses are transmitted from the transmission circuit 102 toward the ultrasonic probe 101, and the subject (not shown) is transmitted from the ultrasonic probe 101. ) Ultrasonic waves are transmitted inward. The ultrasonic wave reflected and returned in the subject is received by the ultrasonic probe 101, and the received signal is input to the receiving circuit 103. The receiving circuit 103 receives the received signal so that a scanning line is formed in the subject. Are delayed and added. A signal output from the receiving circuit 103 is input to the signal processing circuit 104. The signal processing circuit 104 includes a log amplifier, a detection circuit, a filter processing circuit, an A / D converter, etc., and performs necessary signal processing and converts it into a digital signal (image data) for output. Are input to a digital scan converter (DSC) 105. The DSC 105 converts ultrasonic scanning line direction data into TV screen scanning line data. The image data output from the DSC 105 is input to display means 111 having a CRT, and a tomographic image is displayed on the display screen. The image data output from the DSC 105 is also input to the initial contour extraction means 107 (an example of the contour calculation means referred to in the present invention). In this initial contour extracting means 107, for example, the contour of the left ventricle of the heart is obtained for the frame specified by the frame specifying means 112, for example, in the same manner as the contour extracting method in the first ultrasonic diagnostic apparatus of the present invention described above. It is done. The frame designation means 112 designates a frame based on an electrocardiogram or a Doppler waveform. Alternatively, the display screen may be frozen and the operator may directly specify the frame. The contour obtained by the initial contour extraction means 107 is input to the display means 111, and the contour is displayed superimposed on the tomographic image. The contour obtained by the initial contour extracting means 107 is also input to the contour deforming means 109 described later.

  The image data output from the DSC 105 is also input to the preprocessing unit 106. The pre-processing means 106 performs pre-processing for calculating the motion in the motion detection means 108 for obtaining the motion of each point of the subject in the tomographic plane, and the motion detection means 108 performs the motion using the optical flow method. If it is to be obtained, the preprocessing means 106 performs smoothing processing, and if the motion detection means 108 is to obtain motion using the cross-correlation method, the preprocessing means 106 performs binarization processing. Is called.

  The image data preprocessed by the preprocessing unit 106 is input to the motion detection unit 108, and the motion of the subject in the tomographic plane is obtained. This movement information is input to the contour deformation means 109. As described above, the contour obtained by the initial contour extracting unit 107 is also input to the contour deforming unit 109, and the contour is deformed so as to match the contour of the frame currently processed by the contour deforming unit 109. The The contour deforming unit 109 does not return to the contour extracted by the initial contour extracting unit 107 for each frame and deforms the contour, but by further sequentially deforming the contour once deformed by the contour deforming unit 109. The contour of the frame being processed is determined. The contour deformed by the contour deforming unit 109 is input to the display unit 111 and displayed on the display screen so as to be superimposed on the tomographic line of the corresponding frame. The contour output from the contour deforming unit 109 is also input to the guidance amount calculating unit 110. Based on the input contour, the guidance amount calculation unit 110 obtains the guidance amount derived from the contour, such as the area inside the contour, the center of gravity of the contour, the volume inside the contour, and the like. Further, if necessary, a difference value over a plurality of frames of the guidance amount is obtained. These induction amounts or difference values are input to the display unit 111, and the display unit 111 displays the induction amounts or difference values by numerical values, graphs, or the like, for example.

  In the configuration shown in FIG. 31, the tomographic image output from the DSC 105 and input to the display unit 111, the contour output from the initial contour extraction unit 107, the contour input to the display unit 111, and the display output from the contour deformation unit 109 The contour input to the means 111 and the guidance amount output from the guidance amount calculation means 110 and input to the display means 111 or the difference value thereof are adjusted so that the quantities relating to the same frame are simultaneously input to the display means 111. Has been made.

  FIG. 32 is a block diagram showing a specific example of the preprocessing means 106 of the ultrasonic diagnostic apparatus shown in FIG.

  The image data output from the DSC 105 is input to a two-dimensional FIR filter 106a, which is an example of the preprocessing unit 106, and is smoothed. If the motion detection unit 108 detects motion by performing a two-dimensional cross-correlation calculation, smoothing processing is not necessarily required, but the motion detection unit 108 detects motion by performing an optical flow calculation. If it is, the optical flow method is a method on the assumption that the input image is spatially smooth. Therefore, it is preferable to perform a smoothing process as a preprocessing by the two-dimensional FIR filter 107a or the like.

  FIG. 33 is a block diagram showing another specific example of the preprocessing means 106.

  33 includes a two-dimensional FIR filter, a histogram analysis unit, a memory (MEM), and a binarization processing unit. A two-dimensional FIR filter is not always necessary.

  The histogram analysis unit analyzes how the luminance values of the image are distributed by creating a histogram of the image data, and processes the data so that the luminance distribution falls within a predetermined range and stores it in the memory belongs to. The binarization processing unit binarizes the image data that has been subjected to such processing at a predetermined threshold level. Such binarization processing is useful in that the calculation speed can be improved when the subsequent motion detection means 108 detects motion by cross-correlation calculation.

  FIG. 34 is a block diagram showing an example of a detailed configuration of a part of the ultrasonic diagnostic apparatus shown in FIG.

  As described above, the DSC 105 converts ultrasonic scanning line data into TV scanning line data, and includes an address generation unit, an interpolation calculation unit, and a memory (MEM). The address generation unit designates data used for the interpolation calculation and designates the memory (MEM) address. In the interpolation calculation unit, TV scanning line data is generated by performing interpolation calculation on the inputted ultrasonic scanning line data. The generated TV scanning line data is temporarily stored in the memory. This memory has a capacity capable of storing, for example, 64 frames of image data. When data exceeding the number of frames is input, the memory is overwritten sequentially, and old data is sequentially deleted.

  The preprocessing means 106 performs appropriate spatial filtering processing, etc., and the two-dimensional motion vector calculation unit 108a, which is an example of the motion detection means 108 shown in FIG. 31, uses the image data of temporally adjacent frames. A two-dimensional motion vector of each of a plurality of points in the plane is calculated.

  The initial contour extraction unit 107 includes a luminance gradient (gradient) calculation unit and a contour extraction unit. In the luminance gradient (gradient) calculation unit, the frame specified by the frame specification unit 112 (this is expressed as (Hereinafter referred to as “initial frame”), a spatial luminance gradient (gradient) of each point in the initial frame is calculated, and the contour extraction unit calculates a predetermined intensity in the initial frame based on the calculated gradient. The outline of the tissue, for example the left ventricle, is extracted. The contour in the initial frame is hereinafter referred to as “initial contour”.

  The contour deforming unit 109 sequentially obtains contours in frames that are sequentially adjacent to the initial frame using the calculated initial contour and the two-dimensional motion vector.

  FIG. 35 is a schematic diagram showing a tomographic image of the heart for illustrating the calculation contents in the two-dimensional motion vector calculation unit 108a and the contour deforming unit 109 shown in FIG.

  A contour indicated by a thick line in FIG. 35 is an initial contour or a contour deformed in the immediately preceding frame. The black circle points indicate the points inside the myocardium in the vicinity of the contour where the two-dimensional motion vector calculation unit 108a calculates the two-dimensional motion vectors, and the arrows extending from the black circle points indicate the calculated two-dimensional motion vectors. . A contour indicated by a dotted line represents a contour deformed based on the calculated two-dimensional motion vector. In FIG. 35, white circles are also seen, but the points of the white circles are out of the myocardium and are very weak points of reflected ultrasound. Therefore, the two-dimensional motion vector obtained with respect to this point shows an accurate motion. Since there is a high possibility that it is not represented, it is not involved in the deformation of the contour based on the image luminance information. The deformed contour is preferably obtained by fitting to a spline function or the like.

  FIG. 36 is a block diagram showing another example of the detailed configuration of a part of the ultrasonic diagnostic apparatus shown in FIG. 31, and FIG. 37 is an explanatory diagram of a method for designating a motion detection direction in advance.

  36 is different from the configuration shown in FIG. 34 in that the motion detecting unit 108b includes a direction specifying unit and a specified direction motion calculating unit.

  The direction designating unit designates a predetermined center point inside the contour as shown in FIG. The center point may be specified manually by the operator, or the center of gravity of the initial contour already obtained or the contour of the previous frame is obtained and the center of gravity is automatically specified as the center point. May be.

  In the designated direction motion calculation unit shown in FIG. 36, the magnitude of the motion in the direction toward the designated center point is obtained as shown in FIG. In this case, unlike the calculation of the two-dimensional motion vector in the two-dimensional motion vector calculation unit 108a shown in FIG. 34, a one-dimensional calculation in the direction toward the center point is sufficient. The method of deforming the contour is the same as described above.

  FIG. 38 is a block diagram showing another example of the motion detection means 108 shown in FIG. 31, and FIG. 39 is an explanatory diagram thereof.

  The motion detection unit 108c shown in FIG. 38 includes a motion detection unit, a region division unit, and a region average processing unit. For example, as shown in FIG. 39, the region dividing means performs designation so as to divide into eight in the radial direction in the vicinity of the contour. The motion detection unit calculates a motion vector from adjacent frames at a plurality of points in each region. In the area average processing unit, the motion vector is calculated with reference to the image data output from the preprocessing unit 106 (the image data may be output from the DSC 105 without passing through the preprocessing unit 106). If the luminance value at a point is greater than a predetermined value, it is determined that the motion vector at that point is valid (corresponding to the black dot in FIG. 39), while the luminance value is smaller than the predetermined value Is determined that the motion vector at that point is invalid (corresponding to the white circled point in FIG. 39), and an average motion vector of the valid motion vectors is obtained (corresponding to the arrow in FIG. 39). . The contour deforming means 109 deforms the initial contour or the contour of the immediately preceding frame based on these average motion vectors, and obtains a contour suitable for the frame.

  FIG. 40 is a partial block diagram showing a configuration in which an electrocardiogram means for inputting an electrocardiogram of a patient (subject) is added to the ultrasonic diagnostic apparatus shown in FIG. 31, and FIG. 41 is a schematic diagram of an electrocardiogram waveform.

  In the electrocardiogram means 113 shown in FIG. 40, an electrocardiogram as shown in FIG. 41 is inputted and sent to the display means 111. On the display screen, an electrocardiogram as shown in FIG. 41 is displayed together with a normal ultrasonic tomogram. The

  The electrocardiogram input by the electrocardiogram means 113 is input to the frame designation means 112, and the frame designation means 112 detects an R wave (see FIG. 41) on the electrocardiogram, and a predetermined time has elapsed from the R wave. A point A frame is designated. The initial contour extracting unit 107 obtains the initial contour of the frame specified by the frame specifying unit 112 as described above.

  FIG. 42 is a partial block diagram showing a configuration in which a Doppler analyzer is added to the ultrasonic diagnostic apparatus shown in FIG.

  The ultrasonic diagnostic apparatus is usually provided with a function (Doppler analysis unit 114) for calculating, for example, a blood flow velocity by Doppler calculation. The Doppler analysis unit 114 receives information representing a temporal change in blood flow velocity. Is output. The temporal change in the blood flow velocity is synchronized with the heartbeat, and information on the blood flow velocity is input to the frame designating unit 112. The frame designating unit 112 replaces the electrocardiogram in FIG. Based on the information, the frame from which the initial contour is to be extracted is designated.

  FIG. 43 is a block diagram in which a freeze unit is added to the ultrasound diagnostic apparatus shown in FIG.

  As described above, the DSC 105 is provided with a memory for storing, for example, image data for 64 frames (see FIG. 34). When the operator performs a predetermined operation, the freeze means 115 saves the image data currently stored in the memory as it is to the DSC 105 and outputs a freeze command for prohibiting overwriting of new image data. The frame designating unit 112 designates a frame for which an initial contour is to be obtained based on an instruction from the operator or based on the electrocardiogram and blood flow velocity information described with reference to FIGS. From the DSC 105, the image data of the initial frame is output and sent to the display means 111, and the tomographic image of the initial frame is displayed on the display screen. The operator inputs a point (contour point) on the contour in the initial frame by using the contour input unit constituting the initial contour extraction unit 107a.

  The input contour point is input to the contour generation unit, and the contour generation unit obtains an initial contour based on the input contour point using, for example, a polygonal line or a spline curve. The obtained initial contour is input to the contour deforming unit 110, and the contour deforming unit 110 outputs the contour for the frames other than the initial frame out of the 64 frames stored in the DSC 105 and output from the DSC 105. Is required.

  FIG. 44 is a schematic diagram illustrating an example of how to obtain the volume inside the contour in the guidance amount calculation unit 110 of the ultrasonic diagnostic apparatus illustrated in FIG. 31.

  FIG. 44A shows an ellipse that approximates the outline of the left ventricle. By rotating the ellipse, a spheroid is obtained, and the volume inside the spheroid is obtained.

  FIG. 44 (B) shows an example in which the inside of the outline of the left ventricle is approximated by a large number of rectangles, and by rotating each rectangle around the central axis of each rectangle, a large number of disks are obtained, The sum of the volumes inside these disks is determined.

  For example, as shown in this example, the volume inside the contour is obtained for each frame, and the change in volume is displayed on the display screen as shown in FIG. 24, for example.

It is the figure which showed the flow which extracts and displays the outline in one Example of the 1st ultrasonic diagnosing device of this invention. It is a figure which shows the multiple (16 in this example) designation | designated line extended to a radiation from a predetermined | prescribed center point (X0, Y0). It is a figure which shows each example of a gradient calculating means. It is a figure which shows a differential filter. It is the schematic diagram which showed and graphed the value of the image data of each point (X, Y) on a tomogram. It is a figure which shows each example of a scalar quantity calculating means. It is a flowchart of the data processing in a maximum point calculating means. FIG. 8 is an explanatory diagram of data processing shown in the flowchart in FIG. 7. It is a figure which shows the various feature-value calculated | required in an outline extraction means. It is a calculation block diagram for calculating an average value. It is explanatory drawing of the calculation of inside / outside determination. It is a calculation block diagram for inside / outside determination. It is explanatory drawing of a calculation of an angle difference. It is a calculation block diagram for calculating | requiring an angle difference. It is explanatory drawing of calculation of distance. It is a calculation block diagram for calculating | requiring the distance r and the coordinate (Xr, Yr) of an intersection. It is a calculation block diagram for calculating | requiring the distance dr. It is a circuit block diagram for extracting one outline point for each designated line based on a calculated feature amount N. It is a figure which shows one of the methods of preventing the fall of the outline extraction precision by presence of a valve. It is a figure which shows another method which prevents the fall of the outline extraction precision by presence of a valve. It is a figure which shows the outline of the left ventricle calculated | required by connecting the outline point on each designation | designated line. It is a figure which shows an example of how to obtain | require the area of the left ventricle which appeared in the tomogram. It is a figure which shows an example of how to obtain | require the volume of a left ventricle. It is a figure which shows an example of the display method of an area thru | or volume. It is a figure which shows the other example of a designation | designated line. It is the figure which showed the flow which extracts and displays the outline in one Example of the 2nd ultrasonic diagnosing device of this invention. It is explanatory drawing which shows how to determine a calculation effective area | region. It is explanatory drawing which shows how to obtain | require the outline in the outline calculating means shown in FIG. It is the figure which showed the other flow which extracts and displays the outline in the Example of the 2nd ultrasonic diagnosing device of this invention. It is explanatory drawing which shows how to obtain | require the outline based on a binary image. It is a schematic block diagram of an embodiment of a third ultrasonic diagnostic apparatus of the present invention. It is a block diagram which shows the specific example of the pre-processing means of the ultrasonic diagnosing device shown in FIG. It is a block diagram which shows the other specific example of a pre-processing means. FIG. 32 is a block diagram showing an example of a detailed configuration of a part of the ultrasonic diagnostic apparatus shown in FIG. 31. FIG. 35 is a schematic diagram showing a tomographic image of the heart illustrating the calculation contents in the two-dimensional motion vector calculation unit and the contour deforming unit shown in FIG. 34. FIG. 32 is a block diagram showing another example of a detailed configuration of a part of the ultrasonic diagnostic apparatus shown in FIG. 31. It is explanatory drawing of the method of designating the detection direction of a motion beforehand. FIG. 32 is a block diagram showing another example of the motion detection means shown in FIG. 31. It is explanatory drawing of the motion detection means shown in FIG. FIG. 32 is a partial block diagram showing a configuration in which an electrocardiogram means for inputting an electrocardiogram of a patient (subject) is added to the ultrasonic diagnostic apparatus shown in FIG. 31. It is a schematic diagram of an electrocardiogram waveform. FIG. 32 is a partial block diagram illustrating a configuration in which a Doppler analysis unit is added to the ultrasonic diagnostic apparatus illustrated in FIG. 31. FIG. 32 is a block diagram in which a freeze unit is added to the ultrasonic diagnostic apparatus shown in FIG. 31 and a configuration example of an initial contour extraction unit is shown in blocks. FIG. 32 is a schematic diagram showing an example of how to obtain the volume inside the contour in the guidance amount calculation means of the ultrasonic diagnostic apparatus shown in FIG. 31. It is a figure which shows a mode that the tomogram displayed on the display is traced with a light pen. It is a figure which shows the area | region designated on the tomogram and its tomogram.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Display 2 Light pen 4 Left ventricle 11 Comparator 12,14 Selector 13,15 Register 101 Ultrasonic probe 105 Digital scan converter 106 Preprocessing means 107 Initial contour extraction means 108 Motion detection means 109 Contour deformation means 110 Guided amount calculation means 111 Display means 112 Frame designation means

Claims (13)

In the ultrasonic diagnostic apparatus for obtaining image data corresponding to each point in the tomographic plane extending in the subject by receiving the ultrasound reflected in the subject,
Motion calculating means for calculating the motion of a plurality of points in the tomographic plane based on the image data of a plurality of frames;
Contour computing means for obtaining a contour of a predetermined tissue in the subject based on the image data;
An ultrasonic diagnostic apparatus comprising: a contour deforming unit that obtains a contour deformed from the contour obtained by the contour calculating unit based on the motion detected by the motion calculating unit. The motion calculation means uses either one of an optical flow method and a cross-correlation method, and the motion includes a two-dimensional motion vector in a tomographic plane and a magnitude of motion in a predetermined direction in the tomographic plane. The ultrasonic diagnostic apparatus according to claim 1, wherein one of the two is calculated. The ultrasonic diagnostic apparatus according to claim 2, wherein the predetermined direction is a direction of a plurality of designated lines extending radially in a tomographic plane from a predetermined center point inside the contour. The motion calculation means calculates the motion of each of a plurality of points for each of a plurality of regions set in the tomographic plane, and extracts the calculated motion having a high probability of representing an accurate motion for each of the plurality of regions. The ultrasonic diagnostic apparatus according to claim 1, wherein a representative value of the extracted motion is obtained for each of the plurality of regions. Pre-processing means for performing one of smoothing and binarization on the image data;
The ultrasonic diagnostic apparatus according to claim 1, wherein the motion calculation unit calculates the motion based on the image data subjected to the one processing by the preprocessing unit. The contour calculating means is
A gradient calculation means for obtaining a gradient of the image data for a plurality of points in the tomographic plane;
Scalar quantity calculation means for obtaining a scalar quantity corresponding to the gradient for a plurality of points in the tomographic plane;
A maximum point calculation means for obtaining a plurality of maximum points in the fault plane, wherein the scalar quantity is a maximum;
The ultrasonic diagnostic apparatus according to claim 1, further comprising contour extraction means for obtaining a contour of a predetermined tissue in the subject based on the plurality of maximum points. A contour point designating operator for designating a contour point located on the contour of the predetermined tissue;
The ultrasonic diagnostic apparatus according to claim 1, wherein the contour calculation means obtains the contour based on the contour point designated by the contour point designating operator. Synchronization signal acquisition means for obtaining a heartbeat synchronization signal synchronized with the heartbeat of the subject,
The ultrasonic diagnostic apparatus according to claim 1, wherein the contour calculation means obtains the contour for a frame designated based on the heartbeat synchronization signal. A storage means for storing the image data of a plurality of frames in an overwriteable manner, and a freeze means for prohibiting overwriting to the storage means,
The contour calculation means obtains the contour for at least one predetermined frame among a plurality of frames stored in the storage means in a frozen state in which overwriting is prohibited by the freeze means;
The motion detection means calculates the motion for other frames excluding the predetermined frame among the plurality of frames stored in the storage means in the frozen state;
The ultrasonic diagnostic apparatus according to claim 1, wherein the contour deforming unit obtains a deformed contour from the contour obtained by the contour calculating unit for the other frame. The ultrasonic diagnostic apparatus according to claim 1, further comprising display means for displaying the contour. The guidance amount calculation means for obtaining a guidance amount calculated based on at least one of the contour obtained by the contour calculation means and the contour deformed by the contour deformation means. The ultrasonic diagnostic apparatus according to 1. 12. The guidance amount calculation unit obtains at least one of an area inside the contour, a center of gravity position of the contour, and a volume inside the contour as the guidance amount. The ultrasonic diagnostic apparatus as described. The ultrasonic diagnostic apparatus according to claim 11, further comprising display means for displaying at least one of the induction amount and a change amount of the induction amount over a plurality of frames.

Images