Ultrasonic image processing method and device, electronic equipment and readable storage medium

Abstract

The invention discloses an ultrasonic image processing method. The method comprises the steps that a time-phase section recognition model is acquired; the time phase section recognition model comprises a left ventricle segmentation unit; inputting the cardiac ultrasound video into a time-phase section recognition model to trigger the time-phase section recognition model to obtain a heart standard section corresponding to at least one time phase based on the detected image features and the left ventricle segmentation result; processing the heart standard section under each time phase by using the image segmentation model to obtain a segmentation result; the segmentation result comprises point-line features of each target tissue in the heart standard section; and determining the detection result of the heart standard section in each time phase in parallel according to the point-line characteristics of each target tissue. By applying the technical scheme provided by the invention, the heart standard section can be measured more accurately and efficiently. The invention also discloses an ultrasonic image processing device, electronic equipment and a computer readable storage medium, which also have the above technical effects.

Description

Technical Field

[0001] This application relates to the field of ultrasound technology, and particularly to an ultrasound image processing method, apparatus, electronic device, and computer-readable storage medium. Background Art

[0002] During the process of human heart scanning, it is usually necessary to obtain specific standard sections and perform corresponding measurements. In traditional technologies, doctors mainly obtain standard sections manually, manually select sections at different time phases, and finally perform manual measurements of corresponding time-phase items. Obviously, this implementation method is more dependent on doctors' experience, with low accuracy and low work efficiency.

[0003] Therefore, how to measure the heart standard sections more accurately and efficiently is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide an ultrasound image processing method, which can measure the heart standard sections more accurately and efficiently; another purpose of this application is to provide an ultrasound image processing apparatus, electronic device, and computer-readable storage medium, all of which have the above beneficial effects.

[0005] In a first aspect, this application provides an ultrasound image processing method, including:

[0006] Obtain a time-phase section recognition model; the time-phase section recognition model includes a left ventricle segmentation unit;

[0007] Input a cardiac ultrasound video into the time-phase section recognition model to trigger the time-phase section recognition model to obtain at least one heart standard section corresponding to each time phase based on the detected image features and the left ventricle segmentation result;

[0008] Process each heart standard section at each time phase by using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the heart standard section;

[0009] Parallelly determine the detection results of the heart standard section at each time phase according to the point-line features of each target tissue.

[0010] Optionally, the at least one time phase includes the end-diastolic phase, the mid-systolic phase, and the end-systolic phase of the heart;

[0011] For the heart standard sections in the end-diastolic phase and the end-systolic phase of the heart, the point-line features of each target tissue in the heart standard section include the contour information and the first key point information of the first target tissue;

[0012] For the standard cardiac section in mid-systole, the dot-line features of each target tissue in the standard cardiac section include the edge arc information and the second key point information of the second target tissue.

[0013] Optionally, the contour information of the first target tissue includes the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour; the first key point information includes the mitral valve point.

[0014] Optionally, for the standard cardiac section in end-diastole, determining the detection results of the standard cardiac section at each time phase according to the dot-line features of each target tissue includes:

[0015] Determining the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall of the standard cardiac section in end-diastole according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the left ventricular posterior wall contour, and the mitral valve point.

[0016] Optionally, determining the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall of the standard cardiac section in end-diastole according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the left ventricular posterior wall contour, and the mitral valve point includes:

[0017] Calculating the slope of the measurement line according to the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour;

[0018] Drawing a measurement line through the mitral valve point according to the slope of the measurement line, and the measurement line intersects with the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour;

[0019] Determining adjacent contour measurement line segments in the measurement line, and the adjacent contour measurement line segments include the measurement line segment between the right ventricular contour and the interventricular septum contour, the measurement line segment between the interventricular septum contour and the left ventricular contour, and the measurement line segment between the left ventricular contour and the left ventricular posterior wall contour;

[0020] Determining the midpoints of each adjacent contour measurement line segment, and using each midpoint to correct the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour to obtain the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour;

[0021] Determining the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour.

[0022] Optionally, calculating the slope of the measurement line based on the ventricular septum profile, the left ventricular profile, and the left ventricular posterior wall profile includes:

[0023] Determine the minimum area circumscribed rectangles of the ventricular septum profile, the left ventricular profile, and the left ventricular posterior wall profile, and determine the slopes of the short sides of each of the minimum area circumscribed rectangles;

[0024] Take the mean of the slopes of the short sides of each of the minimum area circumscribed rectangles as the slope of the measurement line.

[0025] Optionally, using each of the midpoints to correct the right ventricular profile, the ventricular septum profile, the left ventricular profile, and the left ventricular posterior wall profile to obtain the corrected right ventricular profile, the corrected ventricular septum profile, the corrected left ventricular profile, and the corrected left ventricular posterior wall profile includes:

[0026] For each of the midpoints, determine the two line segment endpoints of the adjacent contour measurement segments corresponding to the midpoint;

[0027] Take the region formed by the contour edges where the two line segment endpoints are located as the gradient calculation region, and calculate the image gradient at each point position within the gradient calculation region;

[0028] Take the point position corresponding to the maximum image gradient as the new center point, and extend the contour edges where the two line segment endpoints are located to the new center point to obtain the corrected tissue contour; the corrected tissue contour includes the corrected right ventricular profile, the corrected ventricular septum profile, the corrected left ventricular profile, and the corrected left ventricular posterior wall profile.

[0029] Optionally, determining the anteroposterior diameter of the right ventricle, the thickness of the ventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall based on the corrected right ventricular profile, the corrected ventricular septum profile, the corrected left ventricular profile, and the corrected left ventricular posterior wall profile includes:

[0030] Take the length of the intersecting line segment between the corrected right ventricular profile and the measurement line as the anteroposterior diameter of the right ventricle;

[0031] Take the length of the intersecting line segment between the corrected ventricular septum profile and the measurement line as the thickness of the ventricular septum;

[0032] Take the length of the intersecting line segment between the corrected left ventricular profile and the measurement line as the anteroposterior diameter of the left ventricle;

[0033] Take the length of the intersecting line segment between the corrected left ventricular posterior wall profile and the measurement line as the thickness of the left ventricular posterior wall.

[0034] Optionally, the edge arc information of the second target tissue includes the aortic and left ventricular anterior wall arcs, and the posterior aortic wall arc; the second key point information includes the first aortic root and the second aortic root.

[0035] Optionally, for the standard cardiac section at the mid-systole of the heart, the parallel determination of the detection results of the standard cardiac section at each time phase according to the point-line characteristics of each target tissue includes:

[0036] Determine the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section at the mid-systole of the heart according to the aortic and left ventricular anterior wall arcs, the posterior aortic wall arc, the first aortic root, and the second aortic root.

[0037] Optionally, the determination of the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section at the mid-systole of the heart according to the aortic and left ventricular anterior wall arcs, the posterior aortic wall arc, the first aortic root, and the second aortic root includes:

[0038] Generate a corresponding skeleton diagram based on the aortic and left ventricular anterior wall arcs and the posterior aortic wall arc;

[0039] In the skeleton diagram, draw a connection line through the first aortic root and the second aortic root, and determine the first intersection point of the connection line and the aortic and left ventricular anterior wall arc, and the second intersection point with the posterior aortic wall arc;

[0040] Correct the first intersection point and the second intersection point respectively to obtain the corrected first intersection point and the corrected second intersection point;

[0041] Take the connection line between the corrected first intersection point and the corrected second intersection point as the aortic annulus diameter measurement line segment, and take the length of the aortic annulus diameter measurement line segment as the aortic annulus diameter;

[0042] Determine the midpoint of the aortic annulus diameter measurement line segment;

[0043] Translate the midpoint of the aortic annulus diameter measurement line segment by a preset length in a preset direction to obtain a translation point;

[0044] Draw a left ventricular outflow tract diameter line through the translation point according to the slope of the aortic annulus diameter measurement line segment, and determine the third intersection point of the left ventricular outflow tract diameter line and the aortic and left ventricular anterior wall arc, and the fourth intersection point with the posterior aortic wall arc;

[0045] Correct the third intersection point and the fourth intersection point respectively to obtain the corrected third intersection point and the corrected fourth intersection point;

[0046] Take the length of the line connecting the corrected third intersection point and the corrected fourth intersection point as the inner diameter of the left ventricular outflow tract.

[0047] Optionally, the respectively correcting the first intersection point and the second intersection point to obtain a corrected first intersection point and a corrected second intersection point includes:

[0048] Calculate the image gradient of each point position on the aortic and left ventricular anterior wall arcs, and take the point position corresponding to the maximum image gradient on the aortic and left ventricular anterior wall arcs as the corrected first intersection point;

[0049] Calculate the image gradient of each point position on the posterior wall arc of the aorta, and take the point position corresponding to the maximum image gradient on the posterior wall arc of the aorta as the corrected second intersection point.

[0050] Optionally, for the standard cardiac section at the end of cardiac systole, the parallel determination of the detection results of the standard cardiac section at each time phase according to the point-line characteristics of each target tissue includes:

[0051] Determine the interventricular septum thickness, left ventricular anteroposterior diameter, and left ventricular posterior wall thickness of the standard cardiac section at the end of cardiac systole according to the right ventricular contour, interventricular septum contour, left ventricular contour, left ventricular posterior wall contour, and mitral valve point.

[0052] In a second aspect, the present application also discloses an ultrasonic image processing device, including:

[0053] An acquisition module for acquiring a time-phase section recognition model; the time-phase section recognition model includes a left ventricular segmentation unit;

[0054] An identification module for inputting a cardiac ultrasound video into the time-phase section recognition model to trigger the time-phase section recognition model to obtain at least one standard cardiac section corresponding to each time phase based on the detected image features and left ventricular segmentation results;

[0055] A processing module for processing each standard cardiac section at each time phase by using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line characteristics of each target tissue in the standard cardiac section;

[0056] A determination module for parallelly determining the detection results of the standard cardiac section at each time phase according to the point-line characteristics of each target tissue.

[0057] In a third aspect, the present application also discloses an electronic device, including:

[0058] A memory for storing a computer program;

[0059] A processor, configured to implement the steps of any of the ultrasonic image processing methods described above when executing the computer program.

[0060] In a fourth aspect, the present application also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the ultrasonic image processing methods described above are implemented.

[0061] The present application provides an ultrasonic image processing method, including: obtaining a phase section recognition model; the phase section recognition model includes a left ventricle segmentation unit; inputting a cardiac ultrasound video into the phase section recognition model to trigger the phase section recognition model to obtain at least one cardiac standard section corresponding to each phase based on the detected image features and the left ventricle segmentation result; using an image segmentation model to process the cardiac standard section at each phase to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the cardiac standard section; determining the detection result of the cardiac standard section at each phase in parallel according to the point-line features of each target tissue.

[0062] Applying the technical solution provided by the present application, a phase section recognition model including a left ventricle segmentation unit is pre-created. The phase section recognition model can be used to automatically recognize the cardiac standard sections at specific phases, so as to obtain the point-line features of each target tissue in the cardiac standard sections at different phases. Finally, the detection result of the cardiac standard section can be determined according to the point-line features of each target tissue, realizing the automatic measurement of the cardiac standard section, getting rid of the dependence on manual experience, effectively improving the accuracy of the detection result of the cardiac standard section, and improving the measurement efficiency of the cardiac standard section.

[0063] The ultrasonic image processing device, electronic device, and computer-readable storage medium provided by the present application also have the above technical effects, which will not be repeated here. Description of the Drawings

[0064] In order to more clearly illustrate the technical solutions in the prior art and the embodiments of the present application, the drawings required for description in the prior art and the embodiments of the present application will be briefly introduced below. Of course, the following drawings described for the embodiments of the present application are only a part of the embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts, and the other drawings obtained also belong to the protection scope of the present application.

[0065] Figure 1 It is a schematic flowchart of an ultrasonic image processing method provided by the present application;

[0066] Figure 2Schematic diagram of segmentation and annotation of parasternal left ventricular long-axis section at end-diastole provided by this application;

[0067] Figure 3 Schematic diagram of segmentation and annotation of parasternal left ventricular long-axis section at mid-systole provided by this application;

[0068] Figure 4 Schematic diagram of segmentation and annotation of parasternal left ventricular long-axis section at end-systole provided by this application;

[0069] Figure 5 Schematic diagram of the structure principle of an ultrasonic image processing device provided by this application;

[0070] Figure 6 Schematic diagram of the structure of an electronic device provided by this application. Detailed implementation manners

[0071] The core of this application is to provide an ultrasonic image processing method, which can measure the parasternal left ventricular long-axis section more accurately and efficiently; another core of this application is to provide an ultrasonic image processing device, an electronic device and a computer-readable storage medium, all of which have the above beneficial effects.

[0072] In order to describe the technical solutions in the embodiments of this application more clearly and completely, the following will introduce the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0073] The embodiments of this application provide an ultrasonic image processing method, which can be applied to electronic devices such as ultrasonic devices.

[0074] Please refer to Figure 1 , Figure 1 Flow chart of an ultrasonic image processing method provided by this application. This ultrasonic image processing method may include the following S101 to S104.

[0075] S101: Obtain a phase-section recognition model; the phase-section recognition model includes a left ventricle segmentation unit.

[0076] This step aims to obtain a phase-plane recognition model, which has both phase recognition and plane recognition functions and can be used to recognize cardiac ultrasound videos to obtain at least one cardiac standard plane corresponding to each phase. Therefore, this phase-plane recognition model may incorporate a target detection unit and a left ventricular segmentation unit. Among them, the target detection unit is used to implement the plane recognition function, that is, to identify and determine the cardiac standard plane, and the left ventricular segmentation unit is used to implement the phase recognition function, that is, to identify and determine each different phase (corresponding to different cardiac states, such as end-diastolic phase, mid-systolic phase, end-systolic phase, etc.).

[0077] Among them, the image recognition model is obtained by pre-training an initial recognition model using sample data and can be pre-stored in the corresponding storage space and directly called when in use. Among them, the sample data may include cardiac standard plane samples, non-cardiac standard plane samples, and their respective corresponding label information. During the training process, ultrasound video samples can be collected in advance, and each ultrasound image sample can be obtained through video frame segmentation. Then, each ultrasound image sample is labeled with the cardiac standard plane and phase using annotation software. Finally, the labeled ultrasound image sequence is input into the initial recognition model for iterative training to obtain a high-precision phase-plane recognition model.

[0078] S102: Input the cardiac ultrasound video into the phase-plane recognition model to trigger the phase-plane recognition model to obtain at least one cardiac standard plane corresponding to each phase based on the detected image features and the left ventricular segmentation result.

[0079] This step aims to realize the recognition of cardiac standard planes in different phases based on the phase-plane recognition model and obtain at least one cardiac standard plane corresponding to each phase. Specifically, for the obtained cardiac ultrasound video, it can be directly input into the phase-plane recognition model for processing. As mentioned above, the phase-plane recognition model may incorporate a target detection unit and a left ventricular segmentation unit. Therefore, the target detection unit is used to extract image features, and the left ventricular segmentation unit is used to perform left ventricular segmentation. Thus, the phase-plane recognition model can further determine the cardiac standard planes in different phases using the detected image features and the left ventricular segmentation result. Specifically, the current phase of the heart can be determined based on the left ventricular features segmented by the left ventricular segmentation unit, and it can be analyzed whether the current plane is the cardiac standard plane in the current phase based on the image features extracted by the target detection unit.

[0080] In a possible implementation, the above at least one phase may include end-diastolic phase, mid-systolic phase, end-systolic phase of the heart; the cardiac standard plane may specifically be the parasternal long-axis view of the left ventricle of the heart.

[0081] Among them, the cardiac ultrasound video can be an ultrasound video pre-acquired by an ultrasound device or an ultrasound video detected in real-time ultrasound mode. This application does not limit this.

[0082] Thus, the automatic acquisition of cardiac standard sections at different time phases can be achieved without manual operation, effectively improving work efficiency.

[0083] S103: Process the cardiac standard sections at each time phase using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the cardiac standard section.

[0084] This step aims to achieve the segmentation process of the cardiac standard section. Specifically, an image segmentation model can be created in advance. After the cardiac standard section is identified based on the cardiac ultrasound video, the cardiac standard section can be input into the image segmentation model for processing, and the output of the model is the point-line features of each target tissue in the heart.

[0085] It can be understood that for the cardiac standard sections at different time phases, the point-line features of the target tissues obtained may vary. In a possible implementation, for the cardiac standard sections at the end-diastole and end-systole of the heart, the point-line features of each target tissue in the cardiac standard section include the contour information of the first target tissue and the first key point information; for the cardiac standard section at the mid-systole of the heart, the point-line features of each target tissue in the cardiac standard section include the edge arc information of the second target tissue and the second key point information.

[0086] Furthermore, when the cardiac standard section is the parasternal long-axis section of the left ventricle of the heart, the contour information of the first target tissue may include the right ventricle contour, the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour, and the first key point information may include the mitral valve point; the edge arc information of the second target tissue may include the aortic and anterior left ventricular wall arcs and the aortic posterior wall arc, and the second key point information may include the first aortic valve root and the second aortic valve root. Among them, the mitral valve point may be the mitral valve tip.

[0087] Among them, the image segmentation model is also pre-trained using sample data on an initial segmentation model, which can be pre-stored in the corresponding storage space and directly called when in use. Among them, the sample data can include cardiac standard section samples at different time phases and the corresponding label information for each sample. Taking the parasternal long-axis section of the left ventricle of the heart as an example, please refer to Figures 2 to 4 , Figure 2 is a schematic diagram of the segmentation and annotation of the parasternal long-axis section of the left ventricle of the heart at the end-diastole of the heart provided by this application, Figure 3 is a schematic diagram of the segmentation and annotation of the parasternal long-axis section of the left ventricle of the heart at the mid-systole of the heart provided by this application, Figure 4This is a schematic diagram of the segmentation and annotation of the parasternal long-axis view of the left ventricle at the end-systole provided by this application. Among them, the label information of the parasternal long-axis view sample of the left ventricle at the end-diastole of the heart can include the right ventricular contour, the interventricular septum contour, the left ventricular contour, the left ventricular posterior wall contour, the mitral valve tip, and the corresponding measurement lines; the label information of the parasternal long-axis view sample of the left ventricle at the mid-systole of the heart can include the aortic and left ventricular anterior wall arcs, the aortic posterior wall arc, aortic valve root 1 (the first aortic valve root), aortic valve root 2 (the second aortic valve root), the aortic annulus inner diameter, and the left ventricular outflow tract inner diameter; the label information of the parasternal long-axis view sample of the left ventricle at the end-systole of the heart can include the right ventricular contour, the interventricular septum contour, the left ventricular contour, the left ventricular posterior wall contour, the mitral valve tip, and the corresponding measurement lines. During the training process, the above-mentioned labeled sample data can be input into the initial recognition segmentation model for iterative training to obtain a high-precision image segmentation model.

[0088] Among them, the initial segmentation model can be a model innovatively improved based on the human pose key point detection model. On the one hand, the label content is improved, and the key point Gaussian heat map is changed to a contour and arc mask; on the other hand, the 17 channels of the original model are changed to 9 channels. Among them, the first four channels are changed to semantic segmentation channels for learning the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour. The fifth to seventh channels are changed to key point detection channels for learning the mitral valve, aortic valve root 1, and aortic valve root 2; the eighth and ninth channels are changed to arc detection channels for learning the aortic and left ventricular anterior wall arcs and the aortic posterior wall arc, so as to complete the multi-task measurement of simultaneously outputting contours, points, and arcs.

[0089] Thus, the automatic segmentation of the standard cardiac section at a specific time phase can be realized without manual operation, effectively improving the work efficiency.

[0090] S104: Parallelly determine the detection results of the standard cardiac section at each time phase according to the point-line features of each target tissue.

[0091] The purpose of this step is to determine the detection results of the standard cardiac section according to the point-line features of each target tissue. As mentioned above, for the standard cardiac sections at different time phases, the point-line features of the target tissues obtained may be different. Correspondingly, the detection results of the standard cardiac sections obtained based on the point-line features of different target tissues may also be different.

[0092] For example, for the standard cardiac section at the end of diastole, the above method of concurrently determining the detection results of the standard cardiac section at each time phase based on the point-line features of each target tissue may include: determining the anteroposterior diameter of the right ventricle (the anteroposterior diameter of the right ventricle is also called the internal diameter of the right ventricle), the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle (the anteroposterior diameter of the left ventricle is also called the internal diameter of the left ventricle), and the thickness of the posterior wall of the left ventricle of the standard cardiac section at the end of diastole according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the posterior wall contour of the left ventricle, and the mitral valve point. For the standard cardiac section at the mid-systole, the above method of concurrently determining the detection results of the standard cardiac section at each time phase based on the point-line features of each target tissue may include: determining the internal diameter of the aortic valve annulus and the internal diameter of the left ventricular outflow tract of the standard cardiac section at the mid-systole according to the aortic arch and the anterior wall arc of the left ventricle, the posterior wall arc of the aorta, the first aortic valve root, and the second aortic valve root. For the standard cardiac section at the end of systole, the above method of concurrently determining the detection results of the standard cardiac section at each time phase based on the point-line features of each target tissue may include: determining the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle of the standard cardiac section at the end of systole according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the posterior wall contour of the left ventricle, and the mitral valve point.

[0093] In a first aspect, the above method of determining the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle of the standard cardiac section at the end of diastole according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the posterior wall contour of the left ventricle, and the mitral valve point may include:

[0094] S10: Calculate the slope of the measurement line according to the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle.

[0095] In one implementation, the above S10 may specifically include:

[0096] S1010: Determine the minimum area circumscribed rectangle of the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle, and determine the slope of the short side of each minimum area circumscribed rectangle;

[0097] S1011: Use the mean value of the slopes of the short sides of each minimum area circumscribed rectangle as the slope of the measurement line.

[0098] Among them, the slope of the short side of the minimum area circumscribed rectangle can be calculated based on the rotation angle of the minimum area circumscribed rectangle relative to the horizontal direction. In addition, before determining the minimum area circumscribed rectangle, the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle output by the image segmentation model can be binarized to effectively simplify the image and facilitate subsequent processing.

[0099] S11: Draw a measurement line through the mitral valve point according to the slope of the measurement line, and the measurement line intersects the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle.

[0100] S12: Determine adjacent contour measurement line segments in the measurement line. The adjacent contour measurement line segments include the measurement line segments between the right ventricular contour and the interventricular septum contour, between the interventricular septum contour and the left ventricular contour, and between the left ventricular contour and the left ventricular posterior wall contour.

[0101] Among them, the measurement line segment between the right ventricular contour and the interventricular septum contour is: the measurement line segment where the measurement line intersects the right ventricular contour, plus, the measurement line segment where the measurement line intersects the interventricular septum contour. The measurement line segment between the interventricular septum contour and the left ventricular contour is: the measurement line segment where the measurement line intersects the interventricular septum contour, plus, the measurement line segment where the measurement line intersects the left ventricular contour. The measurement line segment between the left ventricular contour and the left ventricular posterior wall contour is: the measurement line segment where the measurement line intersects the left ventricular contour, plus, the measurement line segment where the measurement line intersects the left ventricular posterior wall contour.

[0102] S13: Determine the midpoints of each adjacent contour measurement line segment, and use each midpoint to correct the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour to obtain the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour.

[0103] In one implementation, the above S13 may specifically include:

[0104] S1310: For each midpoint, determine the two line segment endpoints of the adjacent contour measurement line segment corresponding to the midpoint;

[0105] S1311: Take the region formed by the contour edges where the two line segment endpoints are located as the gradient calculation region, and calculate the image gradients at the positions of each point within the gradient calculation region;

[0106] S1312: Take the point position corresponding to the maximum image gradient as the new center point, and extend the contour edges where the two line segment endpoints are located to the new center point to obtain the corrected tissue contour; the corrected tissue contour includes the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour.

[0107] Obviously, for the measurement line segment between the right ventricular contour and the interventricular septum contour, the gradient calculation region corresponding to its midpoint is the region between the lower edge line of the right ventricular contour and the upper edge line of the interventricular septum contour; for the measurement line segment between the interventricular septum contour and the left ventricular contour, the gradient calculation region corresponding to its midpoint is the region between the lower edge line of the interventricular septum contour and the upper edge line of the left ventricular contour; for the measurement line segment between the left ventricular contour and the left ventricular posterior wall contour, the gradient calculation region corresponding to its midpoint is the region between the lower edge line of the left ventricular contour and the upper edge line of the left ventricular posterior wall contour.

[0108] It should be noted that the expansion of the contour edge specifically refers to translating the contour edge upward or downward to a new center point so that the new center point falls on the translated edge contour.

[0109] Among them, the positions of the points within the gradient calculation region can specifically be the positions of each pixel point, and the image gradient calculation process can be implemented by calculating with an edge extraction matrix.

[0110] S14: Determine the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle.

[0111] In one implementation, the above S14 can specifically include:

[0112] S1410: Take the length of the intersecting line segment between the corrected right ventricular contour and the measurement line as the anteroposterior diameter of the right ventricle;

[0113] S1411: Take the length of the intersecting line segment between the corrected interventricular septum contour and the measurement line as the thickness of the interventricular septum;

[0114] S1412: Take the length of the intersecting line segment between the corrected left ventricular contour and the measurement line as the anteroposterior diameter of the left ventricle;

[0115] S1413: Take the length of the intersecting line segment between the corrected posterior wall contour of the left ventricle and the measurement line as the thickness of the posterior wall of the left ventricle.

[0116] In the second aspect, the above determination of the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section in the mid-systole of the heart according to the aortic and left ventricular anterior wall arcs, the aortic posterior wall arc, the first aortic root, and the second aortic root can include:

[0117] S20: Generate a corresponding skeleton diagram based on the aortic and left ventricular anterior wall arcs and the aortic posterior wall arc.

[0118] The corresponding arcs can be arranged according to the positional relationship between the aortic and left ventricular anterior wall arcs and the aortic posterior wall arc to obtain the corresponding skeleton diagram.

[0119] Furthermore, the skeleton diagram can be refined to generate a corresponding refined skeleton diagram. Then operations are performed based on the refined skeleton diagram. Among them, the image refinement operation belongs to a morphological processing algorithm, which is a method of representing a planar region as a graphical structure shape.

[0120] In addition, before performing the refinement operation, the aortic and left ventricular anterior wall arcs and the aortic posterior wall arc can be binarized to effectively simplify the image and facilitate subsequent processing.

[0121] S21: In the skeleton diagram, draw a connection line through the first aortic root and the second aortic root, and determine the first intersection point of the connection line with the arc of the aorta and the anterior wall of the left ventricle, and the second intersection point with the arc of the posterior wall of the aorta.

[0122] S22: Respectively correct the first intersection point and the second intersection point to obtain the corrected first intersection point and the corrected second intersection point.

[0123] In one implementation manner, the above S22 may specifically include:

[0124] S2210: Calculate the image gradient of each point position on the arc of the aorta and the anterior wall of the left ventricle, and use the point position corresponding to the maximum image gradient on the arc of the aorta and the anterior wall of the left ventricle as the corrected first intersection point;

[0125] S2211: Calculate the image gradient of each point position on the arc of the posterior wall of the aorta, and use the point position corresponding to the maximum image gradient on the arc of the posterior wall of the aorta as the corrected second intersection point.

[0126] Similarly, the specific positions of each point on the arc of the aorta and the anterior wall of the left ventricle and on the arc of the posterior wall of the aorta may be the positions of each pixel point, and the image gradient calculation process may be implemented by using an edge extraction matrix calculation.

[0127] S23: Use the connection line between the corrected first intersection point and the corrected second intersection point as the measurement line segment of the aortic annulus diameter, and use the length of the measurement line segment of the aortic annulus diameter as the aortic annulus diameter.

[0128] S24: Determine the midpoint of the measurement line segment of the aortic annulus diameter.

[0129] S25: Translate the midpoint of the measurement line segment of the aortic annulus diameter by a preset length in a preset direction to obtain a translated point.

[0130] As Figure 3 shown, specifically, it may be to horizontally translate the midpoint of the measurement line segment of the aortic annulus diameter to the left by a certain length. Of course, the specific value of the preset length does not affect the implementation of this technical solution and can be set by those skilled in the art according to the actual situation.

[0131] S26: Draw a straight line of the left ventricular outflow tract diameter through the translated point according to the slope of the measurement line segment of the aortic annulus diameter, and determine the third intersection point of the straight line of the left ventricular outflow tract diameter with the arc of the aorta and the anterior wall of the left ventricle, and the fourth intersection point with the arc of the posterior wall of the aorta.

[0132] S27: Respectively correct the third intersection point and the fourth intersection point to obtain the corrected third intersection point and the corrected fourth intersection point.

[0133] Corresponding to the above S22, this S27 may specifically include:

[0134] S2710: Calculate the image gradients at the positions of points on the arcs of the aorta and the anterior wall of the left ventricle, and take the position of the point corresponding to the maximum image gradient on the arcs of the aorta and the anterior wall of the left ventricle as the corrected third intersection point;

[0135] S2711: Calculate the image gradients at the positions of points on the arc of the posterior wall of the aorta, and take the position of the point corresponding to the maximum image gradient on the arc of the posterior wall of the aorta as the corrected fourth intersection point.

[0136] S28: Take the length of the line segment between the corrected third intersection point and the corrected fourth intersection point as the internal diameter of the left ventricular outflow tract.

[0137] In a third aspect, similar to the first aspect above, the determination of the interventricular septum thickness, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle of the standard cardiac section at the end of cardiac systole based on the right ventricular contour, the interventricular septum contour, the left ventricular contour, the posterior wall contour of the left ventricle, and the mitral valve point may include:

[0138] S30: Calculate the slope of the measurement line according to the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle.

[0139] S31: Draw a measurement line through the mitral valve point according to the slope of the measurement line, and the measurement line intersects the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle.

[0140] S32: Determine adjacent contour measurement line segments in the measurement line. The adjacent contour measurement line segments include the measurement line segment between the right ventricular contour and the interventricular septum contour, the measurement line segment between the interventricular septum contour and the left ventricular contour, and the measurement line segment between the left ventricular contour and the posterior wall contour of the left ventricle.

[0141] S33: Determine the midpoints of each adjacent contour measurement line segment, and use each midpoint to correct the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle to obtain the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle.

[0142] S34: Determine the interventricular septum thickness, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle.

[0143] It can be seen by comparison that compared with the detection results of the standard cardiac section at the end of cardiac diastole, the detection results of the standard cardiac section at the end of cardiac systole do not need to include the anteroposterior diameter of the right ventricle.

[0144] In addition, the detection results of the above-mentioned cardiac standard sections at different time phases can be output to the ultrasound interface for visual display. Further, standard data for each inspection item can be pre-configured in the ultrasound system, so as to determine whether the detection results are qualified by comparing the detection results of each inspection item with the standard data.

[0145] Taking the parasternal long-axis standard section of the heart as an example, for the parasternal long-axis standard section at the end of diastole of the heart, the standard data for each inspection item are as follows: the anteroposterior diameter of the right ventricle is 14 - 25 mm, the thickness of the interventricular septum is 8 - 11 mm, the anteroposterior diameter of the left ventricle is 37 - 55 mm, and the thickness of the posterior wall of the left ventricle is 8 - 11 mm. For the parasternal long-axis standard section at the mid-systole of the heart, the standard data for each inspection item are as follows: the inner diameter of the aortic valve annulus is 16 - 26 mm, and the inner diameter of the left ventricular outflow tract is 21 - 33 mm for males and 23 - 32 mm for females. For the parasternal long-axis standard section at the end of systole of the heart, the standard data for each inspection item are as follows: the thickness of the interventricular septum is 14 - 16 mm, the anteroposterior diameter of the left ventricle is 23 - 36 mm, and the thickness of the posterior wall of the left ventricle is 14 - 16 mm. Thus, when the ultrasound system detects an unqualified inspection item, an abnormal prompt can also be output on the ultrasound interface to remind the doctor to make corresponding handling.

[0146] It can be seen that the ultrasound image processing method provided by the embodiments of the present application pre-creates a time-phase section recognition model including a left ventricle segmentation unit. The time-phase section recognition model can be used to automatically recognize the cardiac standard sections at specific time phases, so as to obtain the point-line features of each target tissue in the cardiac standard sections at different time phases. Finally, the detection results of the cardiac standard sections can be determined according to the point-line features of each target tissue, realizing the automatic measurement of the cardiac standard sections, getting rid of the dependence on manual experience, effectively improving the accuracy of the detection results of the cardiac standard sections, and improving the measurement efficiency of the cardiac standard sections.

[0147] The embodiments of the present application provide an ultrasound image processing device.

[0148] Please refer to Figure 5 , Figure 5 which is a schematic structural diagram of an ultrasound image processing device provided by the present application. The ultrasound image processing device may include:

[0149] An acquisition module 1, configured to acquire a time-phase section recognition model; the time-phase section recognition model includes a left ventricle segmentation unit;

[0150] An identification module 2, configured to input a cardiac ultrasound video into the time-phase section recognition model, so as to trigger the time-phase section recognition model to obtain at least one corresponding cardiac standard section at each time phase based on the detected image features and the left ventricle segmentation result;

[0151] The processing module 3 is used to process the standard cardiac sections at each phase by using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the standard cardiac section.

[0152] The determination module 4 is used to determine the detection results of the standard cardiac section at each phase in parallel according to the point-line features of each target tissue.

[0153] It can be seen that the ultrasonic image processing device provided by the embodiment of the present application pre-creates a phase-section recognition model including a left ventricle segmentation unit. The phase-section recognition model can be used to automatically recognize the standard cardiac sections at specific phases, so as to obtain the point-line features of each target tissue in the standard cardiac sections at different phases. Finally, the detection results of the standard cardiac section can be determined according to the point-line features of each target tissue, realizing the automatic measurement of the standard cardiac section, getting rid of the dependence on manual experience, effectively improving the accuracy of the detection results of the standard cardiac section, and improving the measurement efficiency of the standard cardiac section.

[0154] In an embodiment of the present application, the at least one phase may include the end-diastolic phase of the heart, the mid-systolic phase of the heart, and the end-systolic phase of the heart.

[0155] For the standard cardiac sections at the end-diastolic phase and the end-systolic phase of the heart, the point-line features of each target tissue in the standard cardiac section may include the contour information of the first target tissue and the first key point information.

[0156] For the standard cardiac section at the mid-systolic phase of the heart, the point-line features of each target tissue in the standard cardiac section may include the edge arc information of the second target tissue and the second key point information.

[0157] In an embodiment of the present application, the contour information of the first target tissue may include the right ventricle contour, the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour; the first key point information may include the mitral valve point.

[0158] In an embodiment of the present application, for the standard cardiac section at the end-diastolic phase of the heart, the above-mentioned determination module 4 may specifically be used to determine the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall of the standard cardiac section at the end-diastolic phase of the heart according to the right ventricle contour, the interventricular septum contour, the left ventricle contour, the left ventricular posterior wall contour, and the mitral valve point.

[0159] In an embodiment of the present application, the above-mentioned determination module 4 may include:

[0160] The calculation unit is used to calculate the slope of the measurement line according to the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour.

[0161] The first drawing unit is configured to draw a measurement line through the mitral valve point according to the slope of the measurement line, and the measurement line intersects with the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour;

[0162] The first determination unit is configured to determine adjacent contour measurement line segments in the measurement line. The adjacent contour measurement line segments include the measurement line segment between the right ventricular contour and the interventricular septum contour, the measurement line segment between the interventricular septum contour and the left ventricular contour, and the measurement line segment between the left ventricular contour and the left ventricular posterior wall contour;

[0163] The first correction unit is configured to determine the midpoints of the respective adjacent contour measurement line segments, and use the midpoints to correct the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour to obtain the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour;

[0164] The second determination unit is configured to determine the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the left ventricular posterior wall according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour.

[0165] In an embodiment of the present application, the above calculation unit may specifically be configured to determine the minimum area circumscribed rectangles of the interventricular septum contour, the left ventricular contour, and the left ventricular posterior wall contour, and determine the short side slopes of the respective minimum area circumscribed rectangles; take the average value of the short side slopes of the respective minimum area circumscribed rectangles as the slope of the measurement line.

[0166] In an embodiment of the present application, the above first correction unit may specifically be configured to, for each midpoint, determine the two line segment endpoints of the adjacent contour measurement line segment corresponding to the midpoint; use the region formed by the contour edges where the two line segment endpoints are located as the gradient calculation region, and calculate the image gradients of the positions of the points within the gradient calculation region; take the point position corresponding to the maximum image gradient as the new center point, and extend the contour edges where the two line segment endpoints are located to the new center point to obtain the corrected tissue contour; the corrected tissue contour includes the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected left ventricular posterior wall contour.

[0167] In an embodiment of the present application, the above second determination unit may specifically be configured to take the length of the intersecting line segment between the corrected right ventricular contour and the measurement line as the anteroposterior diameter of the right ventricle; take the length of the intersecting line segment between the corrected interventricular septum contour and the measurement line as the thickness of the interventricular septum; take the length of the intersecting line segment between the corrected left ventricular contour and the measurement line as the anteroposterior diameter of the left ventricle; take the length of the intersecting line segment between the corrected left ventricular posterior wall contour and the measurement line as the thickness of the left ventricular posterior wall.

[0168] In one embodiment of the present application, the edge arc information of the second target tissue may include the aortic and left ventricular anterior wall arcs, and the posterior aortic wall arc; the second key point information may include the first aortic root and the second aortic root.

[0169] In one embodiment of the present application, for the standard cardiac section in mid-systole, the above-mentioned determination module 4 may be specifically configured to determine the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section in mid-systole according to the aortic and left ventricular anterior wall arcs, the posterior aortic wall arc, the first aortic root, and the second aortic root.

[0170] In one embodiment of the present application, the above-mentioned determination module 3 may include:

[0171] A skeleton diagram generation unit, configured to generate a corresponding skeleton diagram based on the aortic and left ventricular anterior wall arcs and the posterior aortic wall arc;

[0172] A second drawing unit, configured to draw a connection line through the first aortic root and the second aortic root in the skeleton diagram, and determine a first intersection point of the connection line and the aortic and left ventricular anterior wall arcs, and a second intersection point of the connection line and the posterior aortic wall arc;

[0173] A second correction unit, configured to correct the first intersection point and the second intersection point respectively to obtain a corrected first intersection point and a corrected second intersection point;

[0174] A third determination unit, configured to use the connection line between the corrected first intersection point and the corrected second intersection point as the aortic annulus diameter measurement line segment, and use the length of the aortic annulus diameter measurement line segment as the aortic annulus diameter;

[0175] A fourth determination unit, configured to determine the midpoint of the aortic annulus diameter measurement line segment;

[0176] A translation unit, configured to translate the midpoint of the aortic annulus diameter measurement line segment by a preset length in a preset direction to obtain a translation point;

[0177] A third drawing unit, configured to draw a left ventricular outflow tract diameter straight line through the translation point according to the slope of the aortic annulus diameter measurement line segment, and determine a third intersection point of the left ventricular outflow tract diameter straight line and the aortic and left ventricular anterior wall arcs, and a fourth intersection point of the left ventricular outflow tract diameter straight line and the posterior aortic wall arc;

[0178] A third correction unit, configured to correct the third intersection point and the fourth intersection point respectively to obtain a corrected third intersection point and a corrected fourth intersection point;

[0179] A fifth determination unit, configured to use the length of the connection line between the corrected third intersection point and the corrected fourth intersection point as the left ventricular outflow tract diameter.

[0180] In one embodiment of the present application, the second correction unit may specifically be configured to calculate the image gradients of the positions of the points on the aortic and left ventricular anterior wall arcs, and use the position of the point corresponding to the maximum image gradient on the aortic and left ventricular anterior wall arcs as the corrected first intersection point; calculate the image gradients of the positions of the points on the aortic posterior wall arc, and use the position of the point corresponding to the maximum image gradient on the aortic posterior wall arc as the corrected second intersection point.

[0181] In one embodiment of the present application, for the standard cardiac section at the end of cardiac systole, the determining module 4 may specifically be configured to determine the interventricular septum thickness, left ventricular anteroposterior diameter, and left ventricular posterior wall thickness of the standard cardiac section at the end of cardiac systole according to the right ventricular contour, interventricular septum contour, left ventricular contour, left ventricular posterior wall contour, and mitral valve point.

[0182] For the introduction of the device provided in the embodiments of the present application, please refer to the above method embodiments, and the present application will not elaborate herein.

[0183] The embodiments of the present application provide an electronic device.

[0184] Please refer to Figure 6 , Figure 6 , which is a schematic structural diagram of an electronic device provided by the present application. The electronic device may include:

[0185] A memory for storing computer programs;

[0186] A processor, which when executing the computer program can implement the steps of any of the above ultrasonic image processing methods.

[0187] As Figure 6 shown, which is a schematic structural diagram of the composition of an electronic device. The electronic device may include: a processor 10, a memory 11, a communication interface 12, and a communication bus 13. The processor 10, the memory 11, and the communication interface 12 all complete communication with each other through the communication bus 13.

[0188] In the embodiments of the present application, the processor 10 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 devices, etc.

[0189] The processor 10 may call the program stored in the memory 11. Specifically, the processor 10 may execute the operations in the embodiments of the ultrasonic image processing method.

[0190] The memory 11 is used to store one or more programs. The program may include program codes, and the program codes include computer operation instructions. In the embodiments of the present application, the memory 11 stores at least a program for implementing the following functions:

[0191] Obtain a phase plane recognition model; the phase plane recognition model includes a left ventricle segmentation unit;

[0192] Input the cardiac ultrasound video into the phase plane recognition model to trigger the phase plane recognition model to obtain at least one cardiac standard plane corresponding to each phase based on the detected image features and the left ventricle segmentation result;

[0193] Use an image segmentation model to process the cardiac standard planes at each phase to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the cardiac standard plane;

[0194] Parallelly determine the detection results of the cardiac standard planes at each phase according to the point-line features of each target tissue.

[0195] In a possible implementation, the memory 11 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function, etc.; the data storage area may store the data created during use.

[0196] In addition, the memory 11 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device or other volatile solid-state storage devices.

[0197] The communication interface 12 may be an interface of a communication module for connecting to other devices or systems.

[0198] Of course, it should be noted that Figure 6 The structure shown does not constitute a limitation on the electronic device in the embodiments of the present application. In actual applications, the electronic device may include more or fewer components than Figure 6 shown, or combine certain components.

[0199] The embodiments of the present application provide a computer-readable storage medium.

[0200] The computer program stored on the computer-readable storage medium provided by the embodiments of the present application can implement the steps of any of the above ultrasonic image processing methods when executed by a processor.

[0201] The computer-readable storage medium may include: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0202] For the introduction of the computer-readable storage medium provided by the embodiments of the present application, please refer to the above method embodiments, and the present application will not elaborate here.

[0203] The various embodiments in the specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method section.

[0204] Those skilled in the art can further realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of the examples have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0205] The steps of the methods or algorithms described in combination with the embodiments disclosed herein can be directly implemented by hardware, software modules executed by a processor, or a combination of the two. The software modules can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium well-known in the technical field.

[0206] The technical solutions provided in this application have been introduced in detail above. Specific examples are used herein to elaborate on the principles and implementation manners of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. An ultrasonic image processing method, characterized in that Including: Obtaining a phase section recognition model; the phase section recognition model includes a left ventricle segmentation unit; Inputting a cardiac ultrasound video into the phase section recognition model to trigger the phase section recognition model to obtain at least one cardiac standard section corresponding to each phase based on the detected image features and the left ventricle segmentation result; Processing the cardiac standard section at each phase by using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the cardiac standard section; Parallelly determining the detection result of the cardiac standard section at each phase according to the point-line features of each target tissue.

2. The ultrasonic image processing method according to claim 1, wherein The at least one phase includes the end-diastolic phase of the heart, the mid-systolic phase of the heart, and the end-systolic phase of the heart; For the cardiac standard section in the end-diastolic phase of the heart and the end-systolic phase of the heart, the point-line features of each target tissue in the cardiac standard section include the contour information of the first target tissue and the first key point information; For the cardiac standard section in the mid-systolic phase of the heart, the point-line features of each target tissue in the cardiac standard section include the edge arc information of the second target tissue and the second key point information.

3. The ultrasonic image processing method according to claim 2, characterized in that, The contour information of the first target tissue includes the right ventricle contour, the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour; the first key point information includes the mitral valve point.

4. The ultrasonic image processing method according to claim 3, wherein For the cardiac standard section in the end-diastolic phase of the heart, the parallelly determining the detection result of the cardiac standard section at each phase according to the point-line features of each target tissue includes: Determining the anteroposterior diameter of the right ventricle, the interventricular septum thickness, the anteroposterior diameter of the left ventricle, and the left ventricular posterior wall thickness of the cardiac standard section in the end-diastolic phase of the heart according to the right ventricle contour, the interventricular septum contour, the left ventricle contour, the left ventricular posterior wall contour, and the mitral valve point.

5. The ultrasonic image processing method according to claim 4, wherein, The determining the anteroposterior diameter of the right ventricle, the interventricular septum thickness, the anteroposterior diameter of the left ventricle, and the left ventricular posterior wall thickness of the cardiac standard section in the end-diastolic phase of the heart according to the right ventricle contour, the interventricular septum contour, the left ventricle contour, the left ventricular posterior wall contour, and the mitral valve point includes: Calculating the slope of the measurement line according to the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour; Drawing a measurement line through the mitral valve point according to the slope of the measurement line, and the measurement line intersects with the right ventricle contour, the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour; Determining adjacent contour measurement line segments in the measurement line, and the adjacent contour measurement line segments include the measurement line segment between the right ventricle contour and the interventricular septum contour, the measurement line segment between the interventricular septum contour and the left ventricle contour, and the measurement line segment between the left ventricle contour and the left ventricular posterior wall contour; Determining the midpoint of each adjacent contour measurement line segment, and using each midpoint to correct the right ventricle contour, the interventricular septum contour, the left ventricle contour, and the left ventricular posterior wall contour to obtain the corrected right ventricle contour, the corrected interventricular septum contour, the corrected left ventricle contour, and the corrected left ventricular posterior wall contour; Determine the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle.

6. The ultrasonic image processing method according to claim 5, wherein Calculating the slope of the measurement line according to the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle includes: Determine the minimum area circumscribed rectangles of the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle, and determine the short side slopes of the minimum area circumscribed rectangles. Take the mean value of the short side slopes of the minimum area circumscribed rectangles as the slope of the measurement line.

7. The ultrasonic image processing method according to claim 5, characterized in that Using each of the midpoints to correct the right ventricular contour, the interventricular septum contour, the left ventricular contour, and the posterior wall contour of the left ventricle to obtain the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle includes: For each of the midpoints, determine the two line segment endpoints of the adjacent contour measurement line segment corresponding to the midpoint. Take the region formed by the contour edges where the two line segment endpoints are located as the gradient calculation region, and calculate the image gradient at each point position within the gradient calculation region. Take the point position corresponding to the maximum image gradient as the new center point, and extend the contour edges where the two line segment endpoints are located to the new center point to obtain the corrected tissue contour; the corrected tissue contour includes the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle.

8. The ultrasonic image processing method according to claim 5, wherein Determining the anteroposterior diameter of the right ventricle, the thickness of the interventricular septum, the anteroposterior diameter of the left ventricle, and the thickness of the posterior wall of the left ventricle according to the corrected right ventricular contour, the corrected interventricular septum contour, the corrected left ventricular contour, and the corrected posterior wall contour of the left ventricle includes: Take the length of the intersecting line segment between the corrected right ventricular contour and the measurement line as the anteroposterior diameter of the right ventricle. Take the length of the intersecting line segment between the corrected interventricular septum contour and the measurement line as the thickness of the interventricular septum. Take the length of the intersecting line segment between the corrected left ventricular contour and the measurement line as the anteroposterior diameter of the left ventricle. Take the length of the intersecting line segment between the corrected posterior wall contour of the left ventricle and the measurement line as the thickness of the posterior wall of the left ventricle.

9. The ultrasonic image processing method according to claim 2, wherein The edge arc information of the second target tissue includes the aortic and left ventricular anterior wall arcs and the aortic posterior wall arc; the second key point information includes the first aortic valve root and the second aortic valve root.

10. The ultrasonic image processing method according to claim 9, wherein, For the standard cardiac section at mid-systole, determining the detection results of the standard cardiac section at each time phase in parallel according to the point-line characteristics of each target tissue includes: Determine the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section at mid-systole according to the aortic and left ventricular anterior wall arcs, the aortic posterior wall arc, the first aortic valve root, and the second aortic valve root.

11. The ultrasonic image processing method according to claim 10, wherein Determining the aortic annulus diameter and the left ventricular outflow tract diameter of the standard cardiac section at mid - systolic phase according to the aortic and anterior left ventricular wall arcs, the posterior aortic wall arc, the first aortic root, and the second aortic root, includes: Generating a corresponding skeleton diagram based on the aortic and anterior left ventricular wall arcs and the posterior aortic wall arc; In the skeleton diagram, drawing a connection line through the first aortic root and the second aortic root, and determining a first intersection point of the connection line and the aortic and anterior left ventricular wall arc, and a second intersection point of the connection line and the posterior aortic wall arc; Respectively correcting the first intersection point and the second intersection point to obtain a corrected first intersection point and a corrected second intersection point; Taking the connection line between the corrected first intersection point and the corrected second intersection point as the aortic annulus diameter measurement line segment, and taking the length of the aortic annulus diameter measurement line segment as the aortic annulus diameter; Determining the mid - point of the aortic annulus diameter measurement line segment; Translating the mid - point of the aortic annulus diameter measurement line segment by a preset length in a preset direction to obtain a translated point; Drawing a left ventricular outflow tract diameter straight line through the translated point according to the slope of the aortic annulus diameter measurement line segment, and determining a third intersection point of the left ventricular outflow tract diameter straight line and the aortic and anterior left ventricular wall arc, and a fourth intersection point of the left ventricular outflow tract diameter straight line and the posterior aortic wall arc; Respectively correcting the third intersection point and the fourth intersection point to obtain a corrected third intersection point and a corrected fourth intersection point; Taking the length of the connection line between the corrected third intersection point and the corrected fourth intersection point as the left ventricular outflow tract diameter.

12. The ultrasonic image processing method according to claim 11, wherein The respectively correcting the first intersection point and the second intersection point to obtain a corrected first intersection point and a corrected second intersection point includes: Calculating the image gradient of each point position on the aortic and anterior left ventricular wall arc, and taking the point position corresponding to the maximum image gradient on the aortic and anterior left ventricular wall arc as the corrected first intersection point; Calculating the image gradient of each point position on the posterior aortic wall arc, and taking the point position corresponding to the maximum image gradient on the posterior aortic wall arc as the corrected second intersection point.

13. The ultrasonic image processing method according to claim 3, wherein For the standard cardiac section at end - systolic phase, the parallel determination of the detection results of the standard cardiac section at each time phase according to the point - line characteristics of each target tissue includes: Determining the interventricular septum thickness, the left ventricular anteroposterior diameter, and the left ventricular posterior wall thickness of the standard cardiac section at end - systolic phase according to the right ventricular contour, the interventricular septum contour, the left ventricular contour, the left ventricular posterior wall contour, and the mitral valve point.

14. An ultrasonic image processing apparatus, characterized in that, Including: An acquisition module for acquiring a time - phase section recognition model; the time - phase section recognition model includes a left ventricular segmentation unit; A recognition module for inputting a cardiac ultrasound video into the time - phase section recognition model to trigger the time - phase section recognition model to obtain at least one corresponding standard cardiac section at each time phase based on the detected image features and the left ventricular segmentation result; A processing module, configured to process the standard cardiac cross-sections at each time phase by using an image segmentation model to obtain a segmentation result; the segmentation result includes the point-line features of each target tissue in the standard cardiac cross-sections. A determination module, configured to determine the detection results of the standard cardiac cross-sections at each time phase in parallel according to the point-line features of each target tissue.

15. An electronic device, characterized in that, Comprising: A memory, configured to store a computer program; A processor, configured to implement the steps of the ultrasonic image processing method according to any one of claims 1 to 13 when executing the computer program.

16. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the ultrasonic image processing method according to any one of claims 1 to 13 are implemented.

Images