A method and system for visual enhancement of cardiac ultrasound images

Abstract

This invention provides a method and system for visual enhancement of cardiac ultrasound images. The method involves calculating the structural similarity between the cardiac ultrasound image to be enhanced and its left and right adjacent cardiac ultrasound images, and calculating the two-dimensional entropy of both the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the adjacent cardiac ultrasound images. At least one reference image is selected from the adjacent cardiac ultrasound images based on the structural similarity and two-dimensional entropy. A target search window is obtained based on the similarity of the target search window with its neighboring blocks, the similarity with a search window, and the distance between the reference image containing the search window and the cardiac ultrasound image to be enhanced. The enhanced pixel values ​​are obtained based on the target search window, the neighboring blocks of the cardiac ultrasound image to be enhanced, and the search window. This invention enhances cardiac ultrasound images not only by considering the image itself but also by considering adjacent images, resulting in a more significant enhancement effect.

Description

Technical Field

[0001] This invention relates to the field of image processing, and in particular to a method and system for visual enhancement of cardiac ultrasound images. Background Technology

[0002] The heart acts like a pump in the human body, continuously pushing blood to all parts of the body. It is also the guardian of health, maintaining the continuous functioning of life. Through its rhythmic beating, it delivers oxygen- and nutrient-rich blood to every cell, enabling cells to respire and generate the necessary energy. The rhythmic beating of the heart, regulated by a pacemaker, ensures uninterrupted blood flow, thus maintaining normal bodily functions. Simultaneously, the heart assists in the elimination of metabolic waste by carrying carbon dioxide and waste products back to organs such as the lungs and kidneys, maintaining the stability of the body's internal environment. However, in modern society, unhealthy lifestyles, unhealthy diets, and lack of exercise have led to a rise in heart disease. Heart disease, high blood pressure, and arrhythmias have become major factors affecting human health.

[0003] Echocardiography, as an important part of medical imaging, has wide applications and significant clinical value. It uses ultrasound technology to acquire images that allow for real-time, non-invasive observation and assessment of the heart's structure and function. Doctors can use ultrasound images to evaluate parameters such as the heart's size, shape, and wall thickness to diagnose the type and severity of heart disease. For example, echocardiography can detect abnormalities in heart valves, the extent of myocardial infarction, and myocardial contractile function, thus guiding doctors in developing treatment plans. Furthermore, before cardiac surgery, doctors can use ultrasound images to accurately locate lesions, ensuring the precision and safety of the procedure. During interventional procedures, echocardiography can display the position and effects of therapeutic instruments in real time, enabling doctors to perform operations more precisely.

[0004] The quality of cardiac ultrasound images is often limited by a variety of factors, such as the acoustic properties of the tissues, image noise, and limitations of the imaging equipment. Therefore, the application of image enhancement techniques can overcome these limitations, enabling physicians to better interpret the images and make more reliable clinical judgments. Summary of the Invention

[0005] To address the above problems, the present invention provides a method for visual enhancement of cardiac ultrasound images, the method comprising the following steps:

[0006] Acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy.

[0007] The size of the neighborhood block and the size of the search window are determined based on the two-dimensional entropy of the echocardiogram image to be enhanced. The similarity s1 between the neighborhood block of each pixel in the echocardiogram image to be enhanced and the neighborhood block of each pixel in the at least one parameter image is calculated. The similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image is calculated. The target search window is obtained based on the similarity s1, the similarity s2, and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced.

[0008] The values ​​of the enhanced pixels are obtained based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window.

[0009] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0010] The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

[0011] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0012] Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance.

[0013] According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , N is a natural number.

[0014] Preferably, determining the size of the neighborhood block and the search window size based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced specifically involves:

[0015] The ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to the standard value is calculated. The product of the ratio and the first scale is used as the size of the neighborhood block, and the product of the ratio and the second scale is used as the size of the search window, wherein the second scale is larger than the first scale.

[0016] Preferably, the step of obtaining the target search window based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the echocardiogram image to be enhanced specifically involves:

[0017] Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance;

[0018] Select the search window with the largest value (w) as the target search window.

[0019] Preferably, obtaining the enhanced pixel value based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window specifically involves:

[0020] The first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced is calculated based on the nonlocal mean filtering method.

[0021] Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced.

[0022] The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method.

[0023] The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

[0024] In addition, the present invention also provides a visual enhancement system for cardiac ultrasound images, the system comprising the following modules:

[0025] A reference image module is used to acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, and select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy.

[0026] The target search window module is used to determine the size of the neighborhood block and the search window size based on the two-dimensional entropy of the echocardiogram image to be enhanced, calculate the similarity s1 between the neighborhood block of each pixel in the echocardiogram image to be enhanced and the neighborhood block of each pixel in the at least one parameter image, and calculate the similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image. The target search window is obtained based on the similarity s1, the similarity s2 and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced.

[0027] The enhancement module is used to obtain the value of the enhanced pixel based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window.

[0028] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0029] The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

[0030] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0031] Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance.

[0032] According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , N is a natural number.

[0033] Preferably, determining the size of the neighborhood block and the search window size based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced specifically involves:

[0034] The ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to the standard value is calculated. The product of the ratio and the first scale is used as the size of the neighborhood block, and the product of the ratio and the second scale is used as the size of the search window, wherein the second scale is larger than the first scale.

[0035] Preferably, the step of obtaining the target search window based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the echocardiogram image to be enhanced specifically involves:

[0036] Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance;

[0037] Select the search window with the largest value (w) as the target search window.

[0038] Preferably, obtaining the enhanced pixel value based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window specifically involves:

[0039] The first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced is calculated based on the nonlocal mean filtering method.

[0040] Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced.

[0041] The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method.

[0042] The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

[0043] Finally, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described above.

[0044] To address the issue that existing enhancement methods are not very effective in enhancing cardiac ultrasound images, this invention proposes to enhance the image to be enhanced by using adjacent images. A reference image is selected using the image's two-dimensional entropy and structural similarity. Then, a target search window is chosen from the reference image, and the image to be enhanced is further enhanced based on this target search window. This approach considers not only the image to be enhanced itself but also information from adjacent images, resulting in a more significant enhancement effect. Attached Figure Description

[0045] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a flowchart of Example 1.

[0047] Figure 2 This is a diagram illustrating the search window and neighborhood blocks.

[0048] Figure 3 This is a comparison image of the image before and after enhancement.

[0049] Figure 4 This is a schematic diagram of the structure of Example 2. Detailed Implementation

[0050] In this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, without necessarily requiring or implying any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1

[0053] This invention provides a method for visual enhancement of cardiac ultrasound images, such as... Figure 1 As shown, the method includes the following steps:

[0054] S1. Acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy.

[0055] A cardiac ultrasound examination is a continuous process, displayed as a video on the doctor's computer. When generating reports or measuring cardiac data, a relatively clear image is typically used as the basis. The video consists of many frames. When enhancing the cardiac ultrasound image to be enhanced, surrounding ultrasound images can also be used. Based on this, the present invention also refers to the ultrasound images surrounding the cardiac ultrasound image to be enhanced. Specifically, it acquires cardiac ultrasound images that are temporally adjacent to the cardiac ultrasound image to be enhanced. Temporal order refers to the order in which the images are generated. In another embodiment, the temporally adjacent cardiac ultrasound images can also be acquired according to the frame order in the video. The temporally adjacent cardiac ultrasound images refer to the ultrasound images located to the left and right of the cardiac ultrasound image to be enhanced.

[0056] Structural Similarity Index (SSIM) is used to represent the degree of structural similarity between images. Structural similarity helps avoid showing excessive differences between adjacent left and right echocardiogram images. The formula for calculating structural similarity is:

[0057]

[0058] in, , Let x and y represent the average values ​​of the image, respectively. , Let x and y represent the variances of the image, respectively. , For parameters, The covariance of image x and y is used to avoid an excessively small denominator. Other methods exist for calculating structural similarity, such as Multi-Scale SSIM, but this invention does not limit the specific method used for calculating structural similarity.

[0059] The structural similarity between each of the left and right adjacent cardiac ultrasound images and the ultrasound image to be enhanced is obtained through structural similarity.

[0060] Entropy refers to the degree of disorder in information. Between two images, the one with more detail has a higher entropy. The two-dimensional entropy of an image also contains its spatial features, and its calculation formula is:

[0061]

[0062] in, , MN represents the frequency of occurrence of the feature pair (i, j), and MN represents the number of pixels in the image.

[0063] Each image corresponds to a two-dimensional entropy, thus obtaining the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images. If there are 6 left and right adjacent cardiac ultrasound images, plus the cardiac ultrasound image to be enhanced itself, 7 two-dimensional entropies can be obtained.

[0064] Then, based on the obtained structural similarity and the two-dimensional entropy, at least one reference image is selected from the left and right adjacent cardiac ultrasound images.

[0065] S2, determine the size of the neighborhood block and the search window size based on the two-dimensional entropy of the echocardiogram image to be enhanced, calculate the similarity s1 between the neighborhood block of each pixel in the echocardiogram image to be enhanced and the neighborhood block of each pixel in the at least one parameter image, and calculate the similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image. Obtain the target search window based on the similarity s1, the similarity s2, and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced.

[0066] Neighborhood blocks and search windows are concepts in non-local means (NL-means) filtering, such as... Figure 2As shown, a neighborhood block refers to a region centered on a single pixel. The search window is also a region centered on a single pixel, but the size of the search window is larger than the size of the neighborhood block. For example, if a pixel has coordinates (2,5) and its neighborhood block is a 5×5 region centered at (2,5), then the neighborhood block of pixel (2,5) is a 5×5 region. If the search window is a 20×20 region, then the search window for pixel (2,5) is a 20×20 region centered at (2,5).

[0067] In nonlocal mean filtering, the size of the neighborhood block and the search window significantly impact the image enhancement effect. An excessively large search window introduces too much irrelevant information and may even introduce noise into the enhanced image. Conversely, a search window that is too small will result in a less effective enhancement. While a small neighborhood block can preserve image details, it is less effective at filtering noise. The two-dimensional entropy of an image reflects its details; determining the size of the neighborhood block and the search window using the two-dimensional entropy of the echocardiogram image to be enhanced allows for adaptive adjustment of these dimensions.

[0068] After determining the neighborhood block size and search window size, the similarity s1 between the neighborhood blocks of each pixel in the echocardiogram image to be enhanced and the neighborhood blocks of each pixel in the at least one parameter image is calculated, and the similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image is calculated. Then, the target search window is obtained based on the similarity s1, similarity s2, and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced.

[0069] S3, obtain the value of the enhanced pixel based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window.

[0070] Image enhancement is performed pixel-by-pixel; the enhanced image is obtained after enhancing all pixels of the entire image. In this invention, the enhancement of a single pixel is no longer limited to the search window but also utilizes information from the target search window. Through steps S1 and S2, the obtained target search window is most closely similar to the target search window of the pixel to be enhanced. The information in the target search window is helpful in enhancing the pixel to be enhanced. A comparison of the enhancement effects is shown in the figure below. Figure 3 As shown.

[0071] In an optional embodiment, selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0072] The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

[0073] First, it is determined whether the structural similarity is within a preset range. In a more specific embodiment, the preset range is [0.8, 1]. In another specific embodiment, the left and right adjacent echocardiogram images with a structural similarity greater than a threshold are placed into a first image set, where the threshold is 0.8. If the structural similarity is within the preset range, the N images whose two-dimensional entropy is closest to that of the echocardiogram image to be enhanced are preferentially selected as at least one reference image, where N is a pre-selected value.

[0074] In another optional embodiment, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0075] Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance.

[0076] Two-dimensional entropy also indicates the degree of similarity between two images. If two images are completely identical, their two-dimensional entropies are also identical. By calculating the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, the degree of difference between the cardiac ultrasound image to be enhanced and the left and right adjacent ultrasound images can be obtained.

[0077] According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , ,and , N is a natural number.

[0078] Assumption =0.8、 =0.2, the structural similarity between one of the two adjacent cardiac ultrasound images, p1, and the ultrasound image to be enhanced. The normalized variance is 0.9. If the value is 0.2, then substituting it into the above formula yields r = 1.776; the structural similarity between one of the two adjacent cardiac ultrasound images p2 and the ultrasound image to be enhanced. The normalized variance is 0.9. If the value is 0.5, then substituting it into the above formula yields r = 1.673. p1 is more likely to be selected as the reference image than p2.

[0079] A higher two-dimensional entropy indicates more detail in the echocardiogram image. In this case, the size of the neighborhood block and search window can be appropriately increased to reduce the impact of noise. In an optional embodiment, determining the size of the neighborhood block and search window based on the two-dimensional entropy of the echocardiogram image to be enhanced specifically involves:

[0080] The ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to the standard value is calculated. The product of this ratio and a first scale is used as the size of the neighborhood block, and the product of this ratio and a second scale is used as the size of the search window, where the second scale is larger than the first scale. If the ratio is a decimal, it is rounded up.

[0081] The target search window is the search window that most closely approximates the neighborhood block of the pixel to be enhanced in the left and right adjacent ultrasound images. In an optional embodiment, the target search window is obtained based on similarity s1, similarity s2, and the distance between the reference image where the search window is located and the cardiac ultrasound image to be enhanced, specifically as follows:

[0082] Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance;

[0083] Select the search window with the largest value (w) as the target search window.

[0084] In another alternative embodiment, the step of obtaining the target search window based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the echocardiogram image to be enhanced specifically involves:

[0085] Calculate the product of the first weight and similarity s1, and the product of the second weight and similarity s2. Use the ratio of the sum of these two products to the distance as the influence of the search window, and select the search window with the greatest influence as the target search window. The first weight is greater than the second weight, and both the first and second weights are located between (0, 1). The sum of the first and second weights is 1.

[0086] When calculating the value of a pixel to be enhanced, this invention relies not only on the image containing the pixel but also on a target search window. In an optional embodiment, the process of obtaining the enhanced pixel value based on the target search window, the neighborhood block of the echocardiogram image to be enhanced, and the search window specifically involves:

[0087] The first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced is calculated based on the nonlocal mean filtering method.

[0088] For the pixel to be enhanced, a first pixel value is calculated in the image containing the pixel using a nonlocal means filtering method. The pixel to be enhanced is located in the echocardiogram image to be enhanced. After all points in the echocardiogram image are enhanced, the resulting image is the enhanced echocardiogram image. It should be noted that when performing enhancement using the nonlocal means filtering method, the boundary points of the search window can be handled using padding or other methods, or the method used in nonlocal means filtering can be employed. This invention does not specifically limit this approach.

[0089] Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced.

[0090] The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method.

[0091] The pixel to be enhanced corresponds to a neighborhood block and a search window, and the pixel to be enhanced is located at the center of the corresponding neighborhood block and search window. The target search window and the search window of the pixel to be enhanced are the same size. The neighborhood block corresponding to the center position of the target search window is replaced with the neighborhood block of the pixel to be enhanced, and then the second pixel value of the pixel to be enhanced in the target search window after replacement is calculated.

[0092] The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

[0093] Then, the average of the first pixel value and the second pixel value is taken as the enhanced pixel value. In an alternative approach, the enhanced pixel value is calculated based on the first pixel value, the second pixel value, and similarity s1 and / or similarity s2. For example, according to the formula... .

[0094] Example 2

[0095] The present invention also provides a visual enhancement system for cardiac ultrasound images, such as Figure 4 As shown, the system includes the following modules:

[0096] The reference image module 101 is used to acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, and select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy.

[0097] The target search window module 102 is used to determine the size of the neighborhood block and the size of the search window based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced, calculate the similarity s1 between the neighborhood block of each pixel in the cardiac ultrasound image to be enhanced and the neighborhood block of each pixel in the at least one parameter image, and calculate the similarity s2 between the search window of each pixel in the cardiac ultrasound image to be enhanced and the search window of each pixel in the at least one parameter image, and obtain the target search window based on the similarity s1, the similarity s2 and the distance between the reference image where the search window is located and the cardiac ultrasound image to be enhanced.

[0098] Enhancement module 103 is used to obtain the value of the enhanced pixel based on the target search window and the neighborhood block and search window of the cardiac ultrasound image to be enhanced.

[0099] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0100] The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

[0101] Preferably, the step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves:

[0102] Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance.

[0103] According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , N is a natural number.

[0104] Preferably, determining the size of the neighborhood block and the search window size based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced specifically involves:

[0105] The ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to the standard value is calculated. The product of the ratio and the first scale is used as the size of the neighborhood block, and the product of the ratio and the second scale is used as the size of the search window, wherein the second scale is larger than the first scale.

[0106] Preferably, the step of obtaining the target search window based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the echocardiogram image to be enhanced specifically involves:

[0107] Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance;

[0108] Select the search window with the largest value (w) as the target search window.

[0109] Preferably, obtaining the enhanced pixel value based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window specifically involves:

[0110] The first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced is calculated based on the nonlocal mean filtering method.

[0111] Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced.

[0112] The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method.

[0113] The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

[0114] Example 3

[0115] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a computer, implements the method described in Embodiment 1.

[0116] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of a necessary general-purpose hardware platform, or by a combination of hardware and software. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a computer product. The present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for visual enhancement of cardiac ultrasound images, characterized in that, The method includes the following steps: Acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced. Select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy. The size of the neighborhood block and the size of the search window are determined based on the two-dimensional entropy of the echocardiogram image to be enhanced. The similarity s1 between the neighborhood block of each pixel in the echocardiogram image to be enhanced and the neighborhood block of each pixel in the at least one parameter image is calculated. The similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image is calculated. The target search window is obtained based on the similarity s1, the similarity s2, and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced. The values ​​of the enhanced pixels are obtained based on the target search window, the neighborhood block of the cardiac ultrasound image to be enhanced, and the search window. The step of determining the size of the neighborhood block and the search window size based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced specifically involves: calculating the ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to a standard value, using the product of the ratio and a first scale as the size of the neighborhood block, and using the product of the ratio and a second scale as the size of the search window, wherein the second scale is larger than the first scale. The target search window is obtained based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the cardiac ultrasound image to be enhanced. Specifically: Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance; select the search window with the largest w as the target search window; The step of obtaining the enhanced pixel value based on the target search window and the neighborhood block and search window of the cardiac ultrasound image to be enhanced specifically involves: calculating the first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced using a nonlocal mean filtering method. Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced. The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method. The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

2. The method as described in claim 1, characterized in that, The step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves: The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

3. The method as described in claim 1, characterized in that, The step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves: Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance. According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , N is a natural number.

4. A visual enhancement system for cardiac ultrasound images, characterized in that, The system includes the following modules: A reference image module is used to acquire the cardiac ultrasound image to be enhanced, and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the structural similarity between the cardiac ultrasound image to be enhanced and the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, calculate the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced, and select at least one reference image from the cardiac ultrasound images that are chronologically adjacent to the cardiac ultrasound image to be enhanced based on the structural similarity and the two-dimensional entropy. The target search window module is used to determine the size of the neighborhood block and the search window size based on the two-dimensional entropy of the echocardiogram image to be enhanced, calculate the similarity s1 between the neighborhood block of each pixel in the echocardiogram image to be enhanced and the neighborhood block of each pixel in the at least one parameter image, and calculate the similarity s2 between the search window of each pixel in the echocardiogram image to be enhanced and the search window of each pixel in the at least one parameter image. The target search window is obtained based on the similarity s1, the similarity s2 and the distance between the reference image where the search window is located and the echocardiogram image to be enhanced. The enhancement module is used to obtain the values ​​of the enhanced pixels based on the target search window, the neighborhood blocks of the cardiac ultrasound image to be enhanced, and the search window itself. The step of determining the size of the neighborhood block and the search window size based on the two-dimensional entropy of the cardiac ultrasound image to be enhanced specifically involves: calculating the ratio of the two-dimensional entropy of the cardiac ultrasound image to be enhanced to a standard value, using the product of the ratio and a first scale as the size of the neighborhood block, and using the product of the ratio and a second scale as the size of the search window, wherein the second scale is larger than the first scale. The target search window is obtained based on similarity s1, similarity s2, and the distance between the reference image containing the search window and the cardiac ultrasound image to be enhanced. Specifically: Calculate the average of the similarity scores s1 and s2. Calculated according to the formula Calculate the weights, where d is the distance; select the search window with the largest w as the target search window; The step of obtaining the enhanced pixel value based on the target search window and the neighborhood block and search window of the cardiac ultrasound image to be enhanced specifically involves: calculating the first pixel value of the pixel to be enhanced in the neighborhood block and search window of the cardiac ultrasound image to be enhanced using a nonlocal mean filtering method. Obtain the neighborhood block corresponding to the location of the pixel to be enhanced in the target search window, and replace the neighborhood block corresponding to the location with the neighborhood block of the pixel to be enhanced. The second pixel value of the pixel to be enhanced in the target search window is calculated using the nonlocal mean filtering method. The average of the first and second pixel values ​​is used as the value of the enhanced pixel.

5. The system as described in claim 4, characterized in that, The step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves: The left and right adjacent echocardiogram images with structural similarity within a preset range are placed into a first image set, and the N images whose two-dimensional entropy of the images in the first image set is closest to the two-dimensional entropy of the echocardiogram image to be enhanced are used as at least one reference image; where N is a natural number.

6. The system as described in claim 4, characterized in that, The step of selecting at least one reference image from the left and right adjacent echocardiogram images based on structural similarity and the two-dimensional entropy specifically involves: Calculate the variance of the two-dimensional entropy of the cardiac ultrasound image to be enhanced and the two-dimensional entropy of the left and right adjacent cardiac ultrasound images, and normalize the variance. According to the formula The correlation degree is calculated, and the left and right adjacent echocardiogram images are sorted according to the correlation degree. N images are selected from the sorted left and right adjacent echocardiogram images as the at least one reference image; wherein, The structural similarity is... The variance is the normalized variance. , It is the adjustment coefficient, and , N is a natural number.

Images