Automatic Estimation Method for Module Size in Speckle Tracking of Ultrasonic Sequential Images

Abstract

The present invention provides a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images, including: estimating the trend of tracking error by using the two-way speckle tracking accuracy of a series of template blocks of different sizes; and automatically determining the size of the template block required for block matching by using the two-way speckle tracking accuracy of a series of template blocks of different sizes. The present invention estimates the speckle tracking error through the results of two-way speckle tracking, thereby realizing the adaptive estimation and selection of the size of the template block, and further ensuring the accuracy and stability of speckle tracking, providing more reliable relevant parameters for the clinical diagnosis and treatment of cardiac function.

Description

Technical Field

[0001] The present invention relates to the field of medical image processing, and particularly to a method for automatically estimating the size of a template block for speckle tracking by block matching of cardiac ultrasound sequence images. Background Art

[0002] Speckle tracking technology is a new ultrasound diagnostic technology proposed to avoid the deficiencies of tissue Doppler imaging technology in the direction of medical ultrasound imaging technology, and has been widely used in clinical diagnosis and analysis of cardiac function.

[0003] Currently, speckle tracking algorithms mainly include methods based on block matching, optical flow field, elastic registration, and machine learning.

[0004] The block matching method is the most common and effective calculation method for quantitatively tracking and analyzing tissue movement in ultrasound images. The block matching method mainly involves four aspects of parameters or strategies: 1) the size of the image template block; 2) the size of the search area; 3) the search strategy used by the template block to find the most similar block (the corresponding block of the image module) in the search area; 4) the matching criterion for judging the similarity of two blocks. A large number of experiments have shown that the size of the template block has a great influence on the tracking accuracy of the speckle tracking algorithm based on block matching. When the selected template size is relatively large, the correlation of ultrasound speckles is relatively high and the stability is relatively good. However, when performing motion tracking at this time, some local fine changes or rapid movements cannot be well reflected, and the computational complexity is relatively large. If the selected template block size is relatively small, local fine changes or rapid movements can be better reflected. However, at this time, the correlation and stability of ultrasound speckles will decrease, and the tracking results are easily affected by noise. In addition, parameters such as the imaging resolution of the ultrasound image sequence obtained by different ultrasound devices are different, and the imaging quality of the ultrasound image sequence obtained at different times and the range of organ tissues contained in the images are also different. Therefore, the selection of the template block size is a very important issue. In the existing published articles or relevant materials (for example: Neda Azarmehr, Xujiong Ye, Joseph D. Howes, et al., An optimisation-based iterative approach for speckle tracking echocardiography, Medical & Biological Engineering & Computing, 58: 1309 - 1323, 2020.), it is generally determined by experience and a large number of experimental attempts. Obviously, such a selection lacks self-adaptability. Summary of the Invention

[0005] Aiming at the defects in the prior art, the purpose of the present invention is to provide a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images.

[0006] According to one aspect of the present invention, there is provided a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images, including:

[0007] Obtaining the two-way speckle tracking accuracy of the template blocks of the series sizes;

[0008] Estimating the trend of tracking error by using the two-way speckle tracking accuracy of the template blocks of the series sizes;

[0009] Automatically determining the size of the template block required for block matching by using the two-way speckle tracking accuracy of the template blocks of the series sizes.

[0010] Preferably, the obtaining of the two-way speckle tracking accuracy of the template blocks of the series sizes includes:

[0011] Obtaining cardiac ultrasonic sequence images and the sizes of the series of template blocks;

[0012] Selecting a reference image and a target image from the cardiac ultrasonic sequence images;

[0013] Obtaining the speckles to be tracked in the reference image;

[0014] Forward-tracking the speckles to be tracked through block matching to obtain the corresponding points of the speckles to be tracked in the target image;

[0015] Backward-tracking the corresponding points through block matching to obtain the corresponding tracking points of the corresponding points in the reference image;

[0016] Calculating the distance between the speckles to be tracked and the corresponding tracking points, and calculating the number of speckles within a specified range of the distance according to the size of the module board;

[0017] Obtaining the ratio of the number of speckles to the total number of speckles, which is the two-way speckle tracking accuracy of the template blocks of the series sizes.

[0018] Preferably, the specified range is less than 1.

[0019] Preferably, the reference image is randomly selected from the cardiac ultrasonic sequence images, or an image at the end-diastolic phase is selected as the reference image; the target image is the next frame of the reference image.

[0020] Preferably, the relationship between the two-way speckle tracking accuracy of the template blocks of the series sizes and the tracking error is: the larger the size of the template block, the higher the two-way tracking accuracy and the smaller the forward tracking error; the smaller the size of the template block, the lower the two-way tracking accuracy and the larger the forward tracking error.

[0021] Preferably, the method for automatically determining the size of the template block required for block matching by using the two-way speckle tracking accuracy of a series of size template blocks includes: automatically determining the required template size with the two-way speckle tracking accuracy of a series of size template blocks as the threshold. That is, starting from the smallest template block size, calculate the two-way speckle tracking accuracy of a series of size template blocks under different template block size conditions in turn. When the two-way speckle tracking accuracy of the series of size template blocks is higher than the set accuracy threshold, use the template block size at this time as the finally selected template block size.

[0022] According to a second aspect of the present invention, there is provided a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes any one of the above methods for automatically estimating the size of the template block for speckle tracking of ultrasonic sequence images.

[0023] According to a third aspect of the present invention, there is provided a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements any one of the above methods for automatically estimating the size of the template block for speckle tracking of ultrasonic sequence images.

[0024] According to a fourth aspect of the present invention, there is provided a chip system, including a processor, the coupling of the processor with a memory, and the memory stores program instructions, and when the program instructions stored in the memory are executed by the processor, it implements any one of the above methods for automatically estimating the size of the template block for speckle tracking of ultrasonic sequence images.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention estimates the speckle tracking error through the two-way tracking result of the speckles, thereby realizing the adaptive estimation and selection of the size of the template block, and further ensuring the accuracy and stability of the speckle tracking, and providing more reliable relevant parameters for the clinical diagnosis and treatment of cardiac function. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, objects, and advantages of the present invention will become more obvious:

[0028] Figure 1 is a flowchart of a method for automatically estimating the size of the template block for speckle tracking of ultrasonic sequence images according to an embodiment of the present invention;

[0029] Figure 2 is an exemplary diagram of an echocardiogram image according to a preferred embodiment of the present invention;

[0030] Figure 3 is an exemplary diagram of the change of the two-way speckle tracking accuracy of a series of size template blocks according to a preferred embodiment of the present invention;

[0031] Figure 4 It is an example diagram of the forward tracking error change trend of the series of size template blocks in a preferred embodiment of the present invention. Specific Embodiments

[0032] The present invention will be described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made. These all belong to the protection scope of the present invention.

[0033] Figure 1 It is a flowchart of a method for automatically estimating the size of a template block for speckle tracking in an ultrasonic sequence image provided in an embodiment of the present invention. Refer to Figure 1 As shown, the method for automatically estimating the size of a template block for speckle tracking in an ultrasonic sequence image in this embodiment includes the following steps:

[0034] S1. Obtain the two-way speckle tracking accuracy of the series of size template blocks;

[0035] S2. Estimate the overall tracking error trend using the two-way speckle tracking accuracy of the series of size template blocks;

[0036] S3. Automatically determine the size of the template block required for block matching using the two-way speckle tracking accuracy of the series of size template blocks.

[0037] The template generally refers to a square image block centered on the speckle to be tracked. The series of size modules refers to template blocks of different sizes used for testing. For example, 5*5, 7*7,..., 21*21, 23*23, and other square template blocks of different sizes. That is, the "series of size template blocks" are image blocks of different sizes on an ultrasonic sequence image.

[0038] In this embodiment, for cardiac ultrasonic sequence images, the two-way speckle tracking accuracy of the series of size template blocks is used to estimate the tracking error, so as to adaptively determine the size of the template block in block matching according to the two-way speckle tracking accuracy of the series of size template blocks of different sizes.

[0039] When performing the above S1, obtaining the two-way speckle tracking accuracy of the series of size template blocks includes:

[0040] S101. Obtain the cardiac ultrasonic sequence image and the series of template block sizes;

[0041] S102. Select a reference image and a target image in the cardiac ultrasonic sequence image;

[0042] S103. Obtain the speckles to be tracked in the reference image;

[0043] S104. Forward track the spot to be tracked through a series of size templates to obtain the corresponding points of the spot to be tracked in the target image;

[0044] S105. Reverse track the corresponding points through a series of size templates to obtain the corresponding tracked points of the corresponding points in the reference figure;

[0045] S106. Calculate the distance between the spot to be tracked and the corresponding tracked points, and calculate the number of spots within the specified range of the distance;

[0046] S107. Calculate the ratio of the number of spots to the total number of spots, that is, obtain the two-way spot tracking accuracy of the series of size templates.

[0047] In this embodiment, the spot tracking error is estimated through the two-way spot tracking result, so as to realize the adaptive estimation and selection of the template size, and further ensure the accuracy and stability of spot tracking.

[0048] Specifically, regarding the use of the two-way spot tracking accuracy of a series of size templates, the present invention provides a preferred embodiment. In this preferred embodiment, it is assumed that the cardiac ultrasound sequence image contains n frames of ultrasound images: I i , i = 1, 2,..., n, and the series of template sizes is S j , j = 1, 2,..., m. Generally, according to the application requirements in the art, the series of template sizes are defined by oneself as 5*5, 7*7,.... To estimate the change trend of the tracking error according to the two-way spot tracking accuracy of the series of size templates, in this embodiment, an arbitrary frame of image is taken as the reference image, denoted as: I ref . Of course, in other embodiments, it can also be selected according to certain rules, such as selecting a frame of image at the end-diastolic phase. Taking the next frame of image of I ref as the target image, denoted as: I tag .

[0049] Assume that there are N spots to be tracked in I ref , and the coordinates are: P k = (x k , y k ), k = 1, 2,..., N. Correspondingly, the coordinates in I tag are: T k = (tx k , ty k ), k = 1, 2,..., N.

[0050] Taking S j , j ∈ {1, 2,..., m} as the template size, first forward track through block matching to obtain the corresponding coordinates of the N spots to be tracked in I tag : F jk = (fxjk , fy jk ), k = 1, 2, …, N.

[0051] Then, conversely, taking I tag as the reference image and I ref as the target image, N points: F jk = (fx jk , fy jk ), k = 1, 2, …, N, are speckles, S j is the template block size, and perform block - matching reverse tracking to obtain the corresponding coordinates of the N speckles in I ref : B jk = (bx jk , by jk ), k = 1, 2, …, N.

[0052] Let the distance between point P k and point B jk be: D jk , and for the template block size S j , the number of speckles with D jk less than 1 is N j , then define the corresponding two - way speckle tracking accuracy of the series of size template blocks as ACC j = N j / N. In this embodiment, D jk less than 1 enables the tracking accuracy to be at the sub - pixel level.

[0053] Obviously, ACC j ∈[0, 1], and moreover, the closer ACC j is to 0, the fewer the number of accurately tracked speckles, and the greater the tracking error may be. Conversely, the closer ACC j is to 1, the more the number of accurately tracked speckles, and the smaller the tracking error may be.

[0054] To obtain an overall tracking error trend estimate of bidirectional speckle tracking using series-size template blocks, a preferred embodiment is provided. In this embodiment, an echocardiogram from the publicly available simulation database (Alessandrini M, Chakraborty B, Heyde B, et al., Realistic vendor specific synthetic ultrasound data for quality assurance of 2-D speckle tracking echocardiography: simulation pipeline and open access database. IEEE Trans Ultrason Ferroelectr Freq Control 65(3):411–22, 2018.) is used. As Figure 2 shown in (a) therein is the reference image, Figure 2 shown in (b) therein is the target image, Figure 2 shown in (c) therein is a schematic diagram of 180 speckle positions. In this embodiment, the series-size template blocks use S j = 11, 13, 15, …, 49, 51. As Figure 3 shown, it is an example diagram of the change in bidirectional speckle tracking accuracy based on series-size (S j = 11*11, 13*13, 15*15, …, 49*49, 51*51) template blocks. Figure 4 shown is an example diagram of the change in the average tracking error of forward tracking based on series-size (S j = 11*11, 13*13, 15*15, …, 49*49, 51*51) template blocks.

[0055] From Figure 3 and Figure 4 , it can be seen that for block-matching speckle tracking, the larger the size of the template block, the higher the bidirectional tracking accuracy and the smaller the forward tracking error. Conversely, the smaller the size of the template block, the lower the bidirectional tracking accuracy and the larger the forward tracking error. It can be seen that the change trend of the forward tracking error of the speckle can be estimated using the change trend of the bidirectional tracking accuracy.

[0056] To automatically determine the size of the template block for block matching using the bidirectional speckle tracking accuracy of the series-size template blocks, the present invention provides a preferred embodiment. In this embodiment, considering the change trend of the tracking error of speckle tracking using template blocks of different sizes, and comprehensively considering the bidirectional speckle tracking accuracy and the computational amount of the series-size template blocks, the size of the template block required for block matching is automatically determined.

[0057] Comparing Figure 3 andFigure 4 On the one hand, it can be seen that the change trend of the forward tracking error of the speckle can be estimated by using the change trend of the two-way speckle tracking accuracy of the series of size templates. On the other hand, it can be seen that as the size of the template increases, the increase in the two-way tracking accuracy of the speckle slows down. And as the size of the template increases, the amount of speckle tracking calculation increases rapidly.

[0058] Taking into account the two-way speckle tracking accuracy and the calculation amount of the series of size templates, a preferred embodiment is provided to automatically determine the required template size. That is, starting from the smallest template size, calculate the two-way speckle tracking accuracy of the series of size templates under different template size conditions in turn. When the two-way speckle tracking accuracy of the series of size templates is higher than the set accuracy threshold of 0.8, the template size at this time is the final template size to be selected. Thus, the automatic determination of the template size for the block matching speckle tracking algorithm is realized. In other embodiments, the accuracy threshold is not limited to 0.8. The automatic determination of the two-way speckle tracking accuracy of the series of size templates in this embodiment for the block size required for block matching provides favorable conditions for obtaining better block matching speckle tracking results.

[0059] Based on the same concept of the above embodiments, the present invention provides an embodiment of a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes a method for automatically estimating the template size for speckle tracking of ultrasonic sequence images in any of the above embodiments.

[0060] Based on the same concept of the above embodiments, the present invention provides an embodiment of a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements a method for automatically estimating the template size for speckle tracking of ultrasonic sequence images in any of the above embodiments.

[0061] Based on the same concept of the above embodiments, the present invention provides an embodiment of a chip system, including a processor, the coupling of the processor with a memory, and the memory stores program instructions, and when the program instructions stored in the memory are executed by the processor, it implements a method for automatically estimating the template size for speckle tracking of ultrasonic sequence images in any of the above embodiments.

[0062] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various deformations or modifications within the scope of the claims, which does not affect the essence of the present invention. The above preferred features can be combined arbitrarily without conflict.

Claims

1. An automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images, characterized in that, it includes: Obtaining the two-way speckle tracking accuracy of a series of template blocks of different sizes; Estimating the trend of tracking error by using the two-way speckle tracking accuracy of the series of template blocks of different sizes; Automatically determining the size of the template block required for block matching by using the two-way speckle tracking accuracy of the series of template blocks of different sizes; wherein, the obtaining of the two-way speckle tracking accuracy of the series of template blocks of different sizes includes: Obtaining cardiac ultrasound sequence images and a series of template block sizes; Selecting a reference image and a target image in the cardiac ultrasound sequence images; Obtaining the speckles to be tracked in the reference image; Forward tracking the speckles to be tracked through the series of template blocks of different sizes to obtain the corresponding points of the speckles to be tracked in the target image; Backward tracking the corresponding points through the series of template blocks of different sizes to obtain the corresponding tracking points of the corresponding points in the reference image; Calculating the distance between the speckles to be tracked and the corresponding tracking points, and calculating the number of speckles within a specified range of the distance; Calculating the ratio of the number of speckles to the total number of speckles, that is, obtaining the two-way speckle tracking accuracy of the series of template blocks of different sizes.

2. The automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images according to claim 1, characterized in that: The specified range is less than 1.

3. The automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images according to claim 1, characterized in that: The reference image is randomly selected from one frame of the cardiac ultrasound sequence images or a frame of image is selected at the end-diastolic phase; the target image is the next frame of the reference image.

4. The automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images according to claim 1, characterized in that, The relationship between the two-way speckle tracking accuracy of the series of template blocks of different sizes and the trend of tracking error is: the larger the size of the template block, the higher the two-way speckle tracking accuracy of the series of template blocks of different sizes, and the smaller the forward tracking error; the smaller the size of the template block, the lower the two-way speckle tracking accuracy of the series of template blocks of different sizes, and the larger the forward tracking error.

5. The automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images according to claim 1, characterized in that, The automatically determining the size of the template block required for block matching by using the two-way speckle tracking accuracy of the series of template blocks of different sizes includes: using the two-way speckle tracking accuracy of a certain series of template blocks of different sizes as a threshold to automatically determine the required template size.

6. The automatic estimation method for the size of the template block in speckle tracking of ultrasonic sequence images according to claim 5, characterized in that, Starting from the smallest template size, calculating the two-way speckle tracking accuracy of the series of template blocks of different sizes under different template size conditions in turn; When the two-way speckle tracking accuracy of the series of template blocks of different sizes is higher than the set accuracy threshold, using the template size at this time as the finally selected template size.

7. A computer device, including a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images according to any one of claims 1-6.

8. A computer-readable storage medium, on which a computer program is stored, characterized in that, when the program is executed by a processor, it implements a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images according to any one of claims 1-6.

9. A chip system, characterized in that, it includes a processor, the processor is coupled with a memory, the memory stores program instructions, and when the program instructions stored in the memory are executed by the processor, it implements a method for automatically estimating the size of a template block for speckle tracking in ultrasonic sequence images according to any one of claims 1-6.

Images