Heart valve regurgitation detection method and system and medium

Abstract

The invention provides a heart valve regurgitation detection method and system and a medium. The method comprises the steps that regurgitation area identification caused by heart valve structure or function abnormity is obtained; performing principal component analysis on the regurgitation area, and determining the main extension direction of the regurgitation of the corresponding heart valve; determining at least one morphological characteristic parameter of the reflux region according to the main extension direction; and determining the regurgitation state information of the heart valve according to the morphological characteristic parameters. According to the method, the spatial distribution of a reflux region is subjected to characteristic decomposition through principal component analysis, a complex irregular blood flow form is converted into a main extension direction with definite physical significance, and form characteristic parameters are calculated based on the main extension direction; random errors caused by subjective view angle differences or insufficient experience when valve reflux parameters are manually measured can be eliminated, and valve reflux evaluation results obtained by different operators under different detection conditions have high consistency and repeatability.

Description

Technical Field

[0001] This invention relates to the field of medical imaging, and more specifically to a method, system, and medium for detecting valvular regurgitation. Background Technology

[0002] In the clinical diagnosis of valvular heart disease, accurate detection and quantitative assessment of valvular regurgitation (such as mitral regurgitation) caused by congenital or acquired diseases are crucial, as they are key to developing appropriate treatment plans and assessing patient prognosis. Currently, the clinical detection and quantitative assessment of valvular heart disease mainly relies on the experience of ultrasound physicians for identification and manual measurement, which inevitably has subjective limitations.

[0003] Specifically, physicians must visually identify and manually delineate abnormal blood flow areas at the heart valve orifices using their experience and expertise in color Doppler ultrasound flow mode. This process requires sonographers to accurately identify the turbulent morphology of valvular regurgitation presented in the high-velocity blood flow area on the ultrasound image. Subsequently, using the built-in measurement tools of the ultrasound equipment, they manually measure key parameters such as the radius of the proximal isovelocity surface and the width of the constriction neck of the regurgitation jet on multiple still frames, and complete the data recording after multiple manual calibrations.

[0004] However, this traditional manual measurement method has many significant shortcomings and drawbacks. For example, from an efficiency perspective, the manual measurement process is extremely time-consuming and labor-intensive. Ultrasound physicians need to carefully search for and delineate abnormal blood flow areas in a large number of images, and repeatedly measure and calibrate each key parameter, which severely limits work efficiency and makes it difficult to meet the urgent need for high efficiency in modern clinical diagnosis.

[0005] From an accuracy perspective, measurement results are greatly influenced by the physician's individual experience and judgment. Different physicians, due to variations in their professional level, experience, and visual judgment, may arrive at different results when depicting abnormal blood flow areas and measuring key parameters. This subjective difference leads to a lack of consistency in measurement results among different physicians, consequently significantly reducing the accuracy and reliability of diagnostic findings. Summary of the Invention

[0006] One of the objectives of this invention is to provide a method for detecting valvular regurgitation in the heart, in order to solve the technical problems of existing technologies in detecting valvular regurgitation, which rely heavily on human experience, have high subjectivity, poor consistency and low repeatability.

[0007] To achieve one of the above-mentioned objectives, the present invention provides a method for detecting valvular regurgitation, comprising: identifying a regurgitation region caused by structural or functional abnormalities of the heart valve; performing principal component analysis on the regurgitation region to determine the main extension direction of the corresponding valvular regurgitation; determining at least one morphological feature parameter of the regurgitation region based on the main extension direction; and determining valvular regurgitation status information based on the morphological feature parameter.

[0008] As a further improvement of one embodiment of the present invention, the morphological characteristic parameters characterize the degree of valvular regurgitation.

[0009] As a further improvement of one embodiment of the present invention, the morphological feature parameter includes the neck width of the backflow jet; determining at least one morphological feature parameter of the backflow region according to the main extension direction includes: obtaining a plurality of straight lines that have at least two intersection points with the contour of the backflow region, the plurality of straight lines being perpendicular to the main extension direction; determining the minimum distance as the neck width of the backflow jet based on the distance between the two intersection points corresponding to each straight line.

[0010] As a further improvement of one embodiment of the present invention, the morphological feature parameter includes the proximal isokinetic surface radius; determining at least one morphological feature parameter of the regurgitation region according to the main extension direction includes: determining the position of the heart valves according to the ultrasound image of the heart; projecting several edge points of the regurgitation region onto the main extension direction and determining several corresponding projection values; performing minimum bounding circle fitting on the set of edge points whose projection values ​​fall within a preset range, and determining the proximal isokinetic surface radius according to the minimum bounding circle radius.

[0011] As a further improvement of one embodiment of the present invention, the method includes: determining a maximum projection value and a minimum projection value among a plurality of projection values; determining a preset range based on the maximum projection value and the minimum projection value; wherein the preset range includes a first boundary value and a second boundary value, the first boundary value being less than the second boundary value; the first boundary value being equal to the difference between the maximum projection value and a first intermediate value, the first intermediate value being equal to the product of a preset threshold and the difference between the maximum projection value and the minimum projection value, and the second boundary value being the maximum projection value.

[0012] As a further improvement of one embodiment of the present invention, the method further includes: assessing the severity of cardiac valvular regurgitation based on the morphological characteristic parameters and the area of ​​the valvular regurgitation region.

[0013] As a further improvement of one embodiment of the present invention, the method further includes: obtaining the total number of pixels in the reflux region and the pixel area of ​​each pixel; and determining the area of ​​the valve reflux region based on the product of the total number and the pixel area.

[0014] As a further improvement of one embodiment of the present invention, the step of performing principal component analysis on the regurgitation region to determine the main extension direction of the corresponding heart valve regurgitation includes: determining the coordinate information of several pixels in the regurgitation region, constructing a covariance matrix based on the coordinate information, and determining the main extension direction of the corresponding heart valve regurgitation based on the maximum eigenvalue corresponding to the covariance matrix.

[0015] As a further improvement of one embodiment of the present invention, obtaining the regurgitation region caused by the heart includes: obtaining an ultrasound image of the heart and identifying several corresponding abnormal blood flow regions; determining whether the location information of each abnormal blood flow region meets a first condition and whether the corresponding morphological information meets a second condition; if so, then determining that the corresponding abnormal sequence region is a regurgitation region caused by the heart valve.

[0016] As a further improvement of one embodiment of the present invention, the step of determining whether the location information of each abnormal blood flow region satisfies the first condition and whether the corresponding morphological information satisfies the second condition includes: determining the corresponding atrium and ventricle based on the ultrasound image of the heart; determining whether the abnormal blood flow region is located at the junction of the atrium and ventricle; if so, then determining that the abnormal blood flow region satisfies the first condition.

[0017] As a further improvement of one embodiment of the present invention, the step of determining that the abnormal blood flow region meets the first condition includes: determining the corresponding heart valve orifice based on the ultrasound image of the heart; determining whether the distance between the abnormal blood flow region and the heart valve orifice meets a first preset value; if so, determining that the current location information of the abnormal blood flow region meets the first condition.

[0018] As a further improvement of one embodiment of the present invention, the step of determining whether the location information of each abnormal blood flow region satisfies the first condition and whether the corresponding morphological information satisfies the second condition includes: determining whether the abnormal blood flow region is a connected region; if so, when the shape of the abnormal blood flow region satisfies the preset standard shape, and / or when the area of ​​the abnormal blood flow region satisfies the second preset value, it is determined that the abnormal blood flow region satisfies the second condition.

[0019] To achieve one of the above-mentioned objectives, the present invention also provides a cardiac valve regurgitation detection system, comprising: an input module for obtaining identification of regurgitation regions caused by abnormalities in the structure or function of cardiac valves; a processing module for performing principal component analysis on the regurgitation regions to determine the main extension direction of the corresponding cardiac valve regurgitation; a detection module for determining at least one morphological feature parameter of the regurgitation region based on the main extension direction; and for determining cardiac valve regurgitation status information based on the morphological feature parameter.

[0020] To achieve one of the above-mentioned objectives, the present invention also provides a computer storage medium, comprising: at least one processor; and a memory storing a computer program running on the processor, wherein the processor executes the program to perform the steps of the method for detecting valvular regurgitation.

[0021] Compared with the prior art, the embodiments of the present invention have at least one of the following beneficial effects: This invention employs a method for detecting valvular regurgitation. Principal component analysis is used to decompose the spatial distribution of the regurgitation region, transforming the complex irregular blood flow pattern into a main extension direction with clear physical meaning. This direction accurately reflects the main flow trend of regurgitated blood. Based on this main extension direction, morphological characteristic parameters are calculated, which can eliminate random errors caused by subjective perspective differences or lack of experience when manually measuring valvular regurgitation parameters. This ensures that the valvular regurgitation assessment results obtained by different operators and under different testing conditions have high consistency and repeatability. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the structure of a cardiac valve regurgitation detection system according to one embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of the steps of a method for detecting cardiac valve regurgitation in one embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of step S1 in one embodiment of the present invention.

[0025] Figure 4(a) is a schematic diagram of step S12 in a specific embodiment of an embodiment of the present invention.

[0026] Figure 4(b) is a schematic diagram of step S12 in a specific embodiment of the present invention.

[0027] Figure 5 This is a schematic diagram of step S2 in one embodiment of the present invention.

[0028] Figure 6(a) is a schematic diagram of step S4 in one embodiment of the present invention.

[0029] Figure 6(b) is a schematic diagram of step S3 in one embodiment of the present invention.

[0030] Figure 6(c) is a schematic diagram of step S3 in another embodiment of the present invention.

[0031] Figure 7 This is a schematic diagram of the structure for determining the narrowing neck width of the backflow jet in one embodiment of the present invention.

[0032] Figure 8This is a schematic diagram illustrating the steps of determining a preset range in one embodiment of the present invention.

[0033] Figure 9 This is a flowchart illustrating a preferred embodiment of the cardiac valve regurgitation detection method of the present invention. Detailed Implementation

[0034] The present invention will now be described in detail with reference to the specific embodiments shown in the accompanying drawings. However, these embodiments do not limit the present invention, and any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the scope of protection of the present invention.

[0035] It should be noted that the term "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0036] like Figure 1 As shown, a cardiac valve regurgitation detection system 100 is provided in one embodiment of the present invention.

[0037] The described cardiac valvular regurgitation detection system 100 is an intelligent diagnostic system based on medical image analysis, primarily used to detect and assess the severity of cardiac valvular regurgitation through automated algorithms. Its core function is to provide clinicians with objective and accurate quantitative analysis results of cardiac valvular regurgitation by replacing traditional manual measurements with computer-aided methods.

[0038] The cardiac valve regurgitation detection system 100 includes an input module 11.

[0039] Input module 11 is used to identify regurgitation regions caused by abnormalities in the structure or function of heart valves.

[0040] Specifically, the input module 100, as the "sensing layer" of the system, is responsible for interacting with external data sources (such as medical imaging equipment) and capturing images or signals of the heart valve regurgitation area through imaging technologies such as echocardiography to obtain raw data.

[0041] The cardiac valve regurgitation detection system 100 includes a processing module 12.

[0042] Processing module 12 is used to perform principal component analysis on the reflux region to determine the main extension direction of the corresponding heart valve reflux.

[0043] Specifically, the processing module 12, as the "analysis layer" of the system, is responsible for in-depth processing and analysis of the input data. Through in-depth analysis, the raw data is transformed into meaningful information to support subsequent decision-making.

[0044] The cardiac valve regurgitation detection system 100 includes a detection module 13.

[0045] On the one hand, the detection module 13 is used to determine at least one morphological feature parameter of the reflux region based on the main extension direction.

[0046] On the other hand, the detection module 13 is used to determine the valvular regurgitation status information based on the morphological feature parameters.

[0047] Specifically, the detection module 13, as the "decision layer" of the system, is responsible for making specific detections or evaluations based on the information provided by the processing module 12 and presenting them to the user.

[0048] In one embodiment, the cardiac valve regurgitation detection system 100 further includes an output module 14.

[0049] The output module 14 is used to output the test results to the display screen. Specifically, it can present the morphological feature parameters to the user in an intuitive and easy-to-understand way, such as presenting the results dynamically in real time in the form of numerical values, annotations, charts, reports, etc., which is convenient for doctors to interpret.

[0050] The following will combine Figure 1 The content shown describes the working process of the heart valve regurgitation detection system 100.

[0051] The input module 11, as the data acquisition front end, first interfaces with a medical imaging device with cardiac ultrasound examination function to accurately capture image data of the heart valve regurgitation area and complete preliminary denoising, image enhancement and other preprocessing operations to ensure the quality of input data and lay the foundation for subsequent analysis.

[0052] The processing module 12 performs in-depth mining on the preprocessed data transmitted by the input module 11, accurately extracting the key feature information of the main extension direction of the heart valve regurgitation. At the same time, it can further calculate features such as regurgitation velocity and regurgitation area, providing multi-dimensional analysis basis for the detection module 13.

[0053] Based on the main extension direction and characteristic information of regurgitation output by the processing module 12, the detection module 13 calculates the morphological characteristic parameters related to valvular regurgitation through a preset algorithm model, and integrates these parameters into an intuitive detection report or visualization chart, which is directly fed back to the clinician to help him accurately judge the severity of valvular regurgitation and specify a scientific treatment plan.

[0054] Throughout the entire testing process, the input module 11, processing module 12, and testing module 13 work closely together, with data flowing efficiently and being processed in depth between the modules, together forming a complete closed-loop system from data acquisition to clinical decision support.

[0055] like Figure 2 As shown, the present invention provides a method for detecting valvular regurgitation.

[0056] The method for detecting valvular regurgitation is applied to a system for detecting valvular regurgitation.

[0057] In one embodiment, the cardiac valve regurgitation detection system can be as follows: Figure 1 The configuration described above is applied, and the corresponding technical solutions are referenced in the detection method provided by this invention. However, the cardiac valve regurgitation detection system used in the cardiac valve regurgitation detection method provided by this invention is not limited to this structural configuration.

[0058] like Figure 2 As shown, the method for detecting valvular regurgitation provided by the present invention includes, but is not limited to, the following steps.

[0059] Step S1: Identify the regurgitation region caused by abnormalities in the structure or function of the heart valve; Step S2: Perform principal component analysis on the regurgitation region to determine the main extension direction of the corresponding heart valve regurgitation; Step S3: Determine at least one morphological characteristic parameter of the reflux region based on the main extension direction; Step S4: Determine the ventricular regurgitation status information based on the morphological characteristic parameters.

[0060] Thus, by using principal component analysis to decompose the spatial distribution of the valvular regurgitation region, the complex irregular blood flow pattern is transformed into a main extension direction with clear physical meaning. This direction accurately reflects the main flow trend of regurgitated blood. Based on this main extension direction, morphological characteristic parameters can be calculated, which can eliminate random errors caused by subjective perspective differences or lack of experience when manually measuring valvular regurgitation parameters. This ensures that the valvular regurgitation assessment results obtained by different operators and under different testing conditions have high consistency and repeatability.

[0061] In step S1, the valvular regurgitation region refers to the area through which abnormal blood flows backward from the valve orifice during cardiac contraction or relaxation due to valvular insufficiency caused by structural or functional abnormalities of the heart valves. Physiological or mild valvular regurgitation does not cause serious consequences, and patients generally have no obvious discomfort symptoms. However, moderate to severe valvular regurgitation may lead to reduced cardiac output, increased pulmonary congestion, and subsequently cause pulmonary hypertension, cerebral ischemia, arrhythmia, or even heart failure.

[0062] In one embodiment, the regurgitation region caused by the heart valve includes at least one of mitral regurgitation and tricuspid regurgitation.

[0063] Specifically, the mitral valve is located between the left atrium and the left ventricle. Under normal conditions, when the left ventricle contracts, the mitral valve closes to prevent blood from flowing back from the left ventricle to the left atrium. Mitral regurgitation refers to the phenomenon where, during left ventricular systole, due to incomplete closure of the mitral valve, some blood flows back from the left ventricle to the left atrium.

[0064] The tricuspid valve is located between the right atrium and the right ventricle. Its function is to close during right ventricular contraction to prevent blood from flowing back from the right ventricle to the right atrium. Tricuspid regurgitation refers to the phenomenon where, during right ventricular systole, the tricuspid valve fails to close completely, allowing some blood to flow back from the right ventricle to the right atrium.

[0065] In one embodiment, the regurgitation region includes the regurgitation jet and the area surrounding the regurgitation jet. The regurgitation jet refers to a jet of blood flow with a specific color and / or shape that appears in the color Doppler flow pattern of cardiac ultrasound when blood flows back from the ventricle to the atrium due to valvular insufficiency.

[0066] For example, in mitral regurgitation, some blood in the left ventricle during systole flows back into the left atrium through the incompletely closed mitral valve. In cardiac ultrasound color Doppler flow mode, the regurgitation jet appears as a colored blood flow signal from the left ventricle to the left atrium, and the color can be blue or red depending on the blood flow velocity and direction.

[0067] In step S4, the valvular regurgitation status information is determined based on morphological characteristic parameters. This step can be performed by a machine or assessed manually. Of course, some steps in this method can be completed by a machine while others can be performed manually. Furthermore, the machine output is not necessarily the final result; it can also be process data to assist the ultrasound physician in making a judgment.

[0068] Furthermore, information on the status of valvular regurgitation can be understood as a grade or score. Specifically, by comparing quantitative parameters with standard thresholds, the severity of valvular regurgitation is classified into objective grades (e.g., mild / moderate / severe) or a comprehensive score is calculated (e.g., weighted fusion of multiple parameters into a risk score), thereby providing a standardized basis for the diagnostic process. The quantitative parameters include at least one of the following: the area of ​​the valvular regurgitation region, the width of the constriction neck of the regurgitation jet, and the PISA (Proximal Isovelocity Surface Area) radius.

[0069] like Figure 3 As shown, in one embodiment, step S1 may specifically include the following steps.

[0070] Step S11: Obtain ultrasound images of the heart and identify several corresponding abnormal blood flow areas; Step S12: Determine whether the location information of each abnormal blood flow region meets the first condition and whether the corresponding morphological information meets the second condition. If so, proceed to step S13 to determine if the corresponding abnormal sequence region is a regurgitation region caused by abnormal heart valve structure or function.

[0071] Thus, by setting the first condition (location verification) and the second condition (morphological verification), the specific anatomical location (e.g., near the valve) and morphological characteristics of valvular regurgitation can be verified, which can exclude interference from other abnormal blood flow areas caused by non-valvular regurgitation and improve the accuracy of valvular regurgitation detection.

[0072] In step S11, abnormal blood flow regions can be automatically identified based on neural networks and / or threshold-based image segmentation methods.

[0073] In one specific embodiment, a trained neural network model is used to identify the abnormal blood flow region in the image to be detected.

[0074] Specifically, a multi-modal dataset of labeled cardiac ultrasound images is obtained, with the labeled content being abnormal blood flow regions; a convolutional neural network model is trained based on the multi-modal dataset of cardiac ultrasound images to obtain a trained neural network model; the trained neural network model is then used to identify and determine abnormal blood flow regions in the images to be detected.

[0075] In another specific embodiment, the abnormal blood flow region is determined by utilizing the color or brightness characteristics of pixels in the cardiac ultrasound color Doppler pattern image.

[0076] Specifically, the process involves acquiring color Doppler ultrasound images of the heart, transforming these images in color space (e.g., converting RGB images to HSV (Hue, Saturation, Value) or Lab color space), constructing pixel classification indices based on chromaticity (H) and saturation (S) in the transformed color space, and determining a preset hue range [H]. min H max ] and saturation threshold S thresh For any pixel (x, y) in the detected image, if the hue of the pixel belongs to the preset hue range [H], then... min H max That is, H(x, y) ∈ [H min H max And the saturation of this pixel is greater than the saturation threshold S. thresh (That is, S(x,y)>S) threshIf the detected area is marked as an abnormal blood flow area, then the corresponding area of ​​the detected image is marked as such.

[0077] In another specific embodiment, the abnormal blood flow region is determined based on the watershed region segmentation method.

[0078] Specifically, a color detection image of the heart is acquired, converted into a grayscale image, and the gradient magnitude G(x, y) of the grayscale image is calculated to construct an edge intensity image. Based on distance transformation, the foreground and background regions in the edge intensity image are determined, and the foreground region is assigned a unique label, while the background region is marked as 0, forming an initial label map pointing to integers. A boundary is constructed between two different labeled regions to form several region segmentation maps. The abnormal blood flow region is determined based on at least one of the region area, location, and shape.

[0079] In one specific embodiment, the cardiac detection images include at least one of a cardiac ultrasound color Doppler blood flow pattern static image and a blood flow video.

[0080] For example, a sequence of continuous color Doppler ultrasound images of the apical four-chamber view is acquired and preprocessed, including speckle noise removal, histogram equalization, and adaptive filtering at least one of these operations. Image preprocessing can improve image quality, making the boundaries and details of abnormal sequence regions clearer.

[0081] Among them, the color Doppler mode image of cardiac ultrasound utilizes the Doppler effect to superimpose information such as blood flow velocity and direction onto a two-dimensional grayscale ultrasound image in a encoded form. Red indicates blood flow towards the probe, and blue indicates blood flow away from the probe; the intensity of the color indicates the speed of blood flow. Through this image, the distribution and flow of blood within the heart can be visually determined, including the presence of abnormal blood flow phenomena such as reflux.

[0082] Blood flow video can also include color Doppler blood flow video, which is a dynamic representation of color Doppler images in cardiac ultrasound. It continuously records the dynamic changes in blood flow within the heart. By examining blood flow video, ultrasound physicians can gain a more comprehensive understanding of the dynamic characteristics of blood flow, such as changes in regurgitation areas during different cardiac cycles.

[0083] It should be noted that, on the one hand, the detection method described in this invention can complete automatic measurement by relying solely on color Doppler blood flow images or videos, without relying on color Doppler blood flow curve data, and without needing to place pulsed Doppler or continuous Doppler sampling lines on the image to obtain parameters such as blood flow velocity at specific locations, thus avoiding errors caused by improper placement of measurement points.

[0084] On the other hand, the cardiac ultrasound images used in the detection method of the present invention can cover the entire area and intuitively display the blood flow velocity and direction of the heart valves in the form of color coding, which is more in line with the intuitive perception of clinicians and makes it easier for doctors to quickly understand the overall distribution of blood flow in the heart chambers and the valve regurgitation situation.

[0085] In step S11, the abnormal blood flow region refers to the region in the cardiac ultrasound image that is identified from the image using image segmentation technology and has a different blood flow distribution from the normal blood flow distribution.

[0086] In step S12, the first condition is used to verify whether the location information of the abnormal blood flow region matches the location characteristics of valvular regurgitation. The second condition is used to verify whether the morphological information of the abnormal blood flow region matches the morphological characteristics of the valvular regurgitation region.

[0087] Therefore, by validating the abnormal blood flow region (whether it meets the first and second conditions) and constraining the spatial relationship, we can ensure that only genuine valvular regurgitation is analyzed. This significantly reduces the possibility of mistaking normal blood flow for abnormality and improves the specificity of the detection. In other words, we can not only "find" abnormal regurgitation, but also "distinguish" its authenticity, ensuring that subsequent quantification is meaningful.

[0088] As shown in Figure 4(a), in one specific embodiment, step S12 may specifically include the following steps.

[0089] Step S121: Determine the corresponding atria and ventricles based on the echocardiogram images; Step S122: Determine whether the abnormal blood flow area is located at the junction of the atrium and ventricle; If so, proceed to step S123 to determine that the abnormal blood flow area meets the first condition.

[0090] Thus, by detecting the boundaries of the central cavity structure in cardiac ultrasound images, it is possible to initially locate whether abnormal blood flow is located in a high-incidence area of ​​reflux, reducing the computational load of full-image search and avoiding misjudgment.

[0091] In one specific embodiment, it is determined whether the abnormal blood flow area is located at the mitral valve orifice (the junction of the left atrium and the left ventricle).

[0092] In one specific embodiment, it is determined whether the abnormal blood flow area is located at the tricuspid valve orifice (the junction of the right atrium and the right ventricle).

[0093] In the above embodiments, by separately judging the left and right atria and the junction of the left and right ventricles, the sites of high incidence of valvular regurgitation can be more comprehensively covered, improving the accuracy and comprehensiveness of valvular regurgitation detection and providing a more reliable basis for subsequent diagnosis and treatment.

[0094] In one specific embodiment, step S12 may further include the following steps.

[0095] Step S1221: Determine the corresponding heart valve orifice based on the echocardiogram images; Step S1222: Determine whether the distance between the abnormal blood flow area and the heart valve orifice meets the first preset value; If so, proceed to step S1223 to determine that the current location information of the abnormal blood flow area meets the first condition.

[0096] In this way, by quantifying the distance between the abnormal blood flow area and the heart valve orifice, irrelevant turbulence far from the heart valve orifice (such as spontaneous imaging in the atrium) can be ruled out, further pinpointing the source of regurgitation, forming a hierarchical verification (first a large area and then a finer one), and enhancing the robustness of location determination.

[0097] In step S121', the heart valve orifice is the channel entrance formed when the heart valve opens. The heart valve orifice includes at least one of the mitral valve orifice, tricuspid valve orifice, aortic valve orifice, and pulmonary valve orifice.

[0098] In one specific embodiment, it is determined whether the distance between the abnormal blood flow region and the mitral valve orifice meets a first preset value.

[0099] In one specific embodiment, it is determined whether the distance between the abnormal blood flow region and the tricuspid valve orifice meets a first preset value.

[0100] It should be noted that there may be a variety of abnormal blood flow conditions in the heart, and the location and characteristics of the corresponding abnormal blood flow areas are different from those of valvular regurgitation, such as ventricular septal defect causing left and right ventricular shunting (the shunting occurs at the ventricular septum, not at the junction of the atrium and ventricle).

[0101] In addition, based on the different relationships between different abnormal blood flows and valve orifices, such as the high-speed jet in aortic stenosis occurring above the valve orifice rather than subvalvular regurgitation, it can be distinguished from valvular regurgitation by judging the distance.

[0102] As shown in Figure 4(b), in another specific embodiment, step S12 may specifically include the following steps.

[0103] Step S121': Determine whether the abnormal blood flow region is a connected region; If so, proceed to step S122'. When the shape of the abnormal blood flow region meets the preset standard shape, and / or when the area of ​​the abnormal blood flow region meets the second preset value, it is determined that the abnormal blood flow region meets the second condition.

[0104] In this way, by setting standard shape and area thresholds, artifacts (such as speckle noise) or incomplete segmentation regions can be filtered out, ensuring the quality of input data for subsequent principal component analysis.

[0105] In this embodiment, the true regurgitation region caused by valvular insufficiency is a continuous blood flow signal (i.e., a connected region), and its shape and area conform to hydrodynamic characteristics. By analyzing the continuity, size, and morphological characteristics of the abnormal blood flow region, it is helpful to eliminate invalid regions with irregular shapes or that are too small, thus obtaining regurgitation regions with effective detection.

[0106] In one specific embodiment, the preset standard shape includes at least one of a long fan shape and a long ellipse shape.

[0107] In step S2, Principal Component Analysis (PCA) is a technique used to reduce the dimensionality of multidimensional data while preserving the main information in the data, in order to extract the geometrically dominant direction of the reflux region.

[0108] The main extension direction refers to the main flow direction of valvular regurgitation in space, as determined by principal component analysis. It is the reverse direction of the main axis of the regurgitation region, i.e., the main direction of blood flow ejection, which can point towards the valvular orifice.

[0109] like Figure 5 As shown, in one embodiment, step S2 may specifically include the following steps.

[0110] Step S21: Determine the coordinate information of several pixels in the anti-flow region, and construct a covariance matrix based on the coordinate information; Step S22: Determine the main extension direction of the corresponding heart valve regurgitation based on the largest eigenvalue corresponding to the covariance matrix.

[0111] Thus, by employing principal component analysis and using the covariance matrix to decompose the geometric distribution of the backflow region, the main extension direction can be objectively extracted. This direction represents the spatial expansion trend with the strongest energy within the backflow region, thus avoiding subjective bias.

[0112] In step S21, the coordinate information of several pixels in the regurgitation region refers to the position coordinates of each pixel in the image coordinate system located in the heart valve regurgitation region in the detection image.

[0113] Specifically, when the detected image is a two-dimensional image, its corresponding regurgitation region is segmented into several discrete pixels, with corresponding coordinates of (x, y), representing the row and column positions of the pixels in the image. When the detected image is a three-dimensional image, its corresponding regurgitation region is segmented into several discrete voxel points, with corresponding coordinates of (x, y, z), adding depth information. This coordinate information serves as input data for principal component analysis, used to quantify the spatial distribution characteristics of the valve regurgitation region.

[0114] In step S21, the covariance matrix is ​​used to describe the linear relationship between pixel coordinates and reflects the geometric distribution characteristics of the backflow region. The eigenvectors of the covariance matrix point to the main distribution direction of the backflow region, and the eigenvalues ​​represent the discrete distribution along that direction.

[0115] Specifically, the coordinate information of several pixels within the valve regurgitation region is obtained, and the coordinate information of the center point (i.e., mean coordinates) of the corresponding regurgitation region is determined based on the coordinate information of several pixels. A covariance matrix is ​​constructed based on the coordinate information of several pixels and the coordinate information of the center point. The covariance matrix is ​​decomposed into features, and the feature vector corresponding to the largest eigenvalue is the main extension direction.

[0116] To make it easier to understand, for example, let's assume that several pixels in the recirculation region are denoted as... Construct the covariance matrix according to the following formula (1).

[0117] (1) in, Let n be the average of the coordinates of several pixels within the reflux region, where n is the number of pixels. for The transpose of .

[0118] The covariance matrix C is decomposed into eigenvalues. The main extension direction of the backflow region is the largest eigenvector corresponding to this matrix. In this embodiment, the global statistical properties of the covariance matrix are used to weaken local interference and ensure that the main extension direction is determined by the main blood flow distribution, which is consistent with the momentum-dominant direction of the jet in fluid dynamics.

[0119] In one embodiment, the detection method may further include the following steps.

[0120] Step S4: Assess the severity of cardiac valvular regurgitation based on the morphological characteristic parameters and the area of ​​the valvular regurgitation region.

[0121] Among them, the area of ​​the valvular regurgitation region is used to quantitatively assess the severity of valvular regurgitation. By accurately measuring the cross-sectional area of ​​the regurgitation region, it provides an objective basis for clinical classification (such as mild / moderate / severe).

[0122] As shown in Figure 6(a), in one embodiment, step S4 may specifically include the following steps.

[0123] Step S41: Obtain the total number of pixels in the reflux region and the pixel area of ​​each pixel; Step S42: Determine the area of ​​the valve reflux region based on the product of the total number and the pixel area.

[0124] Thus, by summing the areas of all pixels (number of pixels × area per pixel), the regurgitation cross-sectional area can be accurately calculated, providing an objective quantitative basis for classifying the severity of valvular regurgitation (e.g., mild / moderate / severe). This method avoids the contour errors of traditional manual tracing and is particularly suitable for irregular areas.

[0125] In step S42, the valve regurgitation region area refers to the total area of ​​all pixels within the regurgitation region extracted by image segmentation technology, reflecting the cross-sectional area or projected area of ​​the regurgitation jet. The total number of pixels is used to reflect the count of all pixels within the regurgitation region. The pixel area of ​​a single pixel, also called the physical area of ​​a single pixel, is determined based on the spatial resolution of the imaging device (e.g., ultrasound probe parameters).

[0126] In step S3, morphological characteristic parameters are quantitative indicators used to describe the morphological characteristics of the valvular regurgitation region, such as its geometry and spatial distribution. The quantified morphological characteristic parameters provide an objective and unified standard for assessing valvular regurgitation, facilitating comparisons of the morphological characteristics of the valvular regurgitation region in different patients, the same patient at different times, or at different treatment stages.

[0127] In one embodiment, the morphological characteristic parameters include at least one of the neck width of the backflow jet and the radius of the proximal isokinetic surface.

[0128] In one embodiment, the morphological parameter characterizes the energy of cardiac valve regurgitation.

[0129] In this embodiment, the morphological characteristic parameters indirectly quantify the energy characteristics of regurgitated blood flow through geometric morphology. For example, the larger the radius of the proximal isokinetic surface, the larger the area of ​​the blood convergence zone, the higher the regurgitation energy, indicating that the heart valve regurgitation is more severe.

[0130] As shown in Figure 6(b), in one embodiment, the morphological characteristic parameter includes the neck width of the backflow jet. Step S3 may specifically include the following steps.

[0131] Step S311: Obtain a plurality of straight lines that intersect the contour of the backflow region at least at two points, wherein the plurality of straight lines are perpendicular to the main extension direction; Step S312: Determine the minimum distance as the neck width of the backflow stream based on the distance between the two intersection points corresponding to each straight line.

[0132] Thus, by calculating the distance between the intersection of the normal plane of the main extension direction and the regurgitation profile, the consistency of the measurement direction (always perpendicular to the main blood flow axis) can be ensured. Taking the minimum distance as the narrowing neck width of the regurgitation jet conforms to the physical law in fluid mechanics that "the minimum cross-sectional area determines the flow rate," and the results are more in line with clinical needs.

[0133] In step S311, the contour of the reflux region refers to the boundary line connecting the abnormal blood flow regions in the color Doppler blood flow pattern image of cardiac ultrasound, which can be generated by a pixel-level edge detection algorithm. It is a closed curve in a two-dimensional grayscale ultrasound image and a curved surface in a three-dimensional ultrasound image.

[0134] In step S312, the vena contracta width (VCW) refers to the size of the narrowest point when the reflux jet enters the reflux orifice, reflecting the minimum cross-sectional area through which blood flows through the valve defect.

[0135] For ease of understanding, such as Figure 7 As shown, for example, suppose a set of straight lines L1, L2, L3, ..., Ln are generated perpendicular to the main extension direction. These lines are scanned along the axial direction of the backflow beam. Each line intersects the backflow region contour at two points. Let's assume the intersection points of line L1 and the backflow region contour are A1 and A2, line L2 and the backflow region contour are B1 and B2, line L1 and the backflow region contour are C1 and C2, and line Ln and the backflow region contour are T1 and T2. Calculate the distances d1, d2, d3, ..., dn between the two points of each line. The minimum value among all cross-sectional widths is the neck width of the backflow beam. Assuming d2 is the minimum, then d2 is the neck width of the backflow beam.

[0136] As shown in Figure 6(c), in one embodiment, the morphological characteristic parameter includes the radius of the proximal isokinetic surface. Step S3 may specifically include the following steps.

[0137] Step S321: Determine the location of the heart valves based on the ultrasound images of the heart. Step S322: Project several edge points of the reflux region onto the main extension direction and determine several corresponding projection values. Step S323: Perform minimum bounding circle fitting based on the set of edge points whose projection values ​​fall within the preset range, and determine the radius of the near-end isodynamic surface based on the radius of the minimum bounding circle.

[0138] In this way, by accurately locating the acceleration section of the backflow jet in the main extension direction and calculating the near-end isokinetic surface radius (PISA radius) based on the minimum enclosing circle fitting, the consistency between the target position and the hydrodynamic center is ensured, significantly improving the accuracy of the eccentric backflow assessment.

[0139] In step S323, the proximal isokinetic surface radius refers to the radius corresponding to the hemispherical isokinetic layer formed by the regurgitation jet near the heart valve orifice (i.e., the PISA radius).

[0140] Specifically, as blood flows through the narrow openings of the heart valves, it creates accelerated laminar flow, generating a series of isokinetic surfaces (similar to concentric hemispheres) with the same velocity proximal to the valve openings. The radius of the first isokinetic surface is the PISA radius. Figure 7 The radius R shown is given.

[0141] In steps S322 and S323, a position is selected that falls within a preset range from the valve orifice. This position must be located in the acceleration segment of the isovelocity surface (where the blood flow velocity gradient is maximum). The cross-section of the regurgitation region at the target position is transformed into a standardized geometric model to eliminate the influence of morphological irregularities. The least squares method or the minimum bounding circle algorithm is used to fit the smallest circle covering all edge points. The radius of this circle is the PISA radius.

[0142] like Figure 8 As shown, in one specific embodiment, the method further includes the following steps.

[0143] Step M11: Determine the maximum and minimum projection values ​​among several projection values; Step M12: Determine the preset range based on the maximum projection value and the minimum projection value.

[0144] The preset range includes a first boundary value and a second boundary value, wherein the first boundary value is less than the second boundary value; the first boundary value is equal to the difference between the maximum projection value and the first intermediate value, the first intermediate value is equal to the product of the preset threshold and the difference between the maximum projection value and the minimum projection value, and the second boundary value is the maximum projection value.

[0145] In one specific embodiment, the preset threshold is 0.2, and points in the region 20-25% above the heart valve orifice are taken as the edge point set.

[0146] In one specific embodiment, the direction vector corresponding to the main extension direction is normalized. For example, suppose the direction vector of the main extension direction is V. main The corresponding unit is V main =V main / |V main |

[0147] In one specific embodiment, it is determined whether the main extension direction is towards the heart valve orifice; if not, the main extension direction is adjusted to point towards the heart valve orifice. For example, if the main extension direction is in the opposite direction to the heart valve, then its opposite direction becomes the main extension direction, i.e., V. main =- Vmain .

[0148] To facilitate understanding, let's take an example. Suppose that the principal extension direction of the reflux region is determined to be V by principal component analysis. main Regarding the main extension direction V main The corresponding vectors are normalized and their directions are adjusted.

[0149] Project all edge points of the reflux region onto the main extension direction and determine their projection values ​​ti. Determine their corresponding maximum projection value tmax and minimum projection value tmin. Then, based on the maximum projection value tmax and the minimum projection value tmin, determine the first boundary value tmax-0.2*(tmax-tmin) and the second boundary value tmax.

[0150] Select all points whose projection values ​​are in the interval [tmax-0.2*(tmax-tmin), tmax] as the edge point set, and perform minimum bounding circle fitting on the edge point set to obtain the near-end isotropic surface radius.

[0151] The various embodiments, examples, or specific examples provided by this invention can be combined with each other to ultimately form multiple better embodiments.

[0152] Figure 9 A flowchart illustrating a preferred embodiment of a method for detecting valvular regurgitation is shown. The following will combine... Figure 9 This section summarizes the processing procedure of the preferred embodiment.

[0153] Several frames of multi-mode cardiac ultrasound images (i.e., cardiac ultrasound examination images) are obtained, and contour detection is performed on the cardiac ultrasound examination images to determine whether there are abnormal blood flow areas; if no abnormal blood flow areas are detected, the next frame of cardiac examination image is processed.

[0154] If an abnormal blood flow region is detected, the validity of the detected abnormal blood flow region is verified. Specifically, the location information of the abnormal blood flow region is verified to meet the first condition, and the morphological information of the abnormal blood flow region is verified to meet the second condition. When the characteristic information of the abnormal blood flow region meets both the first and second conditions, the abnormal blood flow region is determined to be a regurgitation region caused by abnormal heart valve structure or function. Principal component analysis is then performed on the regurgitation region to determine the main extension direction of the regurgitation jet.

[0155] Based on the regurgitation region and its main extension direction, morphological characteristic parameters of the regurgitation region are determined. These morphological characteristic parameters include at least one of the following: the area of ​​the valvular regurgitation region, the width of the constriction neck of the regurgitation jet, and the radius of the proximal isokinetic surface. Determining the severity of the user's valvular regurgitation based on these morphological characteristic parameters helps in providing accurate diagnostic results and treatment plans.

[0156] One embodiment of the present invention provides a computer-readable storage medium.

[0157] In one embodiment, a computer-readable storage medium stores a computer program executed by the processor mentioned above, or a cardiac valve regurgitation detection method from any of the aforementioned technical solutions.

[0158] When the processor executes the computer program, it can perform the description of the cardiac valve regurgitation detection method in any of the preceding technical solutions; therefore, it will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated.

[0159] The computer-readable storage medium may include: flash drive, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc.

[0160] In summary, the present invention provides a method, system, and medium for detecting valvular regurgitation. Principal component analysis is used to decompose the spatial distribution of the valvular regurgitation region, transforming the complex irregular blood flow pattern into a principal extension direction with clear physical meaning. This direction accurately reflects the main flow trend of regurgitated blood. Based on this principal extension direction, morphological characteristic parameters are calculated, eliminating random errors caused by subjective perspective differences or lack of experience when manually measuring valvular regurgitation parameters. This ensures high consistency and repeatability of valvular regurgitation assessment results obtained by different operators and under different testing conditions.

[0161] It should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0162] The detailed descriptions listed above are merely specific descriptions of feasible embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. All equivalent embodiments or modifications made without departing from the spirit of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting valvular regurgitation, characterized in that, include: Identify regurgitation regions caused by structural or functional abnormalities of heart valves; Principal component analysis was performed on the regurgitation region to determine the main extension direction of the corresponding heart valve regurgitation; At least one morphological feature parameter of the reflux region is determined based on the main extension direction; The morphological characteristic parameters are used to determine the information on the state of valvular regurgitation.

2. The detection method according to claim 1, characterized in that, The morphological parameters characterize the degree of valvular regurgitation.

3. The method for detecting valvular regurgitation according to claim 1, characterized in that, The morphological characteristic parameters include the neck width of the backflow jet; determining at least one morphological characteristic parameter of the backflow region based on the main extension direction includes: Obtain a plurality of straight lines that intersect the contour of the reflux region at least at two points, the plurality of straight lines being perpendicular to the main extension direction; The minimum distance between the two intersection points of each straight line is determined as the neck width of the reverse flow stream.

4. The method for detecting valvular regurgitation according to claim 1, characterized in that, The morphological characteristic parameters include the proximal isokinetic surface radius; the determination of at least one morphological characteristic parameter of the backflow region based on the main extension direction includes: Determine the location of the heart valves based on ultrasound images of the heart; Project several edge points of the reflux region onto the main extension direction and determine several corresponding projection values; The minimum bounding circle is fitted based on the set of edge points whose projection values ​​fall within the preset range, and the radius of the near-end isodynamic surface is determined based on the radius of the minimum bounding circle.

5. The method for detecting valvular regurgitation according to claim 4, characterized in that, The method includes: Determine the maximum and minimum projection values ​​among a set of projection values; The preset range is determined based on the maximum projection value and the minimum projection value; The preset range includes a first boundary value and a second boundary value, wherein the first boundary value is less than the second boundary value; the first boundary value is equal to the difference between the maximum projection value and the first intermediate value, the first intermediate value is equal to the product of the preset threshold and the difference between the maximum projection value and the minimum projection value, and the second boundary value is the maximum projection value.

6. The method for detecting valvular regurgitation according to claim 1, characterized in that, The method further includes assessing the severity of cardiac valvular regurgitation based on the morphological characteristic parameters and the area of ​​the valvular regurgitation region.

7. The method for detecting valvular regurgitation according to claim 6, characterized in that, The method further includes: Obtain the total number of pixels in the reflux region and the pixel area of ​​each pixel; The area of ​​the valve reflux region is determined by multiplying the total number by the pixel area.

8. The method for detecting valvular regurgitation according to claim 1, characterized in that, The step of performing principal component analysis on the regurgitation region to determine the principal extension direction of the corresponding valvular regurgitation includes: Determine the coordinate information of several pixels within the reflux region, and construct a covariance matrix based on the coordinate information; The main extension direction of the corresponding heart valve regurgitation is determined based on the largest eigenvalue corresponding to the covariance matrix.

9. The method for detecting valvular regurgitation according to claim 1, characterized in that, The process of identifying regurgitation regions caused by structural or functional abnormalities of the heart valves includes: Obtain ultrasound images of the heart and identify several areas of abnormal blood flow. Determine whether the location information of each abnormal blood flow region meets the first condition and whether the corresponding morphological information meets the second condition; If so, then the corresponding abnormal sequence region is identified as a regurgitation region caused by structural or functional abnormalities of the heart valve.

10. The method for detecting valvular regurgitation according to claim 9, characterized in that, The determination of whether the location information of each abnormal blood flow region meets the first condition and whether the corresponding morphological information meets the second condition includes: The corresponding atria and ventricles are determined based on the echocardiogram images. Determine whether the abnormal blood flow area is located at the junction of the atrium and ventricle; If so, the abnormal blood flow area is determined to meet the first condition.

11. The method for detecting valvular regurgitation according to claim 9, characterized in that, The determination that the abnormal blood flow region meets the first condition includes: The corresponding heart valve orifice is determined based on the ultrasound images of the heart. Determine whether the distance between the abnormal blood flow area and the heart valve orifice meets a first preset value; If so, the current location information of the abnormal blood flow area is determined to meet the first condition.

12. The method for detecting valvular regurgitation according to claim 9, characterized in that, The determination of whether the location information of each abnormal blood flow region meets the first condition and whether the corresponding morphological information meets the second condition includes: Determine whether the abnormal blood flow region is a connected region; If so, then when the shape of the abnormal blood flow region meets the preset standard shape, and / or when the area of ​​the abnormal blood flow region meets the second preset value, it is determined that the abnormal blood flow region meets the second condition.

13. A cardiac valve regurgitation detection system, characterized in that, include: The input module is used to identify regurgitation regions caused by structural or functional abnormalities of the heart valves. The processing module is used to perform principal component analysis on the regurgitation region to determine the main extension direction of the corresponding heart valve regurgitation; The detection module is used to determine at least one morphological feature parameter of the reflux region based on the main extension direction; Used to determine the status information of heart valve regurgitation based on the morphological characteristic parameters.

14. A computer storage medium, comprising: At least one processor; A memory storing a computer program running on the processor, characterized in that the processor executes the program to perform the steps of the method for detecting valvular regurgitation as described in any one of claims 1 to 12.

Images