Image processing model training method and device, image processing equipment and medium
By obtaining the hollow map in ultrasonic blood flow imaging and compensating the velocity map, the problem of hollow phenomenon in color ultrasonic blood flow images is solved, the accuracy and accuracy of the image are improved, and the best display effect is ensured in clinical practice .
Patent Information
- Application Number
- CN202310486632.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-04
- Publication Date
- 2024-11-08
AI Technical Summary
In ultrasonic blood flow imaging, due to the interference of human tissue on ultrasonic reflection and the error of ultrasonic front-end circuit components, holes appear in the color ultrasonic blood flow image, affecting the image quality and doctors' judgment of the condition.
By obtaining the ultrasonic blood flow diagram, traverse the speed diagram according to the preset method of judging the hollow point to obtain the hollow map, and compensate the speed diagram based on the hollow map and grayscale map to obtain the compensated ultrasonic blood flow diagram to improve the accuracy of the image degree and accuracy.
It improves the accuracy and accuracy of ultrasonic blood flow images, so that the images maintain the best display results in the clinical process, reduces the hollow phenomenon, and enhances the true reflection of blood flow status.
Smart Images

Figure CN118918040A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of ultrasonic blood flow image processing, and in particular to an image processing model training method, device, image processing equipment and medium. Background Art
[0002] At present, B-type grayscale images and color blood flow imaging are two major aspects of modern medical ultrasound imaging. In ultrasound imaging, the interference of human tissue in the reflection of ultrasound waves will not only cause speckle noise in ultrasound B-type grayscale images, but also in color ultrasound blood flow images, resulting in "holes" in color ultrasound blood flow images. In addition, since the blood flow, tissue, and blood vessel walls in the human body are not completely static reflectors, as well as the errors of ultrasound front-end circuit components, it is easy to form false isolated color image blocks on the color image, namely isolated noise. Therefore, in order to reduce the "hole" phenomenon of color images in color ultrasound blood flow imaging, the image post-processing technology involved is an important link affecting image quality.
[0003] At present, in the process of blood flow imaging, due to the influence of low signal-to-noise ratio or system performance limitation, the phenomenon of holes appears in the ultrasound blood flow image, making the blood flow in the blood vessels incomplete, affecting the appearance and the doctor's overall judgment of the disease. However, due to the different performances of different color ultrasound front ends and color blood flow imaging processors, the threshold values and rule combinations in the threshold rules during hybrid post-processing vary greatly, resulting in the difficulty of having a recognized and systematic hybrid post-processing technology and method. Summary of the invention
[0004] The purpose of the present invention is to provide an image processing model training method, device, image processing equipment and medium, which can traverse the velocity map according to the velocity map according to a preset method for determining void points to obtain a void map, and then compensate the velocity map according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. The image displayed on the interface can better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and maintain the best display result of the image during the clinical process.
[0005] According to one aspect of the present invention, the present invention provides an image processing model training method, comprising:
[0006] Acquiring an ultrasonic blood flow map, wherein the ultrasonic blood flow map includes a grayscale map, a velocity map, a variance map, and an energy map;
[0007] Traversing the velocity graph according to a preset method for determining a hole point, and obtaining a hole graph corresponding to the velocity graph;
[0008] The velocity map is compensated according to the cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
[0009] Optionally, traversing the speed graph according to a preset method for determining a hole point to obtain a hole graph corresponding to the speed graph includes:
[0010] According to the first method for determining a hole point, the velocity graph is traversed to obtain a first hole graph;
[0011] According to the second method for determining a hole point, the velocity graph is traversed to obtain a second hole graph;
[0012] Accordingly, compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the cavity map and the grayscale map includes:
[0013] The velocity map is compensated according to the first cavity map, the second cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
[0014] Optionally, compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the first cavity map, the second cavity map and the grayscale map includes:
[0015] According to the second hole graph, traverse the speed graph to fill the speed graph to obtain a second hole-filling speed graph;
[0016] According to the first hole graph, traverse the second hole filling speed graph to fill the second hole filling speed graph to obtain a third hole filling speed graph;
[0017] Combining the first hole map and the second hole map to obtain a third hole map;
[0018] Determining whether a value of the grayscale image corresponding to the third hole image is a preset threshold;
[0019] If it is the preset threshold, the third hole filling speed map is updated to obtain the compensated ultrasonic blood flow map.
[0020] Optionally, compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the first cavity map, the second cavity map and the grayscale map further includes:
[0021] According to the first hole map, traverse the speed map to fill the speed map to obtain a first hole-filling speed map;
[0022] Determining whether a value of the grayscale image corresponding to the first hole image is a preset threshold;
[0023] If it is the preset threshold, the first hole filling speed map is updated to obtain a compensated ultrasonic blood flow map.
[0024] Optionally, traversing the speed graph according to the first method for determining hole points to obtain a first hole graph includes:
[0025] Traversing the speed graph, determining whether the value of a non-zero point in the speed graph is less than a preset value;
[0026] If it is less than the preset value, the non-zero point is determined to be a hole point, and all the hole points are marked to obtain the first hole map.
[0027] Optionally, before traversing the speed graph according to a preset method for determining a hole point, the method further includes:
[0028] determining whether a value of the speed graph exceeds a preset speed value;
[0029] Determining whether the value of the variance map exceeds a preset variance value;
[0030] Determining whether the value of the energy graph is within a preset energy value range;
[0031] If it exceeds the preset speed value, does not exceed the preset variance value, and is within the preset energy value range, then the corresponding speed map is extracted.
[0032] The present invention also provides an image processing method, comprising:
[0033] obtaining an ultrasound blood flow image to be processed;
[0034] The ultrasonic blood flow map is compensated by using an image processing model to obtain the compensated ultrasonic blood flow map, wherein the image processing model is constructed according to the image processing model training method as described above.
[0035] The present invention provides an image processing model training device, comprising:
[0036] An image acquisition module, used to acquire an ultrasonic blood flow map, wherein the ultrasonic blood flow map includes a grayscale map, a velocity map, a variance map, and an energy map;
[0037] A hole point module, used for traversing the speed graph according to a preset method for determining a hole point, and obtaining a hole graph corresponding to the speed graph;
[0038] A compensation module is used to compensate the velocity map according to the cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
[0039] The present invention provides an image processing device, comprising:
[0040] Memory for storing computer programs;
[0041] A processor is used to implement the image processing model training method as described above when executing the computer program.
[0042] The present invention provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are loaded and executed by a processor, the steps of the image processing model training method as described above are implemented.
[0043] It can be seen that in the present invention, according to the velocity map, the velocity map can be traversed according to the preset method for determining void points to obtain a void map, and then the velocity map can be compensated according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. This can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0044] The present application also provides an image processing model training device, an image processing equipment and a computer-readable storage medium, which have the above-mentioned beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0046] Figure 1 A flowchart of an image processing model training method provided by an embodiment of the present invention;
[0047] Figure 2 A flowchart of another image processing model training method provided by an embodiment of the present invention;
[0048] Figure 3 A flowchart of an image processing method provided by an embodiment of the present invention;
[0049] Figure 4 A structural block diagram of an image processing model training device provided by an embodiment of the present invention;
[0050] Figure 5 A structural block diagram of an image processing device provided by an embodiment of the present invention;
[0051] Figure 6 A structural block diagram of an image processing device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0053] At present, B-type grayscale images and color blood flow imaging are two major aspects of modern medical ultrasound imaging. In ultrasound imaging, the interference of human tissue in the reflection of ultrasound waves will not only cause speckle noise in ultrasound B-type grayscale images, but also in color ultrasound blood flow images, resulting in "holes" in color ultrasound blood flow images. In addition, since the blood flow, tissue, and blood vessel walls in the human body are not completely static reflectors, as well as the errors of ultrasound front-end circuit components, it is easy to form false isolated color image blocks on the color image, namely isolated noise. Therefore, in order to reduce the "hole" phenomenon of color images in color ultrasound blood flow imaging, the image post-processing technology involved is an important link affecting image quality.
[0054] At present, in the process of blood flow imaging, due to the influence of low signal-to-noise ratio or system performance limitation, the phenomenon of holes appears in the ultrasound blood flow image, making the blood flow in the blood vessels incomplete, affecting the appearance and the doctor's overall judgment of the disease. However, due to the different performances of different color ultrasound front ends and color blood flow imaging processors, the threshold values and rule combinations in the threshold rules during hybrid post-processing vary greatly, resulting in the difficulty of having a recognized and systematic hybrid post-processing technology and method.
[0055] In view of this, the present invention provides an image processing model training method, which can traverse the velocity map according to the velocity map according to a preset method for determining void points to obtain a void map, and then compensate the velocity map according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. This method can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0056] The following is a detailed description, please refer to Figure 1 , Figure 1 A flowchart of an image processing model training method provided by an embodiment of the present invention. An image processing model training method according to an embodiment of the present invention may include:
[0057] Step S101: Acquire the collected ultrasonic blood flow map, which includes a grayscale map, a velocity map, a variance map, and an energy map.
[0058] In the embodiment of the present invention, there is no restriction on the method of acquiring the ultrasonic blood flow map. The image acquisition device can be used to collect images first, and then the image processing device can be used to generate the ultrasonic blood flow map, and finally the image display device can be used to present the ultrasonic blood flow map. It should be noted that the process of using the image acquisition device to collect images can be to first use the ultrasonic probe in the image acquisition device to transmit ultrasonic signals to the human body. For example, a color ultrasonic diagnostic instrument can be used to transmit ultrasonic signals to the human body for image acquisition. Then, the echo signal received by the ultrasonic probe is converted into an electrical signal by the conversion circuit in the image acquisition device, wherein the conversion circuit includes a signal amplification circuit, an A / D conversion circuit, a beam synthesis circuit, and a conversion circuit. It should be noted that after the conversion circuit, it is converted into two signals, one signal is a two-dimensional signal for generating a grayscale image signal, and the other signal is a color blood flow signal. The color blood flow signal is subjected to autocorrelation calculation to obtain a velocity map signal, an energy map signal, and a variance map signal. Specifically, the color blood flow signal can be subjected to orthogonal demodulation, impurity filtering, and autocorrelation estimation. The signal is respectively entered into a velocity calculator, a variance calculator, and an energy calculator for calculation to obtain a velocity map signal, an energy map signal, and a variance map signal.
[0059] It should be noted that, in the embodiment of the present invention, an image processing device can be used to generate an ultrasonic blood flow map. Specifically, an image signal generated by processing two signals can be sent to a pseudo-color coding and fusion display module to generate an ultrasonic blood flow map, which is equivalent to processing the grayscale image signal, the velocity image signal, the energy image signal and the variance image signal to obtain an ultrasonic blood flow map including a grayscale image, a velocity image, an energy image and a variance image.
[0060] Step S102: traverse the velocity graph according to a preset method for determining hole points to obtain a hole graph corresponding to the velocity graph.
[0061] Since the interference of human tissue on ultrasonic reflection not only causes speckle noise in the ultrasonic B-type grayscale image, but also causes speckle noise in the color ultrasonic blood flow image, thereby causing "holes" to appear in the color ultrasonic blood flow image, in the embodiment of the present invention, the velocity map can be traversed according to the preset method for determining the hole point to obtain a hole map corresponding to the velocity map, which is equivalent to obtaining an image containing "holes". There is no limitation on the method for determining the hole point, and the hole map can be obtained by using a binarization method or by using a non-zero point determination method.
[0062] Further, in the embodiment of the present invention, before traversing the velocity map according to the preset method for determining the hollow point, the velocity map obtained in step S101 can be preprocessed, which is equivalent to screening the velocity map in the collected ultrasonic blood flow map. Specifically, it can be determined whether the value of the velocity map exceeds the preset velocity value, and whether the value of the variance map in the ultrasonic blood flow map exceeds the preset variance value, and whether the value of the energy map in the ultrasonic blood flow map is within the preset energy value range. If it exceeds the preset velocity value, does not exceed the preset variance value, and is within the preset energy value range, then the corresponding velocity map is extracted. If any of the above judgment conditions is not met, the velocity map is abandoned. Among them, there is no restriction on the value setting of the preset velocity value, the preset variance value, and the preset energy value range, which can be pre-set by the designer or set according to the actual use. There is no restriction on the order of judgment, which can be to first judge the value of the variance map, then judge the value of the energy map, and finally judge the value of the velocity map, or it can be to first judge the value of the energy map, then judge the value of the velocity map, and finally judge the value of the variance map. For example, threshold processing is performed on the speed map to obtain the first speed map (Vth map), which is equivalent to performing threshold screening on the speed map (V), variance map (T) and energy map (P) to extract the speed map, namely the first speed map. If V>Vt, Pt1<P<Pt2, and T<Tt are satisfied at the same time, the corresponding speed map is extracted as the Vth map, wherein Vt is a preset speed value, and the value can be set to 20, Pt1 is the minimum value of the preset energy value range, and the minimum value can be set to 20, Pt2 is the maximum value of the preset energy value range, and the maximum value can be set to 50000, and Tt is the preset variance value, and the value can be set to 150, wherein V is an 8-bit signed number, and V∈[-128,127], T is an 8-bit unsigned number, and T∈[0,255], and P is a 16-bit unsigned number, and P∈[0,65535].
[0063] Further, in the embodiment of the present invention, the grayscale image obtained in step S101 may be preprocessed before traversing the velocity map according to the preset method for determining the hole point. It should be noted that the obtained original grayscale image may be binarized to obtain the grayscale image, wherein there is no restriction on the method for obtaining the original grayscale image corresponding to the current blood flow frame, and the original grayscale image adjacent to a single blood flow frame may be obtained, the average value of several adjacent blood flow frames may be obtained as the value of the original grayscale image, or the weighted value of several adjacent blood flow frames may be obtained as the value of the original grayscale image, wherein the number of frames includes but is not limited to 1, 2 or 3. N frames of adjacent grayscale images are selected, wherein the value of the original grayscale image may be obtained using the formula, and the formula is as follows:
[0064]
[0065] Among them, coefi is the weight corresponding to the adjacent original grayscale image, G i is the value of the original grayscale image corresponding to the current blood flow frame, G is the value of the original grayscale image calculated, where coef i There is no restriction on the setting of and N. It can be preset by the designer according to the requirements, or it can be set according to the actual usage. For example, the value of N can be set to 4, and the corresponding coef i The value of can be set to 0.25, where i∈[1,2,3,4].
[0066] It should be noted that, in the embodiment of the present invention, after obtaining the original grayscale image, the original grayscale image may be binarized to obtain the grayscale image. Specifically, the binarization may be performed using a formula, which is as follows:
[0067]
[0068] Wherein, MG is the grayscale image after binarization processing, and th is the set threshold value, which can be preset by the designer according to the requirements, or can be set according to the actual usage, for example, th∈[40,60], preferably 50.
[0069] In the embodiment of the present invention, by screening the grayscale image and the speed image, and traversing the speed image according to a preset method for determining the hole points, a hole image corresponding to the speed image is obtained, so that the position of the hole point can be determined more accurately, and the precision and accuracy of subsequent compensation of the hole point can be improved.
[0070] Step S103: According to the cavity image and the grayscale image, the velocity image is compensated to obtain a compensated ultrasonic blood flow image.
[0071] In the embodiment of the present invention, the compensated ultrasonic blood flow map can be obtained according to the hole map and the grayscale map, and the speed map can be compensated to obtain the compensated ultrasonic blood flow map. Specifically, the speed map can be traversed according to the hole map to fill the hole filling speed map, and then the grayscale map and the hole filling speed map are compared, and the hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. The filling process can be to use the hole map as a template, traverse the corresponding speed map area one by one, and then fill the area to obtain the hole filling speed map. For example, it can be assumed that there are I areas in total, wherein there is no restriction on the specific value of I, and the jth point in the i-th area is set as mj, and then the K blood flow points closest to mj are found, and the number of positive and negative values in the speed values of these points is counted. If the number of positive values is greater than the number of negative values, the average of all positive value points is taken as the value of the blood flow point for filling, otherwise, the average of all negative value points is taken as the value of the blood flow point for filling, and so on, all the hole points in the speed map are filled to obtain the hole filling speed map. It should be noted that, in the embodiment of the present invention, the process of updating the hole-filling speed map may be to correspond the points in the hole-filling speed map to the points in the grayscale map one by one, and then update the hole-filling speed map according to the grayscale map. For example, assuming that the value of the point in the i-th row and j-th column in the hole map is 1, determine whether the value in the corresponding grayscale map is 1. If it is 1, update the value of the corresponding position in the hole-filling speed map to 0. If it is not 1, determine whether it exceeds the preset range. If it exceeds the preset range, update the second hole-filling speed map to obtain the compensated ultrasonic blood flow map, otherwise no update is performed.
[0072] Based on the above embodiments, the embodiments of the present invention provide an image processing model training method, which can traverse the velocity map according to the velocity map according to a preset method for determining void points to obtain a void map, and then compensate the velocity map according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. This method can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0073] Based on the above embodiment, the present invention introduces a flowchart of a specific image processing model training method. Figure 2 , Figure 2 A flowchart of another image processing model training method provided by an embodiment of the present invention. An image processing model training method according to an embodiment of the present invention may include:
[0074] Step S201: Acquire the collected ultrasonic blood flow map, which includes a grayscale map, a velocity map, a variance map, and an energy map.
[0075] For the specific content of the above step S201, reference may be made to the corresponding content disclosed in the above embodiments, which will not be repeated here.
[0076] Step S202: traverse the velocity graph according to the first method for determining the hole points to obtain a first hole graph, and traverse the velocity graph according to the second method for determining the hole points to obtain a second hole graph.
[0077] In the embodiment of the present invention, the speed graph can be traversed according to the first method for determining the hole point to obtain the first hole graph. It should be noted that the first method for determining the hole point can be to traverse the speed graph to determine whether the value of the non-zero point in the speed graph is less than a preset value. If it is not less than the preset value, the point is not a hole point. If it is less than the preset value, the non-zero point is determined to be a hole point, and all the hole points are marked to obtain the first hole graph, wherein the preset value can be obtained using coef*E, wherein coef is the weight of the corresponding grayscale image, and E is the mean of the non-zero values in the M*M neighborhood of each traversed point. There is no restriction on the numerical setting of coef, which can be preset by the designer according to the requirements, for example, it can be set to 0.5, or it can be set according to the actual usage. For example, all non-zero points in the preprocessed speed map are traversed, and the mean E of the non-zero values in the M*M neighborhood of each point is calculated. If the value of the point is less than the preset value calculated by coef*E, the corresponding point is determined to be a hole point and the point is marked as 1. If the value of the point is not less than the preset value calculated by coef*E, the corresponding point is determined not to be a hole point and the point is marked as 0. And so on, all the hole points are found to obtain the first hole map, where M can preferably be 5.
[0078] In the embodiment of the present invention, the speed map can be traversed according to the second method for determining the hole point to obtain the second hole map. It should be noted that the second method for determining the hole point can be to perform binarization processing on the speed map and then obtain the second hole map. Specifically, the speed map can be binarized to obtain a marked map, and then the marked map is reversed to obtain a region map, and finally the connected regions in the region map are traversed and marked to obtain the second hole map. Specifically, for example, the speed map is binarized using a formula, and the formula is as follows:
[0079]
[0080] Among them, mask is the label image after binarization, abs(Vth) is the absolute value of the pre-processed velocity image, and then the mask is inverted using the formula to obtain maskP (area map), the formula is as follows:
[0081]
[0082] It should be noted that after obtaining the area map, the connected areas in the area map can be traversed and the areas can be marked to obtain a second hole map. Specifically, traversing the eight-connected neighborhood in maskP is equivalent to traversing 8 connected areas with the same value, and marking the eight-connected neighborhood as 1 or 0, where those marked as 1 can represent hole points, and those marked as 0 can represent non-hole points.
[0083] It should be noted that in the embodiment of the present invention, there is no restriction on the judgment order of the first method for judging void points and the second method for judging void points. The first void map can be obtained first and then the second void map, or the second void map can be obtained first and then the first void map.
[0084] Step S203: According to the first cavity image, the second cavity image and the grayscale image, the velocity image is compensated to obtain a compensated ultrasonic blood flow image.
[0085] In the embodiment of the present invention, the speed map can be compensated only according to the first hole map and the grayscale map to obtain the compensated ultrasonic blood flow map. It should be noted that the speed map can be traversed according to the first hole map to fill the speed map to obtain the first hole filling speed map, and then the grayscale map is compared with the first hole filling speed map, and the first hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. It should be noted that the speed map can be filled according to the first hole map by traversing the speed map to obtain the first hole filling speed map, and then it is determined whether the value of the grayscale map corresponding to the first hole map is a preset threshold value. If it is the preset threshold value, the first hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. If it is not the preset threshold value, the first hole filling speed map is not updated. It can also be determined whether the value of the grayscale map corresponding to the first hole map exceeds the preset range. If it exceeds the preset range, the first hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. Otherwise, the first hole filling speed map is not updated. There is no limitation on this in the embodiment of the present invention. Among them, the filling process can be to use the first hole map as a template, traverse the area of the corresponding speed map one by one, and then fill the area to obtain the first hole-filling speed map. For example, it can be assumed that there are I areas in total, set the jth point in the i-th area as mj, and then find the K blood flow points closest to mj, and count the number of positive and negative values in the speed values of these points. If the number of positive values is greater than the number of negative values, take the average of all positive points as the value of the blood flow point for filling, otherwise, take the average of all negative points as the value of the blood flow point for filling, and so on, fill all the hole points in the speed map to obtain the first hole-filling speed map. It should be noted that, in the embodiment of the present invention, the process of updating the first hole-filling speed map may be to correspond the points in the first hole-filling speed map to the points in the grayscale map one by one, and then update the first hole-filling speed map according to the grayscale map. For example, assuming that the value of the point in the i-th row and j-th column in the first hole map is 1, determine whether the value in the corresponding grayscale map is 1. If it is 1, update the value of the corresponding position of the first hole-filling speed map to 0. If it is not 1, do not update the first hole-filling speed map.
[0086] In the embodiment of the present invention, the speed map can also be compensated only according to the second hole map and the grayscale map to obtain the compensated ultrasonic blood flow map. It should be noted that the speed map can be traversed according to the second hole map to fill the speed map to obtain the second hole filling speed map, and then the grayscale map is compared with the second hole filling speed map, and the second hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. It should be noted that the speed map can be traversed according to the second hole map to fill the speed map to obtain the second hole filling speed map, and then it is determined whether the value of the grayscale map corresponding to the second hole map is a preset threshold. If it is the preset threshold, the second hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. If it is not the preset threshold, the second hole filling speed map is not updated. It can also be determined whether the value of the grayscale map corresponding to the second hole map exceeds the preset range. If it exceeds the preset range, the second hole filling speed map is updated to obtain the compensated ultrasonic blood flow map, otherwise it is not updated. There is no limitation on this in the embodiment of the present invention. Among them, the filling process can be to use the second hole map as a template, traverse the area of the corresponding speed map one by one, and then fill the area to obtain the second hole-filling speed map. For example, it can be assumed that there are I areas in total, set the jth point in the i-th area as mj, and then find the K blood flow points closest to mj, and count the number of positive and negative values in the speed values of these points. If the number of positive values is greater than the number of negative values, take the average of all positive points as the value of the blood flow point for filling, otherwise, take the average of all negative points as the value of the blood flow point for filling, and so on, fill all the hole points in the speed map to obtain the second hole-filling speed map. It should be noted that, in the embodiment of the present invention, the process of updating the second hole-filling speed map may be to correspond the points in the second hole-filling speed map to the points in the grayscale map one by one, and then update the second hole-filling speed map according to the grayscale map. For example, assuming that the value of the point in the i-th row and j-th column in the second hole map is 1, determine whether the value in the corresponding grayscale map is 1. If it is 1, update the value of the corresponding position in the second hole-filling speed map to 0. If it is not 1, do not update the second hole-filling speed map.
[0087] Further, in the embodiment of the present invention, the speed map can be compensated according to the first hole map, the second hole map and the grayscale map to obtain the compensated ultrasonic blood flow map. Specifically, the speed map can be filled according to the second hole map, and the second hole filling speed map can be obtained. Then, according to the first hole map, the second hole filling speed map can be filled according to the second hole filling speed map to obtain the third hole filling speed map, and then the first hole map and the second hole map can be merged to obtain the third hole map, and it is determined whether the value of the grayscale map corresponding to the third hole map is a preset threshold value. If it is the preset threshold value, the third hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. If it is not the preset threshold value, the third hole filling speed map is not updated. It can also be determined whether the value of the grayscale map corresponding to the third hole map exceeds the preset range. If it exceeds the preset range, the third hole filling speed map is updated to obtain the compensated ultrasonic blood flow map. Otherwise, the third hole filling speed map is not updated. There is no limitation on this in the embodiment of the present invention. Among them, in the embodiment of the present invention, the speed map can be filled according to the second hole map through the speed map to obtain the second hole filling speed map. Specifically, for example, taking the second hole map as a template, traverse the areas in the preprocessed speed map one by one, assuming that there are I areas in total, set the jth point in the i-th area to be mj, and then find the K blood flow points closest to mj, and count the number of positive and negative values in the velocity values of these points. If the number of positive values is greater than the number of negative values, take the mean of all positive points as the value of the blood flow point for filling, otherwise, take the mean of all negative points as the value of the blood flow point for filling, and so on, until all the hole points in the speed map are filled and the second hole-filling speed map is obtained.
[0088] It should be noted that, in the embodiment of the present invention, after obtaining the second hole filling speed map, the second hole filling speed map can be traversed according to the first hole map to fill the second hole filling speed map, so as to obtain the third hole filling speed map. Specifically, for example, the corresponding points in the second hole filling speed map corresponding to the points marked as 1 in the first hole map can be traversed. If the value of the corresponding point is a positive value, it is determined whether the values of all positive points in the N1*N1 neighborhood of the corresponding point are greater than the value of the corresponding point. If not greater than the value of the corresponding point, the corresponding point is not recorded. If greater than the value of the corresponding point, the corresponding point is recorded, and the number of all points that meet the conditions is determined, and the average value is calculated as the value of the corresponding point for replacement and filling. If the value of the corresponding point is a negative value, it is determined whether the values of all negative points in the N1*N1 neighborhood of the corresponding point are less than the value of the corresponding point. If not less than the value of the corresponding point, the corresponding point is not recorded. If less than the value of the corresponding point, the corresponding point is recorded, and the number of all points that meet the conditions is determined, and the average value is calculated as the value of the corresponding point for replacement and filling, so as to obtain the third hole filling speed map.
[0089] It should be noted that, in an embodiment of the present invention, the first hole map and the second hole map can be merged to obtain a third hole map. The third hole map can be obtained by merging them using an OR logical operation, and then the third hole filling speed map is updated according to the third hole map and the grayscale map. For example, assuming that the value of the point in the i-th row and j-th column of the third hole map is 1, determine whether the corresponding value in the grayscale map is 1. If it is 1, the value of the corresponding position of the second hole filling speed map is updated to 0. If it is not 1, the third hole filling speed map is not updated.
[0090] Based on the above embodiments, the embodiments of the present invention provide an image processing model training method, which can traverse the velocity map according to the velocity map according to a preset method for determining void points to obtain a void map, and then compensate the velocity map according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. This method can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0091] Based on any of the above embodiments, the present invention also provides a specific image processing method, please refer to Figure 3 , Figure 3 A flowchart of an image processing method provided by an embodiment of the present invention. An image processing method according to an embodiment of the present invention may include:
[0092] Step S301: Acquire the ultrasonic blood flow image to be processed.
[0093] Step S302: using an image processing model to compensate the ultrasonic blood flow map to obtain a compensated ultrasonic blood flow map.
[0094] In the embodiment of the present invention, the image processing model can be trained according to the image processing model training method in any of the above embodiments, and the ultrasonic blood flow map can be input into the image processing model for void point compensation to obtain a compensated ultrasonic blood flow map.
[0095] The method of using an image processing model to compensate the ultrasonic blood flow image in the embodiment of the present invention improves the efficiency and accuracy of image processing, and can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and maintain the best display result of the image during the clinical process.
[0096] An image processing model training device provided by an embodiment of the present invention is introduced below. The image processing model training device described below and the image processing model training method described above can refer to each other.
[0097] Please refer to Figure 4 , Figure 4A structural block diagram of an image processing model training device provided by an embodiment of the present invention, the device may include:
[0098] An image acquisition module 11 is used to acquire an ultrasonic blood flow map, wherein the ultrasonic blood flow map includes a grayscale map, a velocity map, a variance map, and an energy map;
[0099] A hole point module 12, configured to traverse the speed graph according to a preset method for determining a hole point, and obtain a hole graph corresponding to the speed graph;
[0100] The compensation module 13 is used to compensate the velocity map according to the cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
[0101] Based on the above embodiment, the hole point module 12 may include:
[0102] A first hole point unit, used for traversing the speed graph according to a first hole point determination method to obtain a first hole graph;
[0103] The second hole point unit traverses the speed graph according to the second hole point determination method to obtain a second hole graph.
[0104] Based on any of the above embodiments, the compensation module 13 may include:
[0105] A compensation unit is used to compensate the velocity map according to the first cavity map, the second cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
[0106] Based on any of the above embodiments, the compensation unit may include:
[0107] A first filling subunit is used to traverse the speed graph and fill the speed graph according to the second hole graph to obtain a second hole-filling speed graph;
[0108] a second filling subunit, traversing the second hole filling speed graph to fill the second hole filling speed graph according to the first hole filling speed graph, to obtain a third hole filling speed graph;
[0109] a merging subunit, configured to merge the first hole map and the second hole map to obtain a third hole map;
[0110] a value judgment subunit, used to determine whether the value of the grayscale image corresponding to the third hole image is a preset threshold;
[0111] The updating subunit is used to update the third hole filling speed map if it is the preset threshold, so as to obtain the compensated ultrasonic blood flow map.
[0112] Based on any of the above embodiments, the compensation unit may further include:
[0113] A filling subunit, configured to traverse the speed graph and fill the speed graph according to the first hole graph to obtain a first hole-filling speed graph;
[0114] A judging subunit, used to determine whether a value of the grayscale image corresponding to the first hole image is a preset threshold;
[0115] The compensation subunit is used to update the first hole filling velocity map if it is the preset threshold value, so as to obtain a compensated ultrasonic blood flow map.
[0116] Based on any of the foregoing embodiments, the first hole point unit may include:
[0117] A threshold judgment subunit, used for traversing the speed graph and determining whether the value of a non-zero point in the speed graph is less than a preset value;
[0118] The extraction subunit is used to determine that the non-zero point is a hole point if it is less than the preset value, and mark all the hole points to obtain the first hole map.
[0119] Based on any of the above embodiments, before the hole point module 12, the following may also be included:
[0120] The preprocessing module is used to preprocess the grayscale image and the speed image.
[0121] Based on any of the above embodiments, the preprocessing module may include:
[0122] A first judgment unit, used to determine whether the value of the speed graph exceeds a preset speed value;
[0123] A second judgment unit, used to determine whether the value of the variance map exceeds a preset variance value;
[0124] A third judgment unit, used to determine whether the value of the energy graph is within a preset energy value range;
[0125] The extraction unit is used to extract the corresponding speed map if the speed exceeds the preset speed value, does not exceed the preset variance value, and is within the preset energy value range.
[0126] Based on any of the above embodiments, the preprocessing module may further include:
[0127] An acquisition unit, used for acquiring an original grayscale image;
[0128] The binarization unit is used to perform binarization processing on the original grayscale image to obtain a grayscale image.
[0129] In the embodiment of the present invention, the hole point module 12 traverses the velocity map according to the preset hole point judgment method to obtain the hole map, and then uses the compensation module 13 to compensate the velocity map according to the hole map and the grayscale map to obtain the compensated ultrasonic blood flow map. This can make the image displayed on the interface better reflect the real blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0130] An image processing device provided by an embodiment of the present invention is introduced below. The image processing device described below and the image processing method described above can be referred to each other.
[0131] Please refer to Figure 5 , Figure 5 A structural block diagram of an image processing device provided by an embodiment of the present invention, the device may include:
[0132] An acquisition module 21, used for acquiring an ultrasonic blood flow image to be processed;
[0133] The image processing module 22 is used to compensate the ultrasonic blood flow map by using an image processing model to obtain the compensated ultrasonic blood flow map, wherein the image processing model is constructed according to the image processing model training method as described above.
[0134] Please refer to Figure 6 , Figure 6 A structural block diagram of an image processing device provided by an embodiment of the present invention, the device includes:
[0135] A memory 10, used for storing computer programs;
[0136] The processor 20 is used to implement the above-mentioned image processing model training method when executing the computer program.
[0137] like Figure 6 , which is a schematic diagram of the structure of an image processing device, may include: a memory 10 , a processor 20 , a communication interface 31 , an input / output interface 32 , and a communication bus 33 .
[0138] In the embodiment of the present invention, the memory 10 is used to store one or more programs, and the program may include program code, and the program code includes computer operation instructions. In the embodiment of the present application, the memory 10 may store programs for implementing the following functions:
[0139] Acquire the collected ultrasonic blood flow image, which includes a grayscale image and a velocity image;
[0140] The velocity graph is traversed according to a preset method for determining a hole point, and a hole graph corresponding to the velocity graph is obtained;
[0141] According to the cavity image and the grayscale image, the velocity image is compensated to obtain the compensated ultrasonic blood flow image.
[0142] In a possible implementation, the memory 10 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function, etc.; the data storage area may store data created during use.
[0143] In addition, the memory 10 may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include an NVRAM. The memory stores an operating system and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, wherein the operating instructions may include various operating instructions for implementing various operations. The operating system may include various system programs for implementing various basic tasks and processing hardware-based tasks.
[0144] The processor 20 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic device, a microprocessor or any conventional processor, etc. The processor 20 may call a program stored in the memory 10 .
[0145] The communication interface 31 may be an interface for connecting to other devices or systems.
[0146] The input / output interface 32 may be an interface for acquiring external input data or outputting data to the outside world.
[0147] Of course, it should be noted that Figure 6 The structure shown does not constitute a limitation on the image processing device in the embodiment of the present application. In actual applications, the image processing device may include Figure 6 More or fewer components than shown, or combinations of certain components.
[0148] In the embodiment of the present invention, according to the velocity map, the velocity map can be traversed according to a preset method for determining void points to obtain a void map, and then the velocity map can be compensated according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. This can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and enable the image to maintain the best display result during the clinical process.
[0149] The embodiment of the present invention also provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are loaded and executed by the processor, the acquired ultrasonic blood flow map is obtained, and the ultrasonic blood flow map includes a grayscale map and a velocity map; the velocity map is traversed according to a preset method for determining a void point to obtain a void map corresponding to the velocity map; and the velocity map is compensated according to the void map and the grayscale map to obtain a compensated ultrasonic blood flow map. In the embodiment of the present invention, according to the velocity map, according to the preset method for determining a void point, the velocity map is traversed to obtain a void map, and then according to the void map and the grayscale map, the velocity map is compensated to obtain a compensated ultrasonic blood flow map. This method can make the image displayed on the interface better reflect the actual blood flow state, improve the accuracy and precision of the ultrasonic blood flow image, and keep the image in the best display result during the clinical process.
[0150] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.
[0151] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0152] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
[0153] The above is a detailed introduction to an image processing model training method, device, image processing equipment and medium provided by the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.
Claims
1. A method for training an image processing model, characterized in that: include: Acquiring an ultrasonic blood flow map, wherein the ultrasonic blood flow map includes a grayscale map, a velocity map, a variance map, and an energy map; Traversing the velocity graph according to a preset method for determining a hole point, and obtaining a hole graph corresponding to the velocity graph; The velocity map is compensated according to the cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
2. The image processing model training method according to claim 1, characterized in that: The method of traversing the speed graph according to a preset method for determining a hole point to obtain a hole graph corresponding to the speed graph includes: According to the first method for determining a hole point, the velocity graph is traversed to obtain a first hole graph; According to the second method for determining a hole point, the velocity graph is traversed to obtain a second hole graph; Accordingly, compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the cavity map and the grayscale map includes: The velocity map is compensated according to the first cavity map, the second cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
3. The image processing model training method according to claim 2, characterized in that: The compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the first cavity map, the second cavity map and the grayscale map comprises: According to the second hole graph, traverse the speed graph to fill the speed graph to obtain a second hole-filling speed graph; According to the first hole graph, traverse the second hole filling speed graph to fill the second hole filling speed graph to obtain a third hole filling speed graph; Combining the first hole map and the second hole map to obtain a third hole map; Determining whether a value of the grayscale image corresponding to the third hole image is a preset threshold; If it is the preset threshold, the third hole filling speed map is updated to obtain the compensated ultrasonic blood flow map.
4. The image processing model training method according to claim 2, characterized in that: The compensating the velocity map to obtain a compensated ultrasonic blood flow map according to the first cavity map, the second cavity map and the grayscale map further includes: According to the first hole map, traverse the speed map to fill the speed map to obtain a first hole-filling speed map; Determining whether a value of the grayscale image corresponding to the first hole image is a preset threshold; If it is the preset threshold, the first hole filling speed map is updated to obtain a compensated ultrasonic blood flow map.
5. The image processing model training method according to claim 2, characterized in that: The step of traversing the velocity graph according to the first method for determining a hole point to obtain a first hole graph includes: Traversing the speed graph, determining whether the value of a non-zero point in the speed graph is less than a preset value; If it is less than the preset value, the non-zero point is determined to be a hole point, and all the hole points are marked to obtain the first hole map.
6. The image processing model training method according to claim 1, characterized in that: Before traversing the speed graph according to the preset method for determining the hole point, the method further includes: determining whether a value of the speed graph exceeds a preset speed value; Determining whether the value of the variance map exceeds a preset variance value; Determining whether the value of the energy graph is within a preset energy value range; If it exceeds the preset speed value, does not exceed the preset variance value, and is within the preset energy value range, then the corresponding speed map is extracted.
7. An image processing method, characterized in that: include: obtaining an ultrasound blood flow image to be processed; The ultrasonic blood flow map is compensated by using an image processing model to obtain the compensated ultrasonic blood flow map, wherein the image processing model is constructed according to the image processing model training method according to any one of claims 1 to 6.
8. An image processing model training device, characterized in that: include: An image acquisition module, used to acquire an ultrasonic blood flow map, wherein the ultrasonic blood flow map includes a grayscale map, a velocity map, a variance map, and an energy map; A hole point module, used for traversing the speed graph according to a preset method for determining a hole point, and obtaining a hole graph corresponding to the speed graph; A compensation module is used to compensate the velocity map according to the cavity map and the grayscale map to obtain a compensated ultrasonic blood flow map.
9. An image processing device, characterized in that: include: Memory for storing computer programs; A processor, used to implement the image processing model training method as described in any one of claims 1 to 6 when executing the computer program.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are loaded and executed by the processor, the steps of the image processing model training method as described in any one of claims 1 to 6 are implemented.
Citation Information
Patent Citations
Method and system for detecting sensitivity of small blood flow signals based on ultrasonic imaging
CN104127207A
Ultrasonic color blood flow imaging boundary processing method and system
CN105708495A
Ultrasonic blood flow image processing method
CN106127707A
Color Doppler blood flow image denoising method based on improved variance
CN109360173A
