Method and apparatus for identifying human tissue characteristic parameters

Abstract

The present invention discloses a method and device for identifying human tissue characteristic parameters, relating to the technical field of image processing, and mainly aiming to solve the problem of poor accuracy in identifying human tissue characteristic parameters in the prior art. The method includes: obtaining reconstructed image data of a target human tissue, where the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data and second reconstructed image data obtained by reconstructing mask image data; generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; and identifying characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue. It is mainly used for the method of identifying human tissue characteristic parameters.

Description

Technical Field

[0001] The present invention relates to the field of image processing technologies, and in particular, to a method and device for identifying characteristic parameters of human tissues. Background Art

[0002] In the clinical diagnosis of diseases of the pulmonary trachea as human tissues, it is necessary to calculate parameters such as the lumen diameter and wall thickness of the bronchus from CT images, segment the bronchus in the lungs of the CT images, and thus segment the wall of the bronchus on the frontal section to determine various parameters of the bronchus.

[0003] Currently, the existing identification of various parameters of human tissues such as the bronchus based on CT images is usually automatic detection of trachea parameters, including automatically segmenting the wall based on image gray level. However, the wall segmentation method based on image gray level is greatly affected by CT image noise, and only segments based on bronchus-like ellipses, resulting in blurred inner and outer boundaries of the wall, missing segmentation targets, and thus affecting the identification accuracy of various parameters of different tracheas in the whole lung. Summary of the Invention

[0004] In view of this, the present invention provides a method and device for identifying characteristic parameters of human tissues, mainly aiming to solve the problem of poor accuracy in identifying existing characteristic parameters of human tissues.

[0005] According to one aspect of the present invention, a method for identifying characteristic parameters of human tissues is provided, including:

[0006] Obtaining reconstructed image data of a target human tissue, where the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data, and second reconstructed image data obtained by reconstructing mask image data;

[0007] Generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result;

[0008] Identifying characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue.

[0009] Further, the identifying characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue includes:

[0010] Determining the morphological characteristics of the target human tissue, where the morphological characteristics include an inner boundary of the tissue and an outer boundary of the tissue;

[0011] Parse the inner boundary coordinate points and outer boundary coordinate points corresponding to the morphological features from the image segmentation processing results;

[0012] Determine the characteristic parameters based on the average value of the distances between the inner boundary coordinate points and the outer boundary coordinate points.

[0013] Further, the generating of the seed point marking image data matching the second reconstructed image data includes:

[0014] Perform positioning marking processing on the second reconstructed image data based on the dilation algorithm to obtain seed point marking image data containing segmentation seed points, and the segmentation seed points are used to distinguish the tissue characteristics of the human tissue.

[0015] Further, before the generating of the seed point marking image data matching the second reconstructed image data, the method further includes:

[0016] Perform image enhancement processing on the first reconstructed image data based on a filtering sharpening operator, so as to perform segmentation processing based on the first reconstructed image data after the image enhancement processing.

[0017] Further, the performing of the segmentation processing on the first reconstructed image data and the seed point marking image data according to the classification characteristics of the target human tissue to obtain the image segmentation processing result includes:

[0018] Integrate the first reconstructed image data and the seed point marking image data based on the classification characteristics to obtain original segmentation image data, and the original segmentation image data is three-dimensional image data;

[0019] Perform segmentation processing on the original segmentation image data according to the random walk algorithm to obtain the image segmentation processing result.

[0020] Further, determine the scanning direction of the target center point of the target human tissue;

[0021] Perform CT scanning on the target human tissue based on the scanning direction to obtain original image data and mask image data;

[0022] Perform image reconstruction processing on the original image data and the mask image data based on the target center point to obtain the first reconstructed image data and the second reconstructed image data.

[0023] Further, the method further includes:

[0024] Obtain at least one characteristic parameter matching the position information and classification characteristics of the target center point;

[0025] Integrate the feature parameters based on the tissue morphology of the target human tissue to obtain the overall feature parameters corresponding to the overall human tissue.

[0026] According to another aspect of the present invention, there is provided an identification device for human tissue feature parameters, including:

[0027] An acquisition module for acquiring the reconstructed image data of the target human tissue, where the reconstructed image data includes the first reconstructed image data obtained by reconstructing the original image data and the second reconstructed image data obtained by reconstructing the mask image data;

[0028] A generation module for generating seed point marker image data that matches the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification features of the target human tissue to obtain an image segmentation processing result;

[0029] An identification module for identifying the feature parameters in the image segmentation processing result based on the morphological features of the target human tissue.

[0030] Further, the identification module includes:

[0031] A determination unit for determining the morphological features of the target human tissue, where the morphological features include the inner boundary of the tissue and the outer boundary of the tissue;

[0032] An analysis unit for analyzing the inner boundary coordinate points and outer boundary coordinate points corresponding to the morphological features from the image segmentation processing result;

[0033] A determination unit for determining the feature parameters based on the average value of the distances between the inner boundary coordinate points and the outer boundary coordinate points.

[0034] Further, the generation module is specifically configured to perform positioning and marking processing on the second reconstructed image data based on the dilation algorithm to obtain seed point marker image data including segmentation seed points, and the segmentation seed points are used to distinguish the tissue features of the human tissue.

[0035] Further, the device further includes:

[0036] A first processing module for performing image enhancement processing on the first reconstructed image data based on a filtering sharpening operator, so as to perform segmentation processing based on the first reconstructed image data after image enhancement processing.

[0037] Further, the generation module includes:

[0038] A first processing unit, configured to integrally process the first reconstructed image data and the seed point labeled image data based on classification features to obtain original segmentation image data, where the original segmentation image data is three-dimensional image data;

[0039] A second processing unit, configured to perform segmentation processing on the original segmentation image data according to a random walk algorithm to obtain an image segmentation processing result.

[0040] Further, the device further includes:

[0041] A determination module, configured to determine a scanning direction of a target center point of the target human tissue;

[0042] A scanning module, configured to perform CT scanning on the target human tissue based on the scanning direction to obtain original image data and mask image data;

[0043] A second processing module, configured to perform image reconstruction processing on the original image data and the mask image data based on the target center point to obtain first reconstructed image data and second reconstructed image data.

[0044] Further, the device further includes: an integration module,

[0045] The obtaining module is further configured to obtain at least one feature parameter that matches the position information of the target center point and the classification features;

[0046] The integration module is configured to integrate the feature parameters based on the tissue form of the target human tissue to obtain an overall feature parameter corresponding to the overall human tissue.

[0047] According to another aspect of the present invention, there is provided a storage medium storing at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the above-mentioned method for identifying human tissue feature parameters.

[0048] According to still another aspect of the present invention, there is provided a terminal, including: a processor, a memory, a communication interface, and a communication bus, and the processor, the memory, and the communication interface complete communication with each other through the communication bus;

[0049] The memory is configured to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the above-mentioned method for identifying human tissue feature parameters.

[0050] By means of the above technical solution, the technical solution provided by the embodiment of the present invention has at least the following advantages:

[0051] The present invention provides a method and apparatus for identifying human tissue characteristic parameters. Compared with the prior art, in the embodiments of the present invention, by obtaining reconstructed image data of a target human tissue, the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data and second reconstructed image data obtained by reconstructing mask image data; generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; identifying the characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue, realizing image segmentation processing with segmentation marking based on seed points, greatly reducing the blurring degree of the edges of the human tissue, avoiding the missing of the segmentation target, and thus improving the identification accuracy of the characteristic parameters of the human tissue.

[0052] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other objects, features and advantages of the present invention more obvious and understandable, the following specifically gives the specific embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0054] Figure 1 shows a flowchart of a method for identifying human tissue characteristic parameters provided by an embodiment of the present invention;

[0055] Figure 2 shows a schematic diagram of a bronchial image processing flow provided by an embodiment of the present invention;

[0056] Figure 3 shows a flowchart of another method for identifying human tissue characteristic parameters provided by an embodiment of the present invention;

[0057] Figure 4 shows a schematic diagram of the overall process of human tissue image processing provided by an embodiment of the present invention;

[0058] Figure 5 shows a schematic diagram of an image data processing flow provided by an embodiment of the present invention;

[0059] Figure 6 shows a flowchart of yet another method for identifying human tissue characteristic parameters provided by an embodiment of the present invention;

[0060] Figure 7 shows a flowchart of yet another method for identifying human tissue characteristic parameters provided by an embodiment of the present invention;

[0061] Figure 8 shows a block diagram of a device for identifying human tissue characteristic parameters provided by an embodiment of the present invention;

[0062] Figure 9 shows a schematic structural diagram of a terminal provided by an embodiment of the present invention. Detailed implementation manners

[0063] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art.

[0064] Since the identification of various parameters of the bronchus based on CT images is usually automatic detection of tracheal parameters, including automatically segmenting the tube wall using image gray scale, and the tube wall segmentation method based on image gray scale is greatly affected by CT image noise, and only segments the bronchus based on an approximately elliptical shape, resulting in blurred inner and outer boundaries of the tube wall and missing segmentation targets, thereby affecting the identification accuracy of various parameters of different tracheas in the whole lung, an embodiment of the present invention provides a method for identifying human tissue characteristic parameters, as Figure 1 shown, the method includes:

[0065] 101. Obtain reconstructed image data of the target human tissue.

[0066] In an embodiment of the present invention, since it is necessary to obtain image data of human tissue, the human tissue can be scanned and photographed based on a CT scanning method to obtain the original image data of the target human tissue, and then the mask image data is obtained, and image reconstruction is respectively performed to obtain the reconstructed image data for identifying the human tissue characteristic parameters. Wherein, the reconstructed image data includes first reconstructed image data obtained by reconstructing the original image data and second reconstructed image data obtained by reconstructing the mask image data. Specifically, in the process of reconstructing the original image data of the target human tissue, the target center point is used as the center of reconstruction. In an embodiment of the present invention, the human tissue can be tissues such as blood vessels and tracheas in the human body. Preferably, the target human tissue is the bronchus. Therefore, after obtaining the original image data, the target center point is preferably the skeleton center point, and thus image reconstruction is performed according to the skeleton center point to obtain the first reconstructed image data, as Figure 2Schematic diagram of the bronchial image processing flow shown. At the same time, the mask image data is determined based on the original image data, and the mask image data is a regional image formed based on the values of 1 or 0 assigned. Therefore, during the image reconstruction process, the target center point is also used as the center of reconstruction, that is, the mask image data is reconstructed according to the skeleton center point to obtain the second reconstructed image data.

[0067] 102. Generate seed point marker image data matching the second reconstructed image data, and perform segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result.

[0068] In the embodiment of the present invention, since the second reconstructed image data is reconstructed based on the mask image data, and the mask image data is a regional image formed based on the values of 1 or 0 assigned, seed point marker image data matching the second reconstructed image data can be generated to determine the semantic segmentation information of the human tissue located in the human organ. For example, determine the semantic segmentation information of the bronchus in the lung. At the same time, perform segmentation processing on the first reconstructed image data and the obtained seed marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result. For example, image segmentation processing can be performed based on a three-dimensional random walk algorithm to obtain an image segmentation processing result of a pixel probability map. Among them, during the segmentation processing, the random walk algorithm is based on the seed point marker image data as the probability of the type to which the image pixel belongs, so as to calculate the segmentation region probability map of the target human tissue. At the same time, perform multiple threshold iterations on the segmentation region probability map, select the optimal parameters of the algorithm, and obtain the final image segmentation processing result.

[0069] It should be noted that the random walk algorithm in the embodiment of the present invention is a global optimization solution algorithm that is not easily trapped in local minima. During the processing of the pixel probability map, the first reconstructed image data and the seed point marker image data are integrated into three-dimensional data according to the classification characteristics of the target human tissue and used as the input parameters of the random walk algorithm. Therefore, based on the random walk algorithm, the probability of the type to which the image pixel belongs can be directly judged from the seed point marker map to obtain an image segmentation processing result. Among them, the classification characteristics are the classification contents obtained by the target human tissue according to different organizational structures. For example, if the target human tissue is the bronchus, it is classified according to the organizational structure of the bronchus, such as the structure of bronchial segments, to obtain the classification of each bronchial segment. In the embodiment of the present invention, when judging and recognizing the image segmentation processing result of the bronchial wall based on the random walk algorithm, it is processed according to different classifications of each bronchial segment, so as to obtain the image segmentation processing results corresponding to different classified segments of the bronchus.

[0070] 103. Identify the feature parameters in the image segmentation processing result based on the morphological features of the target human tissue.

[0071] In the embodiments of the present invention, the feature parameters are morphological feature parameters of the target tissue features segmented from the target human tissue. For example, if the target human tissue is the bronchus, the target tissue features are the bronchial wall, lumen, etc., and the feature parameters are the thickness of the bronchial wall and the lumen diameter, so as to output the feature parameters in the image as a basis for the user to make identification reference. At this time, since the image data corresponds to the cross-section of the human tissue, when identifying based on the morphological features, it is determined according to the cross-sectional shape of the tissue. For example, the wall thickness in the segmentation processing result is identified according to a transverse scan point of the bronchus. The entire wall is an annular structure, so the thickness of the annular wall can be calculated through the inner and outer boundary position points. The embodiments of the present invention do not make specific limitations.

[0072] It should be noted that in the embodiments of the present invention, since different target human tissues have different morphological features, and the obtained image segmentation processing result is based on a three-dimensional random walk algorithm, when performing feature parameter identification, it is identified based on the two-dimensional morphological features of the target human tissue. For example, through two-dimensional morphological analysis of the image segmentation result of the wall, the morphological analysis application component embedded in the current execution end automatically identifies the inner and outer boundaries of the wall and the morphological feature parameters of the wall, including numerical values such as thickness and diameter. The embodiments of the present invention do not make specific limitations.

[0073] In another embodiment of the present invention, for further limitation, as Figure 3 shown, step 103 of identifying the feature parameters in the image segmentation processing result based on the morphological features of the target human tissue includes: 1031. Determine the morphological features of the target human tissue; 1032. Parse the inner boundary coordinate points and outer boundary coordinate points corresponding to the morphological features from the image segmentation processing result; 1033. Determine the feature parameters based on the average value of the distances between the inner boundary coordinate points and the outer boundary coordinate points.

[0074] To accurately identify the characteristic parameters, when identifying the characteristic parameters based on the morphological features in the image segmentation processing result, specifically, the inner boundary coordinate points and outer boundary coordinate points corresponding to the morphological features can be parsed from the image segmentation processing result, so as to calculate the average value of the distances between the coordinates, which is determined as the characteristic parameter. Among them, the morphological features include the inner boundary of the tissue and the outer boundary of the tissue. In the embodiments of the present invention, based on the morphological analysis application component, the inner boundary and outer boundary of the target human tissue corresponding to the characteristic parameters to be identified are determined. At the same time, the inner boundary coordinate points corresponding to the inner boundary of the tissue and the outer boundary coordinate points corresponding to the outer boundary of the tissue are parsed from the image segmentation processing result. Specifically, when determining the inner and outer boundary coordinate points, the morphological analysis application component can directly extract the contour coordinate points (C) of the human tissue, that is, including the inner boundary coordinate points C internal and the outermost boundary coordinate point C external .

[0075] It should be noted that in the embodiments of the present invention, the average value of the distances between the inner and outer contour points (such as the wall of the bronchus) of the human tissue can be calculated by the tissue thickness calculation formula:

[0076] Thickness = E(min(C internal - C external ))), where min(C internal - C external ) is the minimum value of the distance between the inner boundary coordinate point C internal and the outer boundary coordinate point C external . At the same time, since when determining the inner and outer boundary coordinate points, it is determined based on the cross-section of the human tissue, that is, the minimum value between the inner boundary coordinate point and the outer boundary coordinate point is a cross-section. In order to determine the tissue thickness of the complete human tissue, the average value of the minimum values corresponding to all cross-sections can be calculated, and E(min(C internal - C external )) is the average value of the minimum values between all inner boundary coordinate points and outer boundary coordinate points, so as to obtain the tissue thickness.

[0077] In another embodiment of the present invention, for further limitation, the step of generating the seed point marker image data matching the second reconstructed image data includes: performing positioning and marking processing on the second reconstructed image data based on the dilation algorithm to obtain the seed point marker image data including the segmentation seed points.

[0078] In order to use the seed point marked image data as the basis for processing the pixel probability map of the image segmentation region by the random walk algorithm, so as to obtain an accurate segmentation effect, corresponding to the generation of the seed point marked image data. Specifically, based on the dilation algorithm, the second reconstructed image data is subjected to positioning and marking processing, so as to obtain the seed point marked image data including the segmentation seed points with tissue features for distinguishing human tissues. Among them, the second reconstructed image data is obtained by reconstructing the mask image data. Since the mask image data is a region image formed based on the values of 1 or 0, the second reconstructed image data provides accurate human tissue position information, such as bronchial position information. Therefore, the dilation algorithm is used to perform positioning and marking processing on the second reconstructed image data. Specifically, in the embodiment of the present invention, the segmentation seed points at least include three types of seed points, such as background region seed points, human tissue seed points, and human tissue boundary seed points (or contour seed points), so as to segment human tissues according to three types.

[0079] It should be noted that the specific method for obtaining the seed point marked image data based on the dilation algorithm in the embodiment of the present invention is as follows: 1. Extract the boundary (contour) of the second reconstructed image data corresponding to the human tissue, and obtain the boundary points (contour points) of the human tissue, such as bronchial contour points, and the point value is configured as 1; 2. After dilating the second reconstructed image data corresponding to the human tissue by a pixels, extract the boundary (contour). At this time, select the boundary (contour) located within the edge of the human tissue, such as the contour located within the bronchial wall, and the point value is configured as 2; 3. After dilating the second reconstructed image data corresponding to the human tissue by b pixels, extract the boundary (contour). At this time, select the boundary (contour) located outside the edge of the human tissue, such as the contour located outside the bronchial wall and belonging to the background region, and the point value is configured as 3, completing the positioning and marking processing of the dilation algorithm. Among them, the dilation processing is performed 2 times according to the parameters a and b to obtain the seed point marked image data, and the parameters a and b can be optimized and selected multiple times to determine the optimal parameters, which are not specifically limited in the embodiment of the present invention.

[0080] In another embodiment of the present invention, for further limitation, before the step of generating the seed point marked image data matching the second reconstructed image data, the method further includes: performing image enhancement processing on the first reconstructed image data based on the filtering sharpening operator, so as to perform segmentation processing based on the first reconstructed image data after image enhancement processing.

[0081] To improve the accuracy of image processing for the original image data obtained by scanning and avoid distortion of image pixels, it is necessary to filter the first reconstructed image data obtained based on the original image data. Specifically, image enhancement processing is performed on the first reconstructed pre-data based on a filtering sharpening operator, so as to generate responses and inhibitory effects on different partial regions to be segmented. For example, a filtering sharpening operator is calculated through two-dimensional convolution operation, and enhanced image processing is performed on the wall region of the bronchus through the filtering sharpening operator, so as to generate a strong response near the wall boundary and an inhibitory effect in other regions, such as Figure 4 The overall flowchart of human tissue image processing shown

[0082] In another embodiment of the present invention, for further limitation, the steps include segmenting the first reconstructed image data and the seed point marked image data according to the classification characteristics of the target human tissue, and the obtained image segmentation processing results include: integrating the first reconstructed image data and the seed point marked image data based on the classification characteristics to obtain the original segmentation image data; and segmenting the original segmentation image data according to the random walk algorithm to obtain the image segmentation processing results.

[0083] To implement segmentation processing by combining the first reconstructed image data and the seed point marked image data to obtain an accurate segmented image, specifically, the first reconstructed image data and the seed point marked image data are integrated based on the classification characteristics to obtain the original segmentation image data, where the original segmentation image data is a three-dimensional image number, so as to generate a pixel probability map based on the three-dimensional random walk algorithm, that is, the image segmentation processing results after segmentation processing are obtained. In the embodiments of the present invention, the classification characteristics are the classification contents obtained by classifying the target human tissue according to different tissue structures. For example, if the target human tissue is the bronchus, then it is classified according to the tissue structure of the bronchus, such as the structure of bronchial segments, to obtain the classification of each bronchial segment. Therefore, in order to perform image segmentation processing based on the random walk algorithm, the first reconstructed image data and the seed point marked image data are integrated based on the classification characteristics. For example, the first reconstructed image data corresponding to each bronchial segment and the seed point marked image data are integrated respectively as the original segmentation image data corresponding to one bronchial segment, and the obtained image segmentation processing results of this bronchial segment are obtained by segmenting this original segmentation image data through the random walk algorithm.

[0084] In addition, such as Figure 5In the schematic diagram of the image data processing flow shown, in the recognition scenario of the bronchial wall thickness, after obtaining the wall probability map, thresholding processing can be performed to evaluate the segmentation result. If the evaluation result does not meet the preset standard, the thresholding processing is performed again to achieve parameter update and adjustment, thereby improving the accuracy of image segmentation. The embodiments of the present invention do not make specific limitations on the thresholding processing.

[0085] In another embodiment of the present invention, for further limitation, as Figure 6 shown, before step 101 of obtaining the reconstructed image data of the target human tissue, it further includes: 201. Determining the scanning direction of the target center point of the target human tissue; 202. Performing CT scanning on the target human tissue based on the scanning direction to obtain the original image data and the mask image data; 203. Performing image reconstruction processing on the original image data and the mask image data based on the target center point to obtain the first reconstructed image data and the second reconstructed image data.

[0086] Since the image data acquisition method for human tissue usually uses CT scanning, in order to improve the accuracy of image processing and the recognition accuracy, first, scanning is performed according to the target center point, and then image reconstruction processing is performed on the scanned image data. Among them, the target center point is the intersection point between the cross-section of the scanning direction and the human tissue. For example, for the bronchial tissue of the lungs, the target center point can be each point on the skeleton center line. Thus, taking this point as the scanning direction, CT scanning is performed to obtain the original image data of the target human tissue, and then the mask image data of the human tissue is obtained. Of course, the mask image data can be directly obtained based on the processing of the original image data, or can be directly processed during the scanning process. The embodiments of the present invention do not make specific limitations. After obtaining the original image data and the mask image data, image reconstruction processing is respectively performed to obtain the first reconstructed image data and the second reconstructed image data.

[0087] It should be noted that the first reconstructed image data and the second reconstructed image data are respectively image blocks of 100*100 pixels. At this time, in order to distinguish the reconstructed image data corresponding to different target center points, the classification features (such as bronchial segments or positions) to which the target center points belong are saved in the image blocks to obtain an image matrix, that is, the image matrix contains the 100*100 pixel reconstructed image blocks of the original image data and the mask image data, and the classification features (such as bronchial segments, corresponding tracheal level numbers, etc.) to which the center points belong, so that the classification features corresponding to the image data of different sequence CT scans are different. If the level divided according to the classification features is higher, the greater the influence of the volume effect. For example, for different bronchial levels, the higher the level, the thinner the wall, and the greater the influence of the volume effect. Among them, the classification features can distinguish the organizational structures of different human tissues. For the bronchial scenario, the bronchi can be divided according to segments, branches, tracheal levels, etc. to indicate different positions during scanning, and the wall thicknesses of the bronchi corresponding to different positions are different. For example, the main bronchus is at level 0, and each branch of the bronchial tree below increases sequentially, and the wall thickness of the corresponding bronchus becomes thinner and thinner.

[0088] In another embodiment of the present invention, for further limitation, as Figure 7 shown, it further includes: 301. Obtain at least one feature parameter that matches the position information and classification features of the target center point; 302. Integrate the feature parameters based on the tissue morphology of the target human tissue to obtain the overall feature parameter corresponding to the overall human tissue.

[0089] Since human tissues are three-dimensional and stereoscopic, and when identifying feature parameters, it is usually based on two-dimensional calculations. Therefore, in order to realize the identification of the feature parameters of the overall tissue, the position information of the target center point and the classification features are combined for integration. Among them, since the target center point is the intersection point between the human tissue and the skeleton center line, the feature parameters (such as tissue thickness) of all the intersection points distributed on the entire center line are determined by combining the classification features (such as specific organizational structures). Therefore, the overall feature parameter corresponding to the overall human tissue is obtained. For example, for the bronchial tree structure, the lumen diameter and the mean, maximum, and minimum values of the wall thickness of each level of the trachea in the tracheal tree are identified longitudinally.

[0090] An embodiment of the present invention provides a method for identifying human tissue characteristic parameters. Compared with the prior art, in the embodiment of the present invention, by obtaining the reconstructed image data of the target human tissue, the reconstructed image data includes the first reconstructed image data obtained by reconstructing the original image data, and the second reconstructed image data obtained by reconstructing the mask image data; generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; identifying the characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue, realizing image segmentation processing with segmentation marking based on seed points, greatly reducing the blurring degree of the edges of human tissue, avoiding the missing of segmentation targets, and thus improving the recognition accuracy of the characteristic parameters of human tissue.

[0091] Further, as an implementation of the method described above Figure 1 An embodiment of the present invention provides an apparatus for identifying human tissue characteristic parameters, as Figure 8 shown. The apparatus includes:

[0092] An acquisition module 41, configured to acquire the reconstructed image data of the target human tissue, where the reconstructed image data includes the first reconstructed image data obtained by reconstructing the original image data, and the second reconstructed image data obtained by reconstructing the mask image data;

[0093] A generation module 42, configured to generate seed point marker image data matching the second reconstructed image data, and perform segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result;

[0094] An identification module 43, configured to identify the characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue.

[0095] Further, the identification module includes:

[0096] A determination unit, configured to determine the morphological characteristics of the target human tissue, where the morphological characteristics include the inner boundary of the tissue and the outer boundary of the tissue;

[0097] An analysis unit, configured to analyze the inner boundary coordinate points and the outer boundary coordinate points corresponding to the morphological characteristics from the image segmentation processing result;

[0098] A determination unit, configured to determine the characteristic parameters based on the average value of the distances between the inner boundary coordinate points and the outer boundary coordinate points.

[0099] Further, the generating module is specifically configured to perform positioning and marking processing on the second reconstructed image data based on a dilation algorithm to obtain seed point marked image data containing segmentation seed points, and the segmentation seed points are used to distinguish tissue characteristics of the human tissue.

[0100] Further, the device further includes:

[0101] A first processing module, configured to perform image enhancement processing on the first reconstructed image data based on a filtering and sharpening operator, so as to perform segmentation processing based on the first reconstructed image data after the image enhancement processing.

[0102] Further, the generating module includes:

[0103] A first processing unit, configured to perform integration processing on the first reconstructed image data and the seed point marked image data based on classification features to obtain original segmentation image data, and the original segmentation image data is three-dimensional image data;

[0104] A second processing unit, configured to perform segmentation processing on the original segmentation image data according to a random walk algorithm to obtain an image segmentation processing result.

[0105] Further, the device further includes:

[0106] A determination module, configured to determine a scanning direction of a target center point of the target human tissue;

[0107] A scanning module, configured to perform CT scanning on the target human tissue based on the scanning direction to obtain original image data and mask image data;

[0108] A second processing module, configured to perform image reconstruction processing on the original image data and the mask image data based on the target center point to obtain first reconstructed image data and second reconstructed image data.

[0109] Further, the device further includes an integration module,

[0110] The obtaining module is further configured to obtain at least one feature parameter that matches the position information of the target center point and the classification features;

[0111] The integration module is configured to integrate the feature parameters based on the tissue form of the target human tissue to obtain overall feature parameters corresponding to the overall human tissue.

[0112] An embodiment of the present invention provides a recognition device for human tissue characteristic parameters. Compared with the prior art, in the embodiment of the present invention, by acquiring reconstructed image data of a target human tissue, the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data, and second reconstructed image data obtained by reconstructing mask image data; generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; identifying characteristic parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue, realizing image segmentation processing with segmentation marking based on seed points, greatly reducing the blur degree of the edge of the human tissue, avoiding the missing of the segmentation target, and thus improving the recognition accuracy of the characteristic parameters of the human tissue.

[0113] According to an embodiment of the present invention, there is provided a storage medium storing at least one executable instruction, and the computer executable instruction can execute the recognition method for human tissue characteristic parameters in any of the above method embodiments.

[0114] Figure 9 The schematic structural diagram of a terminal provided by an embodiment of the present invention is shown. The specific implementation of the terminal is not limited in the specific embodiment of the present invention.

[0115] As Figure 9 shown, the terminal may include: a processor 502, a communication interface 504, a memory 506, and a communication bus 508.

[0116] Wherein: the processor 502, the communication interface 504, and the memory 506 communicate with each other through the communication bus 508.

[0117] The communication interface 504 is used for communicating with network elements of other devices such as clients or other servers.

[0118] The processor 502 is configured to execute the program 510, and specifically can execute the relevant steps in the above embodiment of the recognition method for human tissue characteristic parameters.

[0119] Specifically, the program 510 may include program codes, and the program codes include computer operation instructions.

[0120] The processor 502 may be a central processing unit (CPU), or a specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. One or more processors included in the terminal may be of the same type, such as one or more CPUs; or may be of different types, such as one or more CPUs and one or more ASICs.

[0121] A memory 506 for storing a program 510. The memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

[0122] The program 510 is specifically configured to cause the processor 502 to perform the following operations:

[0123] Obtain reconstructed image data of a target human tissue, where the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data, and second reconstructed image data obtained by reconstructing mask image data;

[0124] Generate seed point marker image data that matches the second reconstructed image data, and perform segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result;

[0125] Identify feature parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue.

[0126] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program code executable by the computing device, so that they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules respectively, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0127] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for identifying human tissue characteristic parameters, characterized in that, Including: Obtaining reconstructed image data of a target human tissue, where the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data, and second reconstructed image data obtained by reconstructing mask image data. The target human tissue is a trachea or a blood vessel, and the mask image data is an image of a trachea or blood vessel region formed by assigning a value of 1 or 0 to the original image data; Generating seed point marker image data that matches the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; Identifying feature parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue; Among them, the performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result includes: Performing integration processing on the first reconstructed image data and the seed point marker image data based on classification characteristics to obtain original segmentation image data, where the original segmentation image data is three-dimensional image data; Performing segmentation processing on the original segmentation image data to obtain an image segmentation processing result.

2. The method according to claim 1, wherein The identifying feature parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue includes: Determining the morphological characteristics of the target human tissue, where the morphological characteristics include an inner boundary of the tissue and an outer boundary of the tissue; Parsing inner boundary coordinate points and outer boundary coordinate points corresponding to the morphological characteristics from the image segmentation processing result; Determining feature parameters based on the average value of the distances between the inner boundary coordinate points and the outer boundary coordinate points.

3. The method according to claim 1, characterized in that The generating seed point marker image data that matches the second reconstructed image data includes: Performing positioning and marking processing on the second reconstructed image data based on a dilation algorithm to obtain seed point marker image data containing segmentation seed points, where the segmentation seed points are used to distinguish the tissue characteristics of the human tissue.

4. The method according to claim 1, characterized in that, Before the generating seed point marker image data that matches the second reconstructed image data, the method further includes: Performing image enhancement processing on the first reconstructed image data based on a filtering sharpening operator, so as to perform segmentation processing based on the first reconstructed image data after image enhancement processing.

5. The method according to claim 1, wherein Before the obtaining reconstructed image data of the target human tissue, the method further includes: Determining the scanning direction of the target center point of the target human tissue; Performing CT scanning on the target human tissue based on the scanning direction to obtain original image data and mask image data; Performing image reconstruction processing on the original image data and the mask image data based on the target center point to obtain first reconstructed image data and second reconstructed image data.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Obtaining at least one feature parameter that matches the position information and classification characteristics of the target center point; Integrating the feature parameters based on the tissue morphology of the target human tissue to obtain an overall feature parameter corresponding to the overall human tissue.

7. An identification device for human tissue characteristic parameters, characterized in that, Including: An acquisition module for acquiring reconstructed image data of a target human tissue, where the reconstructed image data includes first reconstructed image data obtained by reconstructing original image data, and second reconstructed image data obtained by reconstructing mask image data. The target human tissue is a trachea or a blood vessel, and the mask image data is an image of a trachea or blood vessel region formed by assigning a value of 1 or 0 to the original image data; A generation module for generating seed point marker image data matching the second reconstructed image data, and performing segmentation processing on the first reconstructed image data and the seed point marker image data according to the classification characteristics of the target human tissue to obtain an image segmentation processing result; An identification module for identifying feature parameters in the image segmentation processing result based on the morphological characteristics of the target human tissue; The generation module includes: A first processing unit for integrally processing the first reconstructed image data and the seed point marker image data according to classification characteristics to obtain original segmentation image data, where the original segmentation image data is three-dimensional image data; A second processing unit for performing segmentation processing on the original segmentation image data to obtain an image segmentation processing result.

8. A storage medium storing at least one executable instruction, where the executable instruction causes a processor to perform operations corresponding to the method for identifying human tissue feature parameters according to any one of claims 1-6.

9. A terminal, comprising: A processor, a memory, a communication interface, and a communication bus, where the processor, the memory, and the communication interface complete communication with each other through the communication bus; The memory is used for storing at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the method for identifying human tissue feature parameters according to any one of claims 1-6.

Images