JP4616711B2 - Ultrasonic diagnostic equipment - Google Patents

Description

  The present invention relates to an ultrasonic diagnostic apparatus, and more particularly to a technique for transferring data to a processor that forms an arbitrary tomographic image.

  In the ultrasonic diagnostic apparatus, a three-dimensional space (three-dimensional echo data capturing space) is formed by two-dimensional scanning with an ultrasonic beam. As a result, volume data is acquired. A three-dimensional image is formed by rendering processing on the volume data. On the other hand, an arbitrary cross section (cut plane) is arbitrarily set with respect to a three-dimensional space at an arbitrary position and orientation, and a tomographic image (arbitrary tomographic image) corresponding to the arbitrary cross section is formed.

  Generally, volume data is composed of a plurality of frame data, and each frame data is composed of a plurality of beam data. Each beam data is constituted by an echo data string existing on the ultrasonic beam. That is, the volume data is configured as an echo data group (set) that exists discretely in the three-dimensional space.

  When an arbitrary slice is specified in the three-dimensional space and the above-mentioned arbitrary tomographic image is formed, the echo data is not necessarily stored in the three-dimensional coordinates corresponding to each pixel (lattice point, pixel point) on the arbitrary section in the three-dimensional space. Does not exist. Therefore, for each pixel on an arbitrary cross section, a plurality of (for example, 8 or 16) echo data existing in the vicinity of the pixel are referred to, and interpolation calculation is performed based on them to obtain a pixel value of the pixel of interest. (Interpolation value) is calculated.

  In the conventional apparatus, when an arbitrary cross section is designated for a three-dimensional space, first, a plurality of neighboring echo data used for the interpolation calculation for each pixel according to the pixel order (raster scan order) on the arbitrary cross section. Three-dimensional coordinates (address) are calculated, and a weighting factor for each neighboring echo data is calculated. That is, interpolation data for interpolation calculation is calculated for each pixel. An interpolation data string composed of a plurality of interpolation data constitutes an interpolation table. Next, the pixels are specified in order according to the pixel order on the arbitrary cross section, the interpolation data for each pixel is acquired from the interpolation table, the plurality of echo data specified by the interpolation data is acquired, and they are used. Pixel values are calculated by interpolation calculation.

  More specifically, in the conventional apparatus, the acquired echo data group and interpolation table are stored in a storage unit (external storage unit). On the other hand, the image forming process is executed at high speed by a device such as a DSP (digital signal processor). In the device, interpolation data is acquired from the external storage unit according to the pixel order on an arbitrary cross section, and separately, a plurality of neighboring echo data specified by the interpolation data (usually they have coordinate continuity). No) was acquired from the external storage unit, and the above interpolation processing was executed. The following Patent Document 1 and Patent Document 2 describe an ultrasonic diagnostic apparatus using an interpolation table. However, it does not describe speeding up of data transfer or rearrangement of table elements. Patent Document 3 below describes an interpolation calculation in units of sections. However, there is no description about efficiently performing the interpolation operation in relation to the cache function described later.

  Note that the external storage unit is generally composed of DRAM, while the internal memory in the image forming processor is generally composed of SRAM. As is well known, the latter SRAM can easily perform high-speed random access, but its capacity cannot be increased due to various constraints. On the other hand, although it is easy to increase the capacity of the DRAM, it is necessary to specify a row address and a column address, so that random access takes time. Originally, if all of the volume data and the interpolation data string can be stored in the internal memory, the random access can be performed to perform rapid interpolation processing. However, in practice, necessary data can be transferred from the external memory to the internal memory. It is necessary to read the data into small pieces, and sequentially proceed with the transfer and interpolation processing. An unpublished application related to the present application is Japanese Patent Application No. 2004-254757 filed on September 1, 2004.

JP 2000-239 A JP 2001-327506 A JP 2000-339441 A

  Usually, a processor such as a DSP has a cache function. That is, when acquiring specific data from the external storage unit, a block (cache block) including the data is acquired and temporarily stored in the internal memory (cache memory). That is, the transfer is performed in units of blocks. Even after the data is used in the interpolation operation, the block is stored as it is until it is evicted from the cache memory (until invalidated) when another block is stored. As a result, even after using the echo data in the block, any echo data in the block can be immediately acquired from the cache memory (if it exists) as needed. It is.

  However, in the prior art, efficient data transfer is not realized on the premise of such a cache function. Even in the prior art, data can be acquired in units of blocks based on the cache function. However, if the interpolation operation is simply advanced in pixel order (raster scan order) on an arbitrary section, the data in the cache memory is There was a problem that it could not be used effectively (increased cache misses). As a result, the time required to form one arbitrary tomographic image has increased.

  An object of the present invention is to enable rapid image processing by utilizing a cache function when an arbitrary tomographic image is formed.

  An object of the present invention is to reduce the number of random accesses to an external memory from a processor that forms an arbitrary tomographic image, that is, to improve the locality of reference in caching and improve the formation rate of the arbitrary tomographic image.

(1) The present invention provides a first storage unit that stores a group of echo data obtained from a three-dimensional space in a living body, and each pixel on a cutting plane set for the three-dimensional space. An interpolation data calculation unit for calculating interpolation data having a plurality of weighting factors to be given to a plurality of neighboring echo data used in the interpolation calculation; a second storage unit for storing interpolation data of each pixel on the cut surface; Transfer of echo data from the first storage unit to the image forming processor is executed in units of cache blocks, and the second storage unit is connected to the second storage unit. The transfer of the interpolation data from the storage unit to the image forming processor is executed in the arrangement order of the cache blocks to which representative neighborhood echo data of the plurality of neighborhood echo data belongs, and the image formation The processor has a cache storage unit that temporarily stores the echo data transferred from the first storage unit, and the necessary echo data according to the interpolation data transferred from the second storage unit. And a pixel value calculation unit that obtains a pixel value by performing an interpolation calculation from the storage unit.

  According to the above configuration, the interpolation data of each pixel on the arbitrary slice is transferred from the second storage unit to the image forming processor in the arrangement order of the cache blocks to which the representative neighborhood echo data belongs. That is, the interpolation operation of each pixel proceeds in the cache block arrangement order. For example, in the interpolation calculation for each pixel, a plurality of spatially separated neighboring echo data are referred to, and therefore, a plurality of data blocks having these are fetched into the image forming processor. When the representative neighborhood echo data for one pixel and the representative neighborhood echo data for another pixel exist in the same cache block, it is possible to share multiple other cache blocks in the interpolation operation for both pixels High nature. As described above, when the representative neighborhood echo data are in the same cache block, the interpolation locality is continuously performed, so that the locality of the reference is improved and the number of times the image forming processor randomly accesses the external memory is determined. Can be reduced.

  In the above configuration, the first storage unit and the second storage unit may be configured as a physically integrated storage device. A plurality of cache storage units (for example, a primary cache and a secondary cache) may exist in stages in the image forming processor. The cache block is preferably a one-dimensional data string in terms of space, but may be a two-dimensional or three-dimensional data string. It is desirable to set the sequence so that a series of cache blocks are as continuous as possible in space.

  Preferably, the pixel value calculation unit obtains the necessary echo data from the cache storage unit when the echo data necessary for the interpolation calculation exists in the cache storage unit, and the echo data necessary for the interpolation calculation. Is not present in the cache storage unit, the necessary echo data is acquired from the first storage unit. Preferably, the cache blocks are composed of echo data strings arranged in time series, and the order of the cache blocks is time series. Preferably, the second storage unit stores an interpolation data string in the arrangement order of the cache blocks. Rearrangement may be performed at the stage of storing a plurality of interpolation data on the second storage unit, or rearrangement may be performed when an interpolation data string is read from the second storage unit. Further, after the interpolation data is generated, the interpolation data itself may be rearranged, or information for calculating the interpolation data may be rearranged before the interpolation data is calculated. One interpolation data may be used as a transfer unit, or a plurality of continuous interpolation data may be used as a transfer unit.

  Desirably, the said interpolation data calculating part determines the effective pixel which belongs in the effective display area on the said arbitrary cut surface, and calculates interpolation data about each effective pixel. According to this configuration, an unnecessary interpolation calculation can be avoided and the image formation rate can be improved.

(2) The present invention includes a transmission / reception unit for transmitting / receiving ultrasonic waves, a first storage unit for storing echo data groups obtained from a three-dimensional space in the living body by the transmission / reception of the ultrasonic waves, and the tertiary Interpolation data having a plurality of weighting factors given to a plurality of neighborhood echo data used in the interpolation calculation for each pixel on the cut surface and a cut surface setting unit for setting a cut surface by the user with respect to the original space An interpolation data calculation unit that calculates the value, an arrangement setting unit that sets an arrangement of an interpolation data string composed of interpolation data for each pixel, and a second storage unit that stores an interpolation data string having the set order An image forming processor connected to the first storage unit and the second storage unit, and transfer of echo data from the first storage unit to the image forming processor is performed in units of cache blocks. The order of the interpolation data sequence stored in the second storage unit corresponds to the order of the cache blocks to which the representative neighborhood echo data in the plurality of neighborhood echo data belongs, and each of the interpolation data sequences constituting the interpolation data sequence Interpolation data is sequentially transferred from the head to the image forming processor, and the image forming processor temporarily stores echo data transferred from the first storage unit in units of the cache blocks, and In accordance with the interpolation data of each pixel sequentially transferred from the two storage units, if the echo data necessary for the interpolation operation of the pixel exists in the cache storage unit, the necessary echo data is acquired from the cache storage unit, and the necessary If the echo data does not exist in the cache storage unit, obtain the necessary echo data from the first storage unit, A pixel value calculation unit that performs an interpolation calculation of the pixel based on the obtained plurality of echo data, thereby obtaining a pixel value of the pixel, and an image storage unit that stores the pixel value calculated for each pixel It is characterized by having.

  As described above, according to the present invention, when an arbitrary tomographic image is formed, a quick image processing can be realized by utilizing the cache function more. Alternatively, according to the present invention, it is possible to reduce the number of random accesses to an external memory from a processor that forms an arbitrary tomographic image, that is, increase the locality of reference in caching and improve the formation rate of the arbitrary tomographic image.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described with reference to the drawings.

  FIG. 1 shows a preferred embodiment of a medical ultrasonic diagnostic apparatus according to the present invention, and FIG. 1 is a block diagram showing the overall configuration thereof.

  The 3D probe 10 is a transducer that acquires volume data by transmitting and receiving ultrasonic waves to and from a three-dimensional space in a living body. In the present embodiment, a 2D array transducer formed by two-dimensionally arranging a plurality of vibration elements is provided in the 3D probe 10. An ultrasonic beam is formed by the 2D array transducer, and the ultrasonic beam is electronically scanned two-dimensionally. Specifically, a scanning plane is formed by electronic scanning of an ultrasonic beam in a first direction (θ direction), and tertiary scanning is performed by electronic scanning of the scanning plane in a second direction (φ direction) orthogonal thereto. An original echo data capturing space (three-dimensional space) is formed. In the present embodiment, the 3D space is formed using the 2D array transducer, but the 3D space can also be formed by mechanically scanning the 1D array transducer. Incidentally, electronic sector scanning, electronic linear scanning, electronic convex scanning, and the like are known as electronic scanning methods.

  The transmission / reception unit 12 is configured as a digital beamformer, that is, the transmission / reception unit 12 functions as a transmission beamformer and a reception beamformer. Specifically, a plurality of transmission signals are supplied from the transmission / reception unit 12 to the plurality of vibration elements, thereby forming a transmission beam. The phasing addition processing is executed in the transmission / reception unit 12 for the plurality of reception signals output from the plurality of vibration elements, and thereby the reception signal after phasing addition corresponding to the reception beam is obtained. This received signal corresponds to an echo data string (beam data) obtained on the received beam.

  The memory 14 is constituted by a DRAM, for example, and the memory 14 functions as a so-called 3D memory in the present embodiment. Volume data corresponding to the three-dimensional space is stored on the memory 14. The volume data is composed of a plurality of frames (frame data), and each frame is composed of a plurality of beam data.

  The 3D image forming unit 16 is a module that constructs a three-dimensional image based on volume data. For example, an ultrasonic three-dimensional image is constructed using an image forming method such as a volume rendering method. The image data is output to the display processing unit 18.

  The arbitrary tomographic image forming unit (image forming processor) 22 is configured by a DSP (digital signal processor) having a cache function in the present embodiment. The arbitrary tomographic image forming unit 22 is a means for forming a tomographic image (corresponding to a B-mode tomographic image) corresponding to an arbitrary cross section (cut plane) set at an arbitrary position and posture with respect to the three-dimensional space by the user. is there. Inside, an internal memory 23 composed of SRAM or the like is provided. On the internal memory 23, data transferred from the memory 14 and data transferred from the interpolation table 24 are stored, and image data formed by interpolation calculation is stored. The image data is read and sent to the display processing unit 18. The display processing unit 18 performs necessary display processing on the input image data and then outputs the image data to the display device 20. A three-dimensional image and an arbitrary tomographic image are displayed on the display 20.

  The interpolation table 24 is constructed on a memory such as a DRAM. The interpolation table 24 is a table that stores a plurality of interpolation data necessary for constructing an arbitrary tomographic image as an interpolation data string. Necessary interpolation data is sequentially sent from the interpolation table 24 to the arbitrary tomographic image forming unit 22, and an arbitrary tomographic image is formed by executing interpolation processing on a pixel basis using such interpolation data. In this embodiment, the arrangement order of interpolation data strings (that is, the transfer order) is devised so as to enhance the caching effect in the DSP.

  The interpolation table creation unit 25 is a module that creates the interpolation table 24 based on coordinate data of an arbitrary cross section set by the user. The interpolation table creation unit 25 includes a rearrangement unit 26 in the present embodiment, and the rearrangement unit 26 executes rearrangement processing so that a plurality of interpolation data are in a predetermined order. That is, on the interpolation table 24, a plurality of interpolation data is stored in an array according to such a predetermined order. This will be described in detail later.

  The control unit 28 is constituted by a CPU, an operation program, and the like, and performs operation control of each configuration shown in FIG. In particular, as shown in FIG. 1, data necessary for creating an interpolation table is passed from the control unit 28 to the interpolation table creating unit 25. The control unit 28 may function as the interpolation table creation unit 25. An input unit 30 constituted by an operation panel or the like is connected to the control unit 28. Using this input unit 30, the user can specify an arbitrary cross section for the three-dimensional space and set interpolation conditions.

  FIG. 2 shows the relationship between the three-dimensional space (3D space) 66 and the arbitrary cross section (cut plane) S. In the example shown in FIG. 2, electronic sector scanning is applied in both the first direction and the second direction. The θ direction is the beam scanning direction, and the φ direction is the frame scanning direction. O represents the apex of the 3D space 66. Reference numeral 68 represents one scanning plane. A polar coordinate system is defined by the beam depth direction r, the beam scanning direction θ, and the frame scanning direction φ, while orthogonal coordinates are defined by the x, y, and z directions. An area included in the 3D space 66 on the arbitrary cross section S is an effective display image area S ′. In this embodiment, as will be described in detail later, interpolation data is calculated for each pixel (pixel point) in the display image area S ′, in other words, in an area other than the display image area S ′. For existing pixels, calculation of interpolation data is omitted. A large number of pixel points (pixel array) are defined on the arbitrary cross section S, and each pixel point is a grid point or pixel specified by the X coordinate and the Y coordinate. In FIG. 2, one pixel point P is shown.

  Each pixel point is specified by the coordinates X, X, Y on the arbitrary section S as described above, is specified by the coordinates x, y, z on the orthogonal coordinates, and further, the coordinates r, θ, Identified by φ. The pixel value of each pixel point is obtained by a three-dimensional interpolation operation based on a plurality of neighboring data (neighboring echo data set) existing in the 3D space 66. That is, for example, as shown in FIG. 3, the pixel value of the pixel point P on the arbitrary cross section S refers to the eight neighboring data A to H existing around the pixel point P, and based on them, the weighting coefficient related to the distance is obtained. It is obtained by performing the interpolation calculation used. As described later, the interpolation data for each pixel point includes address information (intra-volume offset information) indicating a “reference point” serving as a reference or a representative for specifying a plurality of neighboring data, and a plurality of neighboring data. A plurality of weighting factors and an address (number) of the pixel point on an arbitrary cross section are included. If a reference point is specified in the three-dimensional space, it is possible to specify point B if the address is advanced by one in the r direction, and point D can be specified if the address is advanced by one in the θ direction. If the address is advanced by one in the direction, the point E can be specified. Similarly, the point C, the point F, the point G, and the point H can be specified based on the point A.

  FIG. 4 conceptually shows a data transfer system for arbitrary image formation. As described above, volume data is stored in the memory 14. For this, an arbitrary cross section S is set. In the interpolation table 24, interpolation data strings for a plurality of pixels in an effective image display area on the arbitrary slice S are stored in a predetermined order. The memory 14 and the interpolation table 24 are connected to an arbitrary tomographic image forming unit (image forming processor) 22. The arbitrary tomographic image forming unit 22 includes an internal memory 23, a core module 77, a transfer controller 80, and the like. In the illustrated example, the internal memory 23 includes an input storage unit 74 and an output storage unit 78 that constitute a main memory, and a cache storage unit 76. Each storage unit may be independent or may be physically integrated. Various forms of providing a cache are known. The image forming processor constituting the arbitrary tomographic image forming unit 22 is a general-purpose processor that performs a program operation, and its hardware configuration is known.

  The volume data is composed of a large number of beam data (echo data strings). The plurality of beam data is stored on the memory 14 in the order of acquisition, that is, in time series order. In the present embodiment, a plurality of blocks (cache blocks) 70 are set for volume data in order to make the cache function work. Each block 70 constitutes a data transfer unit from the memory 14 to the arbitrary tomographic image forming unit 22. At the same time, the storage unit of the cache storage unit 76 is configured. That is, when any data is to be acquired by the cache function of the arbitrary tomographic image forming unit 22, the entire block including the data is acquired all at once. This means that random access is performed in block units. For example, a plurality of blocks 70 are set for one beam data as illustrated. Actually, not all the blocks are transferred, but only the blocks necessary for the interpolation operation for pixels on an arbitrary cross section are transferred. This is indicated by reference numeral 72 in FIG.

  Therefore, if the block transfer is simply performed in the pixel arrangement order (in the raster scan order), there is a possibility that the echo data already transferred to the arbitrary tomographic image forming unit 22 and temporarily stored can be used afterwards. It cannot be increased. This problem is particularly noticeable when the raster is turned back. In other words, even if a pixel close to the pixel that has been previously calculated becomes the current calculation target, if the previously cached data is expelled and lost, the data is acquired again in the memory 14. The chances of having to do increase.

  Therefore, in the present embodiment, for a plurality of pixels to which a reference point belongs in the same block, interpolation calculation is executed for those pixels that are close in time or successively. Specifically, the transfer order (arrangement order) from the interpolation table 24 to the arbitrary tomographic image forming unit 22 is set to be the arrangement order of the blocks to which the reference point belongs. In this embodiment, the arrangement order of blocks corresponds to the time series order. Each block to be transferred is normally in a jumping relationship in time.

  FIG. 5 shows the interpolated data string after being rearranged by the action of the rearrangement unit 26 shown in FIG. Each interpolation data 82 includes weight coefficients WA to WH 84 given to the points A to H where actual data exists, and an intra-volume offset (in-volume address) 86 as coordinate information for specifying a representative reference point (A point). And a pixel point number 88 on an arbitrary cross section. The rearrangement may be performed after the generation of the interpolation data, but in this embodiment, the rearrangement is performed on the basic information before the generation of the interpolation data as will be described later.

  Returning to FIG. 4, when the interpolation data is transferred from the interpolation table 24 according to a predetermined order, if a plurality of echo data specified by the interpolation data is stored in the cache storage unit 76, the echo data is transferred from there. It is read and sent to the core module 77. That is, random access to the memory 14 is avoided. On the other hand, if the echo data does not exist on the cache storage unit 76, random access to the memory 14 is executed, and a block including the data is acquired. The block is stored on the cache storage unit 76 until it is evicted. In this embodiment, the cache storage unit 76 and the input storage unit 74 are provided in stages as viewed from the core module 77, and data input from the outside passes through the input storage unit 74 to the cache storage unit. It is sent to the core module 77 via it. However, various modes can be adopted as long as the cache function can be exhibited.

  The core module 77 is a calculation unit that obtains a pixel value by interpolation calculation based on a plurality of neighboring data. The calculated pixel value is mapped on the output storage unit 78. When the interpolation calculation is performed for all the pixel points, one tomographic image is formed on the output storage unit 78. It is output to the outside. The transfer controller 80 functions as a cache controller, and controls data transfer, particularly block transfer and storage.

  In the present embodiment, when the interpolation data of the first pixel point is processed, for example, four blocks including eight points of data existing around the pixel point are transferred from the memory 14 to the arbitrary tomographic image forming unit 22. The echo data necessary for the interpolation operation is referred to, and those blocks are stored in the cache storage unit 76. When processing the interpolation data of the next pixel point, as described above, if necessary echo data exists on the cache storage unit 76, it is immediately acquired. In many cases, it is possible to reuse the block cached immediately before, thereby improving the locality of the reference.

  FIG. 6 shows a block array existing in a three-dimensional space. In order to perform an interpolation operation on the pixel point P1 on the arbitrary cross section S, eight neighboring data from the points a1 to h1 are referred to. In this example, they exist in four blocks # 1 to # 4. Even when the interpolation calculation is performed on the pixel point P2 close to the pixel point P1 on the arbitrary cross section S, in this example, the eight neighboring data from the points a1 to h1 are referred to as described above. In the past, interpolation operations were often performed with respect to pixel points having such a relationship, but in this embodiment, a plurality of pixel points having a reference point in the same block are consecutively performed. Since the interpolation calculation is performed, the cached four blocks # 1 to # 4 are reused, and the interpolation calculation can be performed quickly without accessing the external memory.

  FIG. 7 shows a method for creating an interpolation data string as a flowchart. In S101, 0 is substituted for i for counting the number of pixel points (pixels) belonging to the display image area shown in FIG. In S102, 0 is substituted for k for specifying the pixel point to be processed.

  In S103, the coordinates X and Y of the kth pixel point on the arbitrary cross section are converted into orthogonal coordinates x, y, and z. In S104, an operation for converting the orthogonal coordinates x, y, z into polar coordinates is executed. Specifically, the following expressions (1), (2), and (3) are executed. Note that r is a depth address along the beam direction, that is, the line direction, line indicates a line number, and frame indicates a frame number.

r = (x ² + y ² + Z ² ) ^1/2 * C1 (1)

      line = θ * C2 (2)

      frame = φ * C3 (3)

  In the above, C1 is the number of ultrasonic samples existing within the unit pixel length, that is, the number of echo data. C2 is the number of ultrasonic lines existing within a unit angle in the θ direction. C3 is the number of ultrasonic frames existing within a unit angle in the φ direction. In addition, θ = arctan (x / y) and φ = arctan (z / y), and when performing the root calculation of the above equation (1), a function may be actually executed, or a lookup table (LUT) ) Or the like.

  In S105, it is determined whether or not the pixel point of interest is within the volume. If the pixel point is outside the volume, the process proceeds to S108. On the other hand, if the pixel point of interest is within the volume, S106 is executed. In S106, in order to specify the reference point (UsOffset) described above for each pixel point, the following equation (4) is executed.

UsOffset = frame * SamplesPerLine * LinesPerFrame
+ line * SamplesPerLine + r (4)

  In the above, SamplesPerLine is the number of echo data per line (one beam), and LinesPerFrame is the number of lines per frame.

  In S107, i is incremented by one, and in S108, k is incremented by one. In S109, it is determined whether or not k has reached n * n (where n * n represents the total number of pixels on an arbitrary cross section). That is, it is determined whether or not processing has been performed up to the last pixel point. If the process has not been performed up to the last pixel point, each process after S103 is repeatedly executed, and if the process for the last pixel point is completed, S110 is executed.

  In S110, information about each pixel point calculated as described above (including the pixel point number and the address information of the reference point) is sorted in time series. That is, in the present embodiment, the blocks are arranged in time series, and information about each pixel point is rearranged in time series. Incidentally, on each beam (each line), a data string is formed along the depth direction, and each beam is sequentially formed in the θ direction. A plurality of beams are formed in the θ direction at each position in the φ direction. Pixel point information is sorted according to such an order.

  In S111, interpolation data is calculated for each pixel point according to the information described above. That is, in this embodiment, the rearrangement is executed before actually calculating the interpolation data. As a result, the interpolation data string shown in FIG. 5 is constructed. As described above, each interpolation data includes the interpolation coefficient, the offset information in the volume of the reference point, and the pixel point number information for a plurality of neighboring data. Is included. The rearrangement may be performed after the interpolation data is actually calculated for each pixel point. The interpolation data string formed in this way is stored on the interpolation table 24, and when forming an arbitrary tomographic image, the interpolation data is sequentially read out in a predetermined unit from the head in the interpolation table.

  FIG. 8 shows a flowchart of arbitrary tomographic image formation processing. First, in S201, the contents of the output storage unit shown in FIG. 4 are initialized to 0 throughout. In S202, 0 is substituted for j representing the interpolation data number.

  In S203, the jth interpolation data to the j + dth interpolation data on the interpolation table are DMA-transferred to the arbitrary tomographic image forming unit. That is, in this embodiment, DMA transfer is performed for every d pieces of interpolation data. Of course, the interpolation data may be transferred one by one.

  In S204, pixel value interpolation calculation is executed for each transferred interpolation data. In that case, the interpolation calculation based on each interpolation data is executed in accordance with the above-described arrangement order. Thereby, a pixel value is obtained for each pixel point, and is written in the output storage unit in S205.

  In S206, j + d is substituted for j, and in S207, it is determined whether j is smaller than i. That is, it is determined whether or not the interpolation operation has been executed for all the pixel points in the display image area. If the interpolation calculation has not been completed for all the pixel points, each process from S203 is repeatedly executed. If the interpolation calculation has been completed for all the pixel points, whether or not to continue the arbitrary slice processing in S208. Is determined, and if not continued, this routine ends.

  In FIG. 9, the order of writing the pixel values to the output storage unit is represented by numbers. Incidentally, “−” represents pixel points that are not to be interpolated, and 0 is given to them as a luminance value by initialization. As described above, in the present embodiment, the pixel values of the pixel points are not calculated according to the raster scan order as in the prior art, but the pixel values are calculated in such an order that the cache function can be used more effectively. Therefore, pixel values are obtained in a jumping manner for each pixel point in the order indicated by each numeral in FIG.

1 is a block diagram showing an overall configuration of an ultrasonic diagnostic apparatus according to the present invention. It is a figure which shows the relationship between three-dimensional space and arbitrary cross sections. It is a figure for demonstrating a three-dimensional interpolation calculation. It is a figure for demonstrating the data transfer system in the case of forming arbitrary tomographic images. It is a figure which shows the interpolation data row | line | column after rearrangement. It is a figure which shows the relationship between a block array and an interpolation calculation. It is a flowchart which shows the preparation process of interpolation data. It is a flowchart which shows the formation process of arbitrary tomographic images. It is explanatory drawing which shows the order of writing of an output memory | storage part.

Explanation of symbols

  14 memory, 22 arbitrary tomographic image forming unit, 24 interpolation table, 25 interpolation table creating unit, 70 block, 76 cache storage unit, 77 core module, 78 output storage unit.

Claims (6)

A first storage unit in which echo data groups obtained from a three-dimensional space in a living body are stored;
For each pixel on the cutting plane set for the three-dimensional space, an interpolation data calculation unit that calculates interpolation data having a plurality of weighting factors to be given to a plurality of neighboring echo data used in the interpolation calculation;
A second storage unit that stores interpolation data of each pixel on the cut surface;
An image forming processor connected to the first storage unit and the second storage unit;
Including
Transfer of echo data from the first storage unit to the image forming processor is executed in units of cache blocks,
The transfer of the interpolation data from the second storage unit to the image forming processor is executed in the arrangement order of the cache blocks to which the representative neighborhood echo data among the plurality of neighborhood echo data belongs,
The image forming processor includes:
A cache storage unit for temporarily storing echo data transferred from the first storage unit;
In accordance with the interpolation data transferred from the second storage unit, the necessary echo data is acquired from the cache storage unit or the first storage unit, and a pixel value calculation unit that obtains a pixel value by executing an interpolation calculation;
An ultrasonic diagnostic apparatus comprising: The apparatus of claim 1.
The pixel value calculation unit obtains the required echo data from the cache storage unit if the echo data required for the interpolation calculation exists in the cache storage unit, and the echo data required for the interpolation calculation is stored in the cache The ultrasonic diagnostic apparatus according to claim 1, wherein the necessary echo data is acquired from the first storage unit when it does not exist in the storage unit. The apparatus of claim 1.
The cache block is composed of echo data strings arranged in chronological order,
2. The ultrasonic diagnostic apparatus according to claim 1, wherein the cache block is arranged in time series order. The apparatus of claim 1.
The ultrasonic diagnostic apparatus according to claim 2, wherein the second storage unit stores an interpolation data string in the order of the cache blocks. The apparatus of claim 1.
The ultrasonic diagnostic apparatus, wherein the interpolation data calculation unit determines effective pixels belonging to an effective display area on the arbitrary cut surface, and calculates interpolation data for each effective pixel. A transmission / reception unit for transmitting / receiving ultrasonic waves;
A first storage unit in which an echo data group obtained from a three-dimensional space in the living body by transmitting and receiving the ultrasonic waves is stored;
A cutting plane setting unit for setting a cutting plane by the user with respect to the three-dimensional space;
For each pixel on the cut surface, an interpolation data calculation unit for calculating interpolation data having a plurality of weighting factors to be given to a plurality of neighboring echo data used in the interpolation calculation,
An arrangement setting unit for setting an arrangement of an interpolation data string composed of interpolation data for each pixel;
A second storage unit storing an interpolation data string having the set sequence;
An image forming processor connected to the first storage unit and the second storage unit;
Including
Transfer of echo data from the first storage unit to the image forming processor is executed in units of cache blocks,
The order of interpolation data sequence stored in the second storage unit corresponds to the order of cache blocks to which representative neighborhood echo data of the plurality of neighborhood echo data belongs, and each interpolation data constituting the interpolation data sequence Are sequentially transferred from the head to the image forming processor,
The image forming processor includes:
A cache storage unit that temporarily stores echo data transferred from the first storage unit in units of the cache blocks;
According to the interpolation data of each pixel sequentially transferred from the second storage unit, if the echo data necessary for the interpolation operation of the pixel exists in the cache storage unit, the necessary echo data is acquired from the cache storage unit, If the necessary echo data does not exist in the cache storage unit, the necessary echo data is acquired from the first storage unit, and the interpolation operation of the pixel is performed based on the acquired plurality of echo data. A pixel value calculation unit for obtaining a pixel value of the pixel by
An image storage unit for storing pixel values calculated for the respective pixels;
An ultrasonic diagnostic apparatus comprising:

Images