Ultrasound image processing apparatus and computer-readable non-transitory storage medium

Abstract

The present invention relates to an ultrasonic image processing apparatus and a computer-readable non-transitory storage medium. A variation amount information calculation section (46) calculates, for each voxel (V) constituting an ultrasonic volume data (Vol), variation amount information indicating a variation amount (V n ) between a signal intensity of a voxel of interest (Va) and a signal intensity of a nearby voxel (Vb) located in the vicinity of the voxel of interest (Va). An opacity calculation section (48) calculates, for each voxel (V), an opacity α(X n )*g(V n ) so that the smaller the variation amount V n between the signal intensities of the nearby voxels, the smaller the opacity with respect to the voxel (V). A rendering processing section (50) performs volume rendering based on the signal intensity X n of each voxel (V) and the calculated opacity α(X n )*g(V n ) of each voxel (V), thereby forming a three-dimensional image.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority to Japanese Patent Application No. 2022-104353 (filed on June 29, 2022), the entire contents of which, including the description, claims, drawings and abstract, are incorporated herein by reference. Technical Field

[0003] This specification discloses an ultrasonic image processing apparatus and an ultrasonic image processing program, and in particular discloses an improvement in volume rendering processing for ultrasonic volume data. Background Technology

[0004] An ultrasonic image processing device is a device that processes data obtained through the transmission and reception of ultrasonic waves. It consists of an ultrasonic diagnostic device or an information processing device, etc. The data that is processed by the ultrasonic image processing device is obtained by the ultrasonic diagnostic device.

[0005] In the past, ultrasound diagnostic devices generated three-dimensional images based on received signals obtained by transmitting and receiving ultrasound waves onto the subject. For example, in obstetrics, three-dimensional images representing the fetus were generated. Here, a three-dimensional image is an image that represents the subject (including biological organisms) in three dimensions or stereoscopically, but is physically two-dimensional. Based on a three-dimensional image, the subject (e.g., the fetus) can be represented with a sense of depth.

[0006] Three-dimensional images are formed by processing ultrasonic volume data constructed from received ultrasonic signals. Ultrasonic volume data is data in which voxels, representing the signal intensity of reflected waves from the object being examined, are arranged in three dimensions (e.g., orientation, slice direction, depth direction). In the past, as one of the methods for forming three-dimensional images from ultrasonic volume data, techniques such as volume rendering have been proposed (e.g., "Volume Rendering", Robert A. Drebin, et al., Computer Graphics, Volume 2, Number 4, August 1988; "Multi-Dimensional Transfer Functions for Interactive Volume Rendering", Joe Kniss, et al., Scientific Computing and Imaging Institute, School of Computing, University of Utah; and "Variational Classification for Visualization of 3D Ultrasound Data", Raanan Fattal, Dani Lischinski, School of Computer Science and Engineering, The Hebrew University of Jerusalem).

[0007] Figure 6 This is a diagram used to illustrate volume drawing. Volume drawing is mainly performed through the following process. First, multiple lines of sight, called rays R, are established from a certain viewpoint, passing through the ultrasonic volume data Vol. Then, based on each voxel V located on ray R... Figure 6 In the example, parameters called signal strength and opacity (opacity value) of V1, V2, ..., VN are used to perform cumulative processing on each ray R to obtain the calculated value. Specifically, cumulative processing characterized by Equation 1 is performed on each ray R.

[0008]

Mathematical Formula 1

[0009] C n -C n-1 +(1-α n-1 )*α(X n )*X n …(Equation 1)

[0010] In Equation 1, C nThis is the cumulative value (calculated value) when the operation is performed up to the nth voxel (voxel n) on ray R (observed from the viewpoint). α n The cumulative value of opacity up to voxel n, α(X) n The opacity of voxel n is represented by X. The cumulative value of the overall opacity with respect to a single ray R is set to 1. n The signal intensity characterizing voxel n. As shown in Equation 1, in volume rendering processing, the calculated value C up to the previous voxel n-1 on ray R is used. n-1 Add the amount of permeation up to voxel n-1 (in terms of 1-α) n-1 Characterization), the opacity α (X) of voxel n n The signal intensity X of voxel n n The product is used to calculate the operational value C of ray R. n That is, it can be said that the opacity α(X) of voxel n n The larger the value C, the more significant the operation. n The signal intensity X of voxel n n The greater the impact, the stronger the influence.

[0011] The cumulative opacity α up to voxel n n The operation is performed using Equation 2 below.

[0012]

Mathematical Formula 2

[0013] α n =α n-1 +(1-α n-1 )*α(X n ...(Equation 2)

[0014] That is, the cumulative opacity α up to voxel n. n This becomes the cumulative value α of the opacity up to voxel n-1. n-1 Add the value ((1-α) n-1 )*α(X n The value obtained is ((1-α)). n-1 )*α(X n The opacity α(X) is based on voxel n. n ) to perform calculations to obtain the cumulative value α of the ray's opacity. n The value obtained by becoming 1.

[0015] One ray R is equivalent to one pixel, and pixel values ​​are calculated based on this value. By calculating pixel values ​​with respect to multiple rays R, a three-dimensional image composed of multiple pixels is formed.

[0016] Furthermore, JP Patent No. 5622374 discloses an ultrasound diagnostic device that generates three-dimensional images from ultrasound volume data, and the user can use a user interface to adjust the opacity of the ultrasound volume data (containing individual voxels).

[0017] Sometimes, noise is contained in the ultrasound volume data obtained by transmitting and receiving ultrasound waves over a subject. Noise can be generated randomly from the electrical circuitry of the ultrasound diagnostic device, or from the tissues within the subject (e.g., noise from tiny objects such as floaters in amniotic fluid in the case of a fetus). Such noise has a certain signal strength in the voxels that constitute the ultrasound volume data. Therefore, if volume rendering is performed on the noisy ultrasound volume data to form a three-dimensional image, it can sometimes be affected by this noise, resulting in noise in the three-dimensional image as well.

[0018] The purpose of the ultrasonic image processing apparatus disclosed in this specification is to reduce the impact of noise in the three-dimensional image obtained by volume rendering of ultrasonic volume data, even when the ultrasonic volume data is noisy. Summary of the Invention

[0019] The ultrasonic image processing apparatus disclosed in this specification includes: a change amount information calculation unit, which calculates change amount information, representing the change between the signal intensity of a voxel of interest and the signal intensity of a neighboring voxel located near the voxel of interest, for each voxel constituting ultrasonic volume data obtained by transmitting and receiving ultrasonic waves to a subject; an opacity calculation unit, which calculates the opacity of each voxel based on the change amount information, such that the smaller the change in signal intensity between the voxel and the neighboring voxel as represented by the change amount information, the smaller the opacity of that voxel; and a rendering processing unit, which performs volume rendering of the ultrasonic volume data for each line of sight based on the signal intensity of each voxel and the calculated opacity of each voxel to obtain pixel values, and forms an ultrasonic image based on multiple pixel values ​​for each line of sight.

[0020] In ultrasound volume data, voxels corresponding to noise exhibit a smaller variation in signal intensity between the voxel and its neighboring voxels. Based on this structure, the smaller the variation in signal intensity between the voxel and its neighboring voxels, the lower the opacity of the voxel of interest. Therefore, in volume rendering, the influence of the signal intensity (noise component) of the voxel V corresponding to noise on the 3D image is reduced.

[0021] Alternatively, the opacity calculation unit may perform calculations with respect to each voxel based on the correction parameters of the change amount, and correct the opacity of each voxel based on the pre-determined opacity of each voxel and the correction parameters.

[0022] According to this structure, corrections can be made based on the pre-determined opacity of each voxel (based on the pre-determined opacity of each voxel) so that the smaller the change in signal intensity between the voxel and the neighboring voxel, the smaller the opacity of each voxel.

[0023] Alternatively, the opacity calculation unit can calculate the correction parameter by normalizing the change to a value between a given minimum and a given maximum value, where at least one of the maximum or minimum value can be changed in accordance with a user instruction.

[0024] According to this structure, users can obtain the desired 3D image in volume rendering by adjusting at least one of the maximum or minimum values.

[0025] Alternatively, the nearby voxel may be a voxel that is not adjacent to the voxel of interest in the ultrasound volume data.

[0026] Alternatively, the ultrasound data can be obtained by transmitting and receiving ultrasound waves directed at the fetus.

[0027] Furthermore, the ultrasonic image processing program disclosed in this specification enables a computer to function as follows: a change amount information calculation unit, which calculates change amount information representing the change between the signal intensity of a voxel of interest and the signal intensity of a neighboring voxel located near the voxel of interest, for each voxel constituting ultrasonic volume data obtained by transmitting and receiving ultrasonic waves to an object under examination; an opacity calculation unit, which calculates the opacity of each voxel based on the change amount information, such that the smaller the change in signal intensity between the voxel and the neighboring voxel as represented by the change amount information, the smaller the opacity of the voxel V; and a rendering processing unit, which performs volume rendering processing on the ultrasonic volume data for each line of sight to obtain pixel values ​​based on the signal intensity of each voxel and the calculated opacity of each voxel, and forms an ultrasonic image based on multiple pixel values ​​for each line of sight.

[0028] According to the ultrasonic image processing apparatus disclosed in this specification, even when the ultrasonic volume data is noisy, the influence of the noise in the three-dimensional image obtained by volume rendering the ultrasonic volume data can be reduced. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the structure of the ultrasound diagnostic device involved in this embodiment.

[0030] Figure 2 This is a schematic diagram of the structure of the three-dimensional image forming part.

[0031] Figure 3 This diagram illustrates the process of generating ultrasonic body data from multiple received frame data.

[0032] Figure 4 This is a conceptual diagram of ultrasound body data.

[0033] Figure 5 This is a conceptual diagram representing the focus voxels and the neighboring voxels.

[0034] Figure 6 It is a diagram used for illustration. Detailed Implementation

[0035] The following description of this embodiment is based on the accompanying drawings.

[0036] Figure 1 This is a schematic structural diagram of the ultrasound diagnostic apparatus 10, which is an ultrasound image processing apparatus according to this embodiment. The ultrasound diagnostic apparatus 10 is installed in medical institutions such as hospitals and is an apparatus that forms and displays ultrasound images based on received signals obtained from the transmission and reception of ultrasound waves directed at a subject as a living organism. The subject may include, for example, a pregnant woman (especially a fetus).

[0037] Specifically, in the ultrasound diagnostic apparatus 10, ultrasound volume data is generated based on the received signal obtained by transmitting and receiving ultrasound waves directed at the subject. This ultrasound volume data is then volumetrically plotted to form a three-dimensional image as an ultrasound image. Especially when the subject is a fetus, ultrasound volume data is obtained by transmitting and receiving ultrasound waves directed at the fetus, and the fetus's face, etc., are stereoscopically represented in the three-dimensional image formed from this ultrasound volume data. Such a three-dimensional image is then provided to pregnant women, etc.

[0038] The ultrasonic probe, or probe 12, is a device for transmitting ultrasonic waves and receiving reflected waves. Specifically, probe 12 is in contact with the surface of the subject, transmitting an ultrasonic beam into the subject and receiving reflected waves reflected by tissue within the subject. Probe 12 contains a vibrating element array comprising multiple vibrating elements. An electrical transmission signal is supplied to each vibrating element in the vibrating element array from the transmitting unit 14 (described later), thereby generating an ultrasonic beam (transmitting beam). Furthermore, each vibrating element in the vibrating element array receives reflected waves from the subject, converts the reflected waves into electrical signals, and transmits them to the receiving unit 16 (described later).

[0039] In this embodiment, probe 12 is a one-dimensional ultrasonic probe with a one-dimensional array of transducers. A beam scanning surface is formed by scanning the ultrasonic beam with probe 12. By scanning the ultrasonic beam with multiple beam scanning surfaces that are slightly different from each other, a received signal for generating ultrasonic volume data is obtained. Alternatively, probe 12 can also be a two-dimensional ultrasonic probe with a two-dimensional array of transducers. In a two-dimensional ultrasonic probe, a two-dimensional ultrasonic beam is formed by a two-dimensional array of transducers, thereby obtaining a received signal for generating ultrasonic volume data.

[0040] The transmitting unit 14 functions as a transmitting beamformer. During ultrasonic wave transmission, the transmitting unit 14 provides multiple transmitting signals side-by-side to the probe 12 (specifically, an array of vibrating elements). Thus, an ultrasonic beam is transmitted from the probe 12.

[0041] The receiving unit 16 functions as a receiving beamformer. During the reception of reflected waves, the receiving unit 16 receives multiple received signals from the probe 12 (specifically, a vibrating element array) in parallel. The receiving unit 16 performs phase modulation addition and other processing on the multiple received signals, thereby generating received beam data. The received beam data consists of multiple signals (data) representing the signal intensity (characterized by values) of the reflected waves, arranged along the depth direction of the object being examined. Received frame data is formed from the multiple received beam data, equivalent to one ultrasonic image, obtained from the beam scanning plane. Furthermore, if the probe 12 is a two-dimensional ultrasonic probe, ultrasonic volume data is formed by arranging the multiple received beam data in a two-dimensional manner.

[0042] The signal processing unit 18 performs various signal processing operations on the received beam data from the receiving unit 16, including detection processing, logarithmic amplification processing, gain correction processing, or filtering processing.

[0043] The image memory 20 is a memory that stores multiple received frame data processed by the signal processing unit 18. The image memory 20 is a FIFO (First In First Out) buffer that outputs data in the order in which the received frame data from the signal processing unit 18 is input. In the case where the probe 12 is a two-dimensional ultrasonic probe, the ultrasonic volume data is stored in the image memory 20.

[0044] The three-dimensional image forming unit 22 performs volume rendering processing on the ultrasonic volume data based on the received signal obtained by transmitting and receiving ultrasonic waves to the subject, thereby generating a three-dimensional image. Details of the processing performed by the three-dimensional image forming unit 22 will be described later.

[0045] The display control unit 24 displays the three-dimensional image generated by the three-dimensional image forming unit 22, for example, on a display 26 composed of a liquid crystal panel or the like.

[0046] The input interface 28 may be composed of, for example, buttons, a trackball, or a touch panel. The input interface 28 is used to input user instructions into the ultrasound diagnostic device 10.

[0047] The memory 30 may be constructed using HDD (Hard Disk Drive), SSD (Solid State Drive), eMMC (embedded Multi Media Card), ROM (Read Only Memory), or RAM (Random Access Memory), etc. The memory 30 stores ultrasound image processing programs used to operate various parts of the ultrasound diagnostic device 10. Alternatively, the ultrasound image processing programs can also be stored on computer-readable, non-transitory storage media such as USB (Universal Serial Bus) memory or CD-ROM. The ultrasound diagnostic device 10 can read and execute the ultrasound image processing programs from such storage media.

[0048] The control unit 32 is configured to include at least one of a general-purpose processor (such as a CPU (Central Processing Unit)) and a dedicated processor (such as a GPU (Graphics Processing Unit), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or programmable logic device). Alternatively, the control unit 32 may not be based on a single processing device, but may be configured by the joint operation of multiple processing devices located in physically separate positions. The control unit 32 controls the various parts of the ultrasound diagnostic apparatus 10 according to an ultrasound image processing program stored in the memory 30.

[0049] Furthermore, each of the transmitting unit 14, receiving unit 16, signal processing unit 18, three-dimensional image forming unit 22, and display control unit 24 is composed of one or more processors, chips, electrical circuits, etc. These units can be implemented through the joint operation of hardware and software.

[0050] Figure 2 This is a schematic diagram of the structure of the three-dimensional image forming unit 22. First, an outline of the processing of the three-dimensional image forming unit 22 in this embodiment will be explained. As shown in Equation 1 above in the conventional example, in volume rendering, the opacity α(X) of voxel n... n The larger the value of C, the greater the operand C. nThe signal intensity X of voxel n below n The greater the impact, the less noise there is. Therefore, if the opacity of the voxel corresponding to the noise (the voxel with the signal strength caused by the noise) is reduced, the impact of noise in the calculated value (cumulative value) of the ray becomes smaller, that is, the noise in the 3D image is reduced.

[0051] In this embodiment, the ultrasound volume data focuses on the following characteristic: for a voxel corresponding to noise, the difference between the signal intensity of that voxel and the signal intensity of a nearby voxel is small (in other words, there are few cases where the difference in signal intensity between the voxel corresponding to noise and its nearby voxel is large). This is believed to be because, due to the increased reflection of ultrasound waves at the boundaries of tissues within the subject (e.g., the surface of a fetus's face), and the reduced reflection at non-boundary areas, the difference in signal intensity between nearby voxels is larger (emphasizing edges in image terms). In contrast, for the noise component, multiple voxels with the same signal intensity are mostly dispersed, which often results in a blurred image (low contrast). Furthermore, in this embodiment, the opacity of each voxel is determined (corrected) based on the difference in signal intensity between voxels. As a result, the opacity of the voxel corresponding to noise is reduced, and the influence of noise in the three-dimensional image is decreased.

[0052] The following is for reference. Figures 2-4 This section will explain the details of the processing performed by the three-dimensional image forming unit 22, and the details of volume rendering in the ultrasound diagnostic apparatus 10 according to this embodiment.

[0053] The 3D scan converter 40 generates ultrasonic body data based on multiple frames of data stored in the image memory 20. Specifically, as... Figure 3 As shown, the 3D scan converter 40 generates ultrasonic volume data Vol by combining multiple received frame data F corresponding to multiple beam scanning surfaces with slightly different positions. Furthermore, if the probe 12 is a two-dimensional ultrasonic probe, since the ultrasonic volume data is already stored in the image memory 20, processing by the 3D scan converter 40 is not required.

[0054] Figure 4 This is a conceptual diagram of ultrasound volume data (Vol). Ultrasound volume data (Vol) is a three-dimensional arrangement of voxels (V) representing the signal intensity of reflected waves from the subject. Figure 4 (as well as Figure 5 , Figure 6 In this diagram, the X-axis represents, for example, the azimuth direction, the Y-axis represents, for example, the slice direction, and the Z-axis represents, for example, the depth direction. Figures 4-6In the diagram, only a portion of voxels V are shown. The signal strength of each voxel V is based on the signal strength at each depth of the received beam data generated by the receiving unit 16, i.e., the received beam data processed by the signal processing unit 18.

[0055] The signal intensity conversion unit 42 performs processing to convert the signal intensity of each voxel V in the ultrasound volume data Vol. Specifically, it performs processing to set the signal intensity of voxels V with signal intensities below a signal intensity threshold to 0. This simplifies the subsequent processing in the three-dimensional image forming unit 22. In addition, the signal intensity threshold can be a predetermined value, or it can be set by a user such as a physician by inputting an instruction from the input interface 28.

[0056] The smoothing processing unit 44 performs a smoothing process to smooth the ultrasonic volume data Vol. The smoothing process is as follows: by applying a smoothing filter consisting of a small number of two-dimensional or three-dimensional pixels with weights assigned to each pixel to each local region of the ultrasonic volume data Vol, the gradient of the signal intensity near the edge portion where the difference in signal intensity in adjacent voxels V is large is smoothed.

[0057] The change information calculation unit 46 calculates change information for each voxel V constituting the ultrasonic volume data Vol, representing the change between the signal intensity of that voxel V (referred to as the voxel of interest Va) and the signal intensity of the neighboring voxel Vb located near the voxel of interest Va.

[0058] Figure 5 This is a conceptual diagram representing a voxel of interest Va and its neighboring voxels Vb. In this embodiment, the voxels V adjacent to Va in the X-axis, Y-axis, and Z-axis directions, respectively, are called neighboring voxels Vb. Figure 5 (The illustration of adjacent neighboring voxels Vb in the Y-axis direction is omitted). Furthermore, not only in the X-axis, Y-axis, and Z-axis directions, but also in the oblique direction, 26 voxels V surrounding the voxel of interest Va can be designated as neighboring voxels Vb. However, in this case, the computational complexity regarding the changes in each voxel V increases (details will be described later). Therefore, in this embodiment, the 6 voxels V located in the X-axis, Y-axis, and Z-axis directions when observed from the voxel of interest Va are designated as neighboring voxels Vb.

[0059] Furthermore, in this embodiment, the voxel V adjacent to the voxel of interest Va is designated as the neighboring voxel Vb, but this is not necessarily required. The neighboring voxel Vb can also be a voxel V that is not adjacent to the voxel of interest Va. For example, the neighboring voxel Vb can also be a voxel V located at a position that is several voxels away from the voxel of interest Va.

[0060] In this embodiment, the change information representing the amount of change in signal intensity between the voxel of interest Va and the neighboring voxel Vb is the difference between the signal intensity of the voxel of interest Va and the neighboring voxel Vb, which directly characterizes the amount of change in signal intensity between the voxel of interest Va and the neighboring voxel Vb. However, the change information is not limited to the difference described above. For example, it can also be the gradient intensity of the signal intensity between the voxel of interest Va and the neighboring voxel Vb, the gradient vector of the signal intensity between the voxel of interest Va and the neighboring voxel Vb, or the first (or second) derivative value in at least one direction of the signal intensity of the voxel of interest Va along the X-axis, Y-axis, or Z-axis.

[0061] Specifically, in this embodiment, the change amount information calculation unit 46 calculates the change in signal intensity between the voxel of interest Va and the neighboring voxel Vb as follows. First, the change amount information calculation unit 46 calculates the center difference of the signal intensity between the voxel of interest Va and the neighboring voxel Vb in each direction of the X-axis, Y-axis and Z-axis using the following equations 3 to 5.

[0062]

Mathematical Expression 3

[0063]

[0064]

[0065]

[0066] In Equations 3 through 5, X(x, y, z) represents the signal intensity of voxel V at coordinates (x, y, z) in the ultrasound volume data Vol. L is an integer greater than or equal to 1. That is, Equation 3 calculates the center difference of the signal intensity along the X-axis with respect to voxel Va (voxel V at coordinates (x, y, z)) (the difference between the signal intensity of the neighboring voxel Vb at coordinates (x+L, y, z) and the signal intensity of the neighboring voxel Vb at coordinates (xL, y, z) divided by 2L). Similarly, Equation 4 calculates the center difference of the signal intensity along the Y-axis with respect to voxel Va, and Equation 5 calculates the center difference of the signal intensity along the Z-axis with respect to voxel Va. Furthermore, if L = 1, then the neighboring voxel Vb becomes a voxel V adjacent to voxel Va; if L ≥ 2, then the neighboring voxel Vb becomes a voxel V not adjacent to voxel Va.

[0067] Then, in this embodiment, the change amount information calculation unit 46 sets the sum of the squares and square roots of the center differences of the signal strength calculated for the three directions as the change amount of signal strength between the voxel of interest Va and the neighboring voxel Vb. That is, the change amount information calculation unit 46 calculates the change amount of signal strength between the voxel of interest Va and the neighboring voxel Vb using the following Equation 6.

[0068]

Mathematical Expression 4

[0069]

[0070] As in Equation 6, the change in signal intensity between voxel V and neighboring voxels is recorded as V. n The above-described calculation is used to calculate the total change in signal intensity V between voxels V and neighboring voxels that constitutes the ultrasonic volume data Vol. n .

[0071] Additionally, regarding the change in signal intensity V between voxel V and its neighboring voxels... n Alternatively, the calculation can be performed using the method described below.

[0072] For example, the change information calculation unit 46 can use one of the center differences of the signal intensity between the voxel of interest Va and the neighboring voxel Vb in each direction of the X-axis, Y-axis and Z-axis to calculate V. n For example, as shown in Equation 7 below, the change information calculation unit 46 can set the maximum value of the center difference of the signal strength in each axial direction as V. n Alternatively, as shown in Equation 8 below, the minimum value among the center differences of signal strength in each axial direction can be set as V. n .

[0073]

Mathematical Expression 5

[0074] V n (xy, z) = max (dX (x, y, z), aY (x, y, z), dZ (x, y, z))... (Formula 7)

[0075] V n (x, y, z) = min (dX (x, y, z), dY (x, y, z), dZ (x, y, z))... (Formula 8)

[0076] Furthermore, the change information calculation unit 46 can also use two of the center differences of the signal intensity between the voxel of interest Va and the nearby voxel Vb in each direction of the X-axis, Y-axis, and Z-axis to calculate V. n For example, as shown in Equation 9 below, the change information calculation unit 46 can set the square root of the sum of the squares of the two largest values ​​of the center difference of the signal strength in the three axial directions as V. n .

[0077]

Mathematical Expression 6

[0078] dX(x,y,z)>dZ(x,y,z) and dY(x,y,z)>dZ(x,y,z)

[0079]

[0080] Furthermore, the change information calculation unit 46 can also, as shown in Equation 10 below, set the average value of the center difference of the signal intensity between the voxel of interest Va and the nearby voxel Vb in each direction of the X-axis, Y-axis, and Z-axis as V. n .

[0081]

Mathematical Expression 7

[0082] V n (xy, z) = aVe (dX (x, y, z), dY (x, y, z), dZ (x, y, z))... (Formula 10)

[0083] Furthermore, the change information calculation unit 46 can also perform the calculation simultaneously with smoothing using a Sobel filter whenever it calculates the center difference of the signal intensity between the voxel of interest Va and the neighboring voxel Vb in each direction of the X, Y, and Z axes. Alternatively, the difference of the signal intensity between the voxel of interest Va and the neighboring voxel Vb in each axis direction can be set not as a center difference, but as the difference of the signal intensity between the voxel of interest Va and one neighboring voxel Vb in each axis direction.

[0084] Back Figure 2 The opacity calculation unit 48 calculates the change information (in this embodiment, the change amount V) about each voxel V based on the change information calculation unit 46. n The opacity calculation unit 48 calculates the opacity of each voxel V. Specifically, the opacity calculation unit 48 performs calculations such that the smaller the change in signal intensity between the voxel V and its neighboring voxels, as represented by the change in voxel V, the smaller the opacity of that voxel V.

[0085] The opacity calculation unit 48 calculates the opacity α(X) for each voxel V, which has been pre-determined for each voxel V in the ultrasound volume data Vol, based on the information about the change in each voxel V. n The correction parameters are used for correction. As will be discussed later, due to the opacity α(X) of each voxel... n The opacity α(X) of each voxel is corrected using the correction parameters. n Therefore, the opacity operation correction parameter of opacity operation unit 48 and the opacity of each voxel operated by opacity operation unit 48 are synonymous.

[0086] In this embodiment, the change information is obtained by varying the signal intensity V between each voxel V and its neighboring voxels. n Characterized, therefore, the opacity calculation unit 48 is based on the change V nThe correction parameters are calculated. Specifically, the correction parameters g(V) for each voxel V are calculated using the following Equation 11. n ).

[0087]

Mathematical Expression 8

[0088]

[0089] As shown in Equation 11, the opacity calculation unit 48 calculates the change in v with respect to voxel V. n For a given maximum value V max In the above cases, the correction parameter g(V) will be applied. n Let ) be 1, and let V be the change in voxel V. n Insufficient to the maximum value V max And it is the given minimum value V min In the above cases, the correction parameter g(V) will be applied. n Let V be the value of V. n / V max Regarding the change in voxel V, V n Insufficient to the predetermined minimum value V min In this case, the correction parameter g(V) will be applied. n The opacity calculation unit 48 sets the value to 0. That is, the opacity calculation unit 48 calculates the change in V by... n Normalized to the minimum value V min to the maximum value V max The values ​​between are used to calculate the correction parameter g(V). n ).

[0090] The correction parameter g(V) shown in Equation 11 above n ) characterizes linear correction, but the correction parameter g(V) n ) can also characterize nonlinear corrections. In this case, the opacity calculation unit 48 calculates the correction parameter g(V) for each voxel V, for example, using the following equation 12. n ).

[0091]

Mathematical Expression 9

[0092]

[0093] In Equation 12, the transformation function f(V) n ) is based on V min ~V max The normalized formula is represented, for example, by the following formula 13.

[0094]

Mathematical Formula 10

[0095]

[0096] In Equation 13, s(V n) is the sigmoid function, characterized by the following equation 14.

[0097]

Mathematical Expression 11

[0098]

[0099] In Equation 14, gain is a parameter that determines the steepness of the sigmoid curve. Gain can be adjusted by the user or be a predetermined fixed value. Furthermore, n(V n ) is in V n For V max When it becomes 1, in V n For V min When it becomes -1, the normalization expression is represented by the following equation 15.

[0100]

Mathematical Expression 12

[0101]

[0102] Then, for each voxel V, the opacity calculation unit 48 calculates the predetermined opacity α(X) for that voxel V. n ) and the correction parameter g(V) calculated for this voxel V n The product of the products of voxels is taken as the corrected opacity of the voxel V. That is, the corrected opacity of each voxel V is characterized by the following equation 16.

[0103]

Mathematical Expression 13

[0104] a(X n )*(V n ...(Equation 16)

[0105] In the correction parameter g(V) n When α(X) is 1, the opacity is not corrected. n In the correction parameter g(V) n When the value of ) is 0, the opacity of voxel V becomes 0. That is, in the volume rendering in the rendering processing unit 50 described later, the signal intensity of voxel V is completely ignored. In the correction parameter g(V) n ) is V n / V max In the case of V min ≤V n ≤V max Therefore, V n / V max It becomes a value smaller than 1. Therefore, the opacity of this voxel becomes smaller than the predetermined value, and the influence of the signal intensity of this voxel V in volume rendering is reduced compared to the case where the opacity is not corrected.

[0106] In addition, in this embodiment, the correction parameter g(V) n The maximum value that can be taken is 1, but the correction parameter g(V) n It can also take values ​​greater than 1. When the correction parameter g(Vn) is greater than 1, the opacity of the voxel V is increased, and the influence of the signal intensity of the voxel V is emphasized in volume rendering.

[0107] The maximum value V contained in Equation 11 or Equation 12 max Or minimum value V min At least one of them can be changed in response to user instructions. If the maximum value V max If it becomes smaller, it will have an effect in the direction that increases the voxel V with more uncorrected opacity. If the maximum value V max If it increases, it will have an effect in the direction of increasing the voxel V that corrects opacity. On the other hand, if the minimum value V... min If it becomes smaller, it will affect the direction in which the voxel V with opacity of 0 (completely ignored in volume rendering) decreases. If the minimum value V... min Increasing the value of V increases the number of voxels with an opacity of 0. Users can adjust the maximum value of V to achieve the desired 3D image in volume rendering. max Or minimum value V min At least one of them.

[0108] The drawing processing unit 50 sets the line of sight (ray R) through the ultrasonic body data Vol from a certain viewpoint (reference). Figure 6 Then, based on the voxels V( located on ray R) Figure 6 The signal intensity X of voxels (V1, V2, ..., VN) is given by the given signal intensity X. n And the corrected opacity α(X) n )*g(V n The drawing processing unit 50 performs cumulative processing on each ray to obtain the calculated value. Specifically, the drawing processing unit 50 performs cumulative processing on each ray, characterized by the following equation 17, where the following equation 17 represents the opacity α(X) of each voxel in the above equation 1 involved in the existing example. n Replaced with the corrected opacity α(X) n )*g(V n (and thus obtained.)

[0109]

Mathematical Expression 14

[0110] C n =C n-1 +(1-α n-1 )*{α(X n )*g(V n )}*X n …(Equation 17)

[0111] The cumulative value α of the opacity up to voxel n located on ray R. n Characterized by the following Equation 18, wherein the following Equation 18 represents the opacity α(X) of each voxel in Equation 2 above involved in the prior art. n Replaced with the corrected opacity α(X) n )*g(V n (and thus obtained.)

[0112]

Mathematical Expression 15

[0113] α n =α n-1 +(1-α n-1 )*{α(X n )*g(V n ...(Equation 18)

[0114] As shown in Equations 17 and 18, the rendering processing unit 50 calculates the signal intensity X of each voxel V based on the signal intensity X. n and the opacity α(X) of each voxel V calculated (corrected) by the opacity calculation unit 48. n )*g(V n The ultrasonic volume data Vol is drawn by performing volume rendering for each line of sight to obtain pixel values, and a three-dimensional image as an ultrasonic image is formed based on multiple pixel values ​​for each line of sight.

[0115] As described above, the voxel V corresponding to the noise (the voxel V with the signal strength caused by the noise) has a change in signal strength V between itself and its neighboring voxels. n This characteristic is that it becomes smaller. According to this embodiment, the opacity calculation unit 48 calculates the change amount V as shown in Equation 11 or Equations 12-15 above. n The smaller the value of the correction parameter g(V), the smaller its value. n That is, the change V n The smaller the value, the higher the corrected opacity α(X) of the voxel V. n )*g(V n The smaller the value, the less impact the signal intensity (noise component) of the voxel V corresponding to noise has on the 3D image during volume rendering.

[0116] On the other hand, in the boundary regions of tissues that are desired to be characterized in a three-dimensional image (e.g., the surface of a fetal face), the amount of signal intensity change V between the tissue and the adjacent voxels... n As it increases, therefore, the correction parameter g(V) nThe opacity of the voxel V corresponding to the boundary of the tissue is not significantly corrected (not reduced), thus reducing the likelihood of the tissue boundary becoming blurred or less bright in the 3D image.

[0117] The ultrasonic image processing apparatus involved in this disclosure has been described above. However, the ultrasonic image processing apparatus involved in this disclosure is not limited to the above-described embodiments. Various modifications can be made as long as they do not depart from the main idea.

[0118] For example, in this embodiment, the ultrasound image processing apparatus is the ultrasound diagnostic apparatus 10, but the ultrasound image processing apparatus is not limited to the ultrasound diagnostic apparatus 10 and may also be other computers. In this case, the computer serving as the ultrasound image processing apparatus functions as the three-dimensional image forming unit 22. Specifically, the computer serving as the ultrasound image processing apparatus receives multiple receive frame data or ultrasound volume data from the ultrasound diagnostic apparatus, and performs calculations on the multiple receive frame data or ultrasound volume data regarding the change amount information of each voxel V, calculations regarding the opacity of each voxel, and volume rendering.

Claims

1. An ultrasonic image processing device, characterized in that, have: The change amount information calculation unit calculates change amount information for each voxel constituting the ultrasonic volume data obtained by transmitting and receiving ultrasonic waves to the subject, which represents the change amount between the signal intensity of the voxel of interest and the signal intensity of a neighboring voxel located near the voxel of interest. The opacity calculation unit calculates the opacity of each voxel based on the change information of each voxel, such that the smaller the change in signal intensity between the voxel and the neighboring voxel as represented by the change information of the voxel, the smaller the opacity of the voxel. and The rendering processing unit, based on the signal intensity of each voxel and the calculated opacity of each voxel, performs volume rendering on each line of sight to obtain pixel values ​​for the ultrasonic volume data, and forms an ultrasonic image based on multiple pixel values ​​for each line of sight. The opacity calculation unit performs calculations with respect to each voxel based on the correction parameters for the change amount, and corrects the opacity of each voxel based on the pre-determined opacity for each voxel and the correction parameters. The opacity calculation unit calculates the correction parameter by normalizing the change to a value between a given minimum and a given maximum value. The opacity calculation unit sets the correction parameter to 1 or a value greater than 1 when the change in the amount of change with respect to the voxel is greater than or equal to the maximum value; sets the correction parameter to the ratio of the change in the amount of change with respect to the voxel to the maximum value when the change in the amount of change with respect to the voxel is less than or equal to the minimum value; and sets the correction parameter to 0 when the change in the amount of change with respect to the voxel is less than or equal to the minimum value. At least one of the maximum or minimum values ​​can be changed in accordance with user instructions.

2. The ultrasonic image processing device according to claim 1, characterized in that, The neighboring voxel is a voxel in the ultrasound volume data that is not adjacent to the voxel of interest.

3. The ultrasonic image processing apparatus according to claim 1 or 2, characterized in that, The ultrasound data is obtained by transmitting and receiving ultrasound waves directed at the fetus.

4. A computer-readable, non-transitory storage medium, which stores commands executable by a computer, characterized in that, The command causes the computer to perform the following steps: The change information calculation step calculates change information for each voxel that constitutes the ultrasonic volume data obtained by transmitting and receiving ultrasonic waves for the subject, representing the change between the signal intensity of the voxel of interest and the signal intensity of a neighboring voxel located near the voxel of interest. The opacity calculation step calculates the opacity of each voxel based on the change information of each voxel, such that the smaller the change in signal intensity between the voxel and the neighboring voxel as represented by the change information of the voxel, the smaller the opacity of the voxel. and The rendering process involves performing volume rendering on the ultrasonic volume data for each viewpoint, based on the signal intensity of each voxel and the calculated opacity of each voxel, to obtain pixel values. An ultrasonic image is then formed based on these multiple pixel values ​​for each viewpoint. In the opacity calculation step, the calculation for each voxel is based on the correction parameter of the change amount, and the opacity of each voxel is corrected based on the pre-determined opacity of each voxel and the correction parameter. In the opacity calculation step, the correction parameter is calculated by normalizing the change to a value between a given minimum and a given maximum value. In the opacity calculation step, if the change in the amount of voxel is greater than or equal to the maximum value, the correction parameter is set to 1 or a value greater than 1; if the change in the amount of voxel is less than the maximum value but greater than or equal to the minimum value, the correction parameter is set to the ratio of the change in the amount of voxel to the maximum value; and if the change in the amount of voxel is less than the minimum value, the correction parameter is set to 0. At least one of the maximum or minimum values ​​can be changed in accordance with user instructions.

Images