Ultrasound image-based method and classification system for identifying multiple injuries in emergency patients

Abstract

The present application relates to the technical field of ultrasonic image processing, and particularly relates to an emergency patient multiple injury identification method and classification system based on an ultrasonic image, wherein the method comprises: performing image analysis on a plurality of first images respectively to determine a key region indicating a pleural line included in each first image; distinguishing a plurality of second images corresponding to a same ultrasonic detection position in the plurality of first images, and analyzing pixel changes between the key region of each second image and the key region of an adjacent second image of the second image to obtain a sliding sign parameter value of each second image; and analyzing the sliding sign parameter value of each second image to obtain a chest cavity feature value corresponding to each second image, wherein a plurality of chest cavity feature values corresponding to the plurality of second images are key data forming a target patient condition identification result. The present application can improve the processing effect of a chest ultrasonic image.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic image processing, and in particular to an ultrasonic image-based multiple injury recognition method and classification system for emergency patients. Background Art

[0002] Polytrauma refers to injuries to at least two organs or anatomical parts of a patient caused by the same injury, with at least one injury seriously endangering the patient's life. Ultrasound testing can be used to screen key areas (such as the chest) in polytrauma patients, enabling rapid and accurate analysis and providing a basis for condition assessment.

[0003] During actual ultrasound examinations, some patients have a high BMI and have more subcutaneous fat in their chests due to obesity. This subcutaneous fat can seriously interfere with the propagation of ultrasonic energy. Conventional solutions such as image noise reduction cannot effectively suppress the problem of decreased clarity in chest ultrasound images. This makes the patient's condition information reflected in the chest ultrasound images inaccurate, and leads to poor results in the condition analysis based on chest ultrasound images, which is prone to missed detections and other problems.

[0004] That is to say, the existing technology has a poor processing effect on chest ultrasound images. Summary of the Invention

[0005] In order to solve the technical problem that the existing technology has poor processing effect on chest ultrasound images, the purpose of the present invention is to provide an ultrasound image-based method and classification system for multiple injuries in emergency patients. The technical solutions adopted are as follows:

[0006] In a first aspect, an embodiment of the present invention provides a method for identifying multiple injuries in emergency patients based on ultrasound images, the method comprising:

[0007] performing image analysis on each of the plurality of first images to determine a key area included in each of the first images, wherein the plurality of first images are a plurality of ultrasound images acquired when a target patient undergoes ultrasound examination, and the key area is used to indicate a pleural line of the target patient;

[0008] distinguishing a plurality of second images from the plurality of first images, wherein the plurality of second images are a plurality of first images corresponding to a same ultrasonic detection position;

[0009] Analyzing pixel changes between a key area of ​​each second image and key areas of adjacent second images in the plurality of second images to obtain a sliding sign parameter value for each second image, the sliding sign parameter value being used to indicate a degree of pleural sliding sign reflected by the corresponding second image;

[0010] The sliding sign parameter value of each second image is analyzed to obtain a chest feature value corresponding to each second image, wherein the chest feature values corresponding to the plurality of second images are key data for forming a disease identification result of the target patient.

[0011] In one embodiment, the image analysis on the plurality of first images respectively to determine a key region included in each first image comprises:

[0012] The plurality of first images are respectively subjected to image segmentation to obtain a plurality of highlighted regions included in each first image;

[0013] The skeleton morphology and pixel concentration trend of each highlighted region included in each first image are analyzed to obtain a pleural line similarity value of each highlighted region included in each first image;

[0014] The region depth and pixel dispersion trend of each highlighted region included in each first image are analyzed to obtain a pleural line difference value of each highlighted region included in each first image;

[0015] The pleural line similarity value and the pleural line difference value of each highlighted region included in each first image are used to obtain a pleural line probability value of each highlighted region included in each first image;

[0016] Among the plurality of highlighted regions included in each first image, the highlighted region with the largest pleural line probability value is determined as the key region included in each first image.

[0017] In one embodiment, the analysis of the skeleton morphology and pixel concentration trend of each highlighted region included in each first image to obtain the pleural line similarity value of each highlighted region included in each first image comprises:

[0018] The plurality of highlighted regions included in each first image are respectively subjected to skeleton processing to obtain a plurality of region skeletons corresponding to each first image;

[0019] The plurality of region skeletons corresponding to each first image are respectively subjected to smoothness analysis to obtain a plurality of skeleton smoothness values corresponding to each first image;

[0020] The plurality of highlighted regions included in each first image are respectively subjected to pixel mean value calculation to obtain a plurality of region pixel mean values corresponding to each first image, wherein the region pixel mean value is the mean value of a plurality of pixel values of a plurality of pixel points in the corresponding highlighted region;

[0021] The skeleton smoothness value and the region pixel mean value corresponding to each highlighted region included in each first image are used to obtain the pleural line similarity value of each highlighted region included in each first image.

[0022] wherein the pleural line similarity value of the kth highlight region is The calculation formula of the pleural line similarity value of the kth highlight region is:

[0023]

[0024] In the above formula, k is a positive integer less than or equal to the total number of the highlight regions included in the corresponding first image, represents the skeleton smoothness of the kth highlight region, represents the region pixel mean value of the kth highlight region.

[0025] In one embodiment, the analysis of the region depth and the pixel dispersion trend of each highlight region included in each first image to obtain the pleural line difference value of each highlight region included in each first image comprises:

[0026] analyzing the difference between the region depth and the reference depth of each highlight region included in each first image to obtain the depth difference coefficient of the kth highlight region included in each first image, wherein the reference depth is the rib depth of the target patient;

[0027] obtaining the region dispersion coefficient of each highlight region included in each first image, wherein the region dispersion coefficient is the dispersion coefficient of the pixel values of the plurality of pixel points in the corresponding highlight region;

[0028] obtaining the pleural line difference value of each highlight region included in each first image according to the depth difference coefficient and the region dispersion coefficient of each highlight region included in each first image;

[0029] wherein the depth difference coefficient of the kth highlight region is The calculation formula of the depth difference coefficient of the kth highlight region is:

[0030]

[0031] In the above formula, k is a positive integer less than or equal to the total number of the highlight regions included in the corresponding first image, represents the region depth of the kth highlight region, represents the reference depth;

[0032] The pleural line probability value of the kth highlight region The calculation formula of the pleural line probability value of the kth highlight region is:

[0033]

[0034] wherein, represents the region dispersion coefficient of the kth highlight region, and exp represents the exponential function with the natural constant as the base number, a pleural line similarity value of the kth highlight region, a pleural line difference value of the kth highlight region.

[0035] In an embodiment, the analyzing, in the plurality of second images, a pixel change between a key region of each second image and a key region of a neighboring second image of the key region of the each second image, to obtain a sliding sign parameter value of the each second image, comprises:

[0036] analyzing a difference of pleural line probability values between the key region of each second image and a key region of a neighboring second image of the key region of the each second image, to obtain a probability difference coefficient corresponding to the each second image;

[0037] clustering a plurality of pixel points included in the key region of each second image, to obtain a plurality of class clusters corresponding to the each second image;

[0038] analyzing a position difference between the plurality of class clusters corresponding to each second image and a plurality of class clusters corresponding to a neighboring second image of the each second image, to obtain a position difference coefficient corresponding to the each second image;

[0039] obtaining the sliding sign parameter value of the each second image according to the probability difference coefficient and the position difference coefficient corresponding to the each second image.

[0040] In an embodiment, the analyzing a position difference between the plurality of class clusters corresponding to each second image and a plurality of class clusters corresponding to a neighboring second image of the each second image, to obtain a position difference coefficient corresponding to the each second image, comprises:

[0041] obtaining a plurality of class cluster critical distances corresponding to each second image, wherein the class cluster critical distance is a distance between a center point of a corresponding class cluster and a region boundary line of a key region in which the corresponding class cluster is located;

[0042] obtaining a plurality of class cluster offset distances corresponding to each second image based on the plurality of class cluster critical distances corresponding to the each second image, wherein the class cluster offset distance is a difference value between the class cluster critical distance of a corresponding class cluster and a class cluster critical distance of an associated class cluster of the corresponding class cluster, and two class clusters indicating a same region in two neighboring second images are associated with each other;

[0043] determining a mean value of the plurality of class cluster offset distances corresponding to each second image as the position difference coefficient corresponding to the each second image;

[0044] wherein the sliding sign parameter value of the ith second image is calculated according to a formula:

[0045]

[0046] wherein i is a positive integer less than or equal to the total number of the second images, a pleural line probability value of a key region of the i th second image, a pleural line probability value of a key region of the i+1 th second image, a probability difference coefficient corresponding to the i th second image, a q th cluster offset distance corresponding to the i th second image, a total number of clusters obtained by clustering in a key region of the i th second image, a position difference coefficient corresponding to the i th second image.

[0047] In an embodiment, the analyzing of the sliding sign parameter value of each second image obtains a chest cavity feature value corresponding to each second image, including:

[0048] The analyzing of the difference between the sliding sign parameter value of each second image and the sliding sign parameter value of the adjacent second image obtains a sliding sign difference coefficient corresponding to each second image;

[0049] The analyzing of the difference between the anechoic region area of each second image and the anechoic region area of the adjacent second image obtains an area difference coefficient corresponding to each second image;

[0050] The chest cavity feature value corresponding to each second image is obtained according to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image.

[0051] In an embodiment, the analyzing of the difference between the anechoic region area of each second image and the anechoic region area of the adjacent second image obtains an area difference coefficient corresponding to each second image, including:

[0052] The anechoic region of each second image is obtained, wherein the anechoic region is a region located below the corresponding key region and having a mean value of pixel values of a plurality of pixel points in the region less than or equal to a set threshold value;

[0053] The number of pixel points included in the anechoic region of each second image is counted to obtain an anechoic region area of each second image;

[0054] The difference between the anechoic region area of each second image and the anechoic region area of the adjacent second image is calculated to obtain an area difference value corresponding to each second image;

[0055] The absolute value of the area difference value corresponding to each second image is determined as the area difference coefficient corresponding to each second image;

[0056] wherein the area difference coefficient corresponding to the i-th second image is may be expressed as:

[0057]

[0058] wherein i is a positive integer less than or equal to the total number of the plurality of second images, denotes the total number of anechoic regions included in the i-th second image, denotes the total number of anechoic regions included in the i+1-th second image, denotes the area of the j-th anechoic region in the i-th second image, denotes the area of the j-th anechoic region in the i+1-th second image.

[0059] In one embodiment, the obtaining, according to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image, of a chest feature value corresponding to each second image comprises:

[0060] obtaining, according to the sliding sign difference coefficient corresponding to each second image, of a sliding sign similarity coefficient corresponding to each second image, wherein the sum of the sliding sign similarity coefficient and the corresponding sliding sign difference coefficient is 1;

[0061] calculating the product of the sliding sign similarity coefficient corresponding to each second image and the corresponding area difference coefficient to obtain a chest feature value corresponding to each second image.

[0062] In a second aspect, another embodiment of the present application provides an emergency patient multiple injury classification system based on ultrasound images, the system comprising:

[0063] a region detection module configured to perform image analysis on a plurality of first images respectively to determine a key region included in each first image, wherein the plurality of first images are a plurality of ultrasound images collected during ultrasound detection of a target patient, and the key region is used to indicate a pleural line of the target patient;

[0064] an image distinguishing module configured to distinguish a plurality of second images from the plurality of first images, wherein the plurality of second images are a plurality of first images corresponding to the same ultrasound detection position;

[0065] a sliding sign detection module configured to analyze, in the plurality of second images, pixel changes between the key region of each second image and the key region of an adjacent second image of the second image to obtain a sliding sign parameter value of each second image, wherein the sliding sign parameter value is used to represent a degree of a pleural sliding sign reflected by the corresponding second image;

[0066] ​​The sliding sign analysis module is used to analyze the sliding sign parameter value of each second image to obtain the chest cavity feature value corresponding to each second image, wherein the multiple chest cavity feature values ​​corresponding to the multiple second images are key data for forming the disease identification result of the target patient.

[0067] In a third aspect, another embodiment of the present invention further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the method described in the first aspect when executed by the processor.

[0068] In a fourth aspect, another embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0069] The present invention has the following beneficial effects:

[0070] The present invention first performs image analysis on multiple first images separately to determine the key area indicating the pleural line in each first image, so as to avoid image interference caused by subcutaneous fat by accurately locating the position of the pleural line, and then identifies multiple second images corresponding to the same ultrasound detection position from the multiple first images, and by analyzing the pixel changes between the key area of ​​each second image and the key area of ​​its adjacent second image, it is identified whether the target patient shows pleural sliding sign in the detection period corresponding to the multiple second images, and the chest cavity characteristic value of each second image is obtained based on this analysis, that is, the key data that can be used to accurately reflect the chest cavity physiological characteristics of the target patient is obtained, so as to assist in identifying the chest cavity disease of the target patient based on the obtained key data. Compared with conventional measures such as image denoising, the present invention extracts key data from multiple ultrasound images to reflect the chest cavity physiological characteristics of the target patient through multiple chest cavity characteristic values ​​rather than ultrasound images, which can effectively avoid various types of image noise existing in ultrasound images under complex working conditions and improve the image processing effect obtained by ultrasound images. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0072] Figure 1 A schematic flow chart of a method for identifying multiple injuries in emergency patients based on ultrasound images provided by one embodiment of the present invention;

[0073] Figure 2 A schematic structural diagram of an ultrasound image-based multiple injury classification system for emergency patients provided by one embodiment of the present invention;

[0074] Figure 3 The present invention provides a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0075] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of an ultrasound-based polytrauma identification method and classification system for emergency patients proposed by the present invention. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0076] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0077] The following describes in detail a method for identifying multiple injuries in emergency patients based on ultrasound images provided by the present invention with reference to the accompanying drawings.

[0078] This paper proposes a method for identifying multiple injuries in emergency patients based on ultrasound images. Figure 1 , which shows a schematic flow chart of a method for identifying multiple injuries of emergency patients based on ultrasound images provided by one embodiment of the present invention, the method comprising the following steps:

[0079] Step S1: performing image analysis on a plurality of first images respectively to determine a key area included in each first image.

[0080] The multiple first images are multiple ultrasound images collected when the target patient undergoes ultrasound testing, and the key area is used to indicate the pleural line of the target patient.

[0081] The above-mentioned target patient can be understood as any patient who undergoes ultrasound examination.

[0082] The pleural line, as used in this document, refers to the hyperechoic, bright line formed by the parietal and visceral pleura on ultrasound images (appearing as a smooth white line on ultrasound images). Under normal circumstances (meaning, when the patient has no chest or lung disease), only one pleural line is visible, located below the patient's ribs.

[0083] Exemplarily, the process of acquiring the multiple first images may be:

[0084] A low-frequency (2-5 MHz) convex array probe is used to detect the internal chest tissue of the target patient to obtain multiple original ultrasound images. It should be understood that the multiple original ultrasound images include not only several original ultrasound images corresponding to the same detection position, but also several original ultrasound images corresponding to different detection positions.

[0085] Noise reduction processing is performed on the multiple original ultrasound images to obtain multiple first images, wherein the multiple original ultrasound images and the multiple first images have corresponding meanings, and the noise reduction processing can be mean filtering processing, Gaussian filtering processing, wavelet transform processing, etc.

[0086] Compared with the solid and hollow organs in the patient's thoracic cavity, the pleural line has significant image features, is less affected by noise, is closely related to the position of the patient's ribs, is easy to identify, and can effectively indicate the actual movement of the patient's thoracic cavity. Therefore, selecting the area where the pleural line is located as the key area for relevant analysis can fully ensure the accuracy and practicality of the analysis results.

[0087] Furthermore, the performing image analysis on the plurality of first images to determine the key areas included in each first image includes:

[0088] Performing image segmentation on each of the plurality of first images to obtain a plurality of highlight regions included in each of the first images;

[0089] Analyzing the skeleton morphology and pixel concentration trend of each highlight region included in each first image to obtain a pleural line similarity value of each highlight region included in each first image;

[0090] analyzing a regional depth and a pixel discrete trend of each highlight region included in each first image to obtain a pleural line difference value of each highlight region included in each first image;

[0091] Obtaining a pleural line probability value for each highlighted area included in each first image according to the pleural line similarity value and the pleural line difference value for each highlighted area included in each first image;

[0092] Among the multiple highlight regions included in each first image, a highlight region with the largest pleural line probability value is determined as a key region included in each first image.

[0093] In one example, the maximum inter-class variance method can be used to complete the above-mentioned image segmentation operation to obtain multiple highlight areas included in each first image, wherein the pixel mean (i.e., grayscale mean) corresponding to the highlight area is greater than or equal to the pixel threshold corresponding to the maximum inter-class variance method (i.e., grayscale threshold, such as 245).

[0094] The pleural line similarity value is used to indicate the similarity between the corresponding highlighted area and the area where the pleural line is located.

[0095] The pleural line difference value is used to indicate the degree of difference between the corresponding highlighted area and the area where the pleural line is located.

[0096] The probability value of each highlighted area being a key area (i.e., the pleural line probability value) is comprehensively evaluated by combining the pleural line similarity value and pleural line difference value of each highlighted area. This can avoid the errors existing in single-dimensional evaluation and make the determined key areas more accurate and reliable.

[0097] Furthermore, the analyzing the skeleton morphology and pixel concentration trend of each highlight region included in each first image to obtain the pleural line similarity value of each highlight region included in each first image includes:

[0098] Performing skeleton processing on each of the multiple highlight regions included in each first image to obtain multiple region skeletons corresponding to each first image;

[0099] Performing smoothness analysis on the multiple regional skeletons corresponding to each first image to obtain the smoothness of the multiple skeletons corresponding to each first image;

[0100] Calculating pixel means for each of the multiple highlight regions included in each of the first images to obtain pixel means for each of the multiple regions corresponding to the first image, wherein the pixel mean for each region is an average of multiple pixel values ​​of multiple pixel points in the corresponding highlight region;

[0101] A pleural line similarity value of each highlight region included in each first image is obtained according to the skeleton smoothness and the regional pixel mean corresponding to each highlight region included in each first image.

[0102] Each regional skeleton corresponding to each first image is composed of a plurality of skeleton points connected in sequence.

[0103] In one example, the skeleton processing may be performed based on a morphological method (such as the Zhang-Suen algorithm) or a distance transformation method.

[0104] The process of obtaining the skeleton smoothness corresponding to a certain regional skeleton can be:

[0105] Slide the multiple skeleton points that constitute the skeleton of the region starting from the first skeleton point to obtain multiple sliding vectors, where each sliding vector is composed of two adjacent skeleton points, and each sliding vector is specifically: a vector from the previous skeleton point to the next skeleton point between the two adjacent skeleton points;

[0106] Among the multiple sliding vectors, the cosine similarity between adjacent sliding vectors is calculated respectively to obtain multiple vector similarities;

[0107] Calculate the mean of multiple vector similarities to obtain the skeleton smoothness corresponding to the skeleton in the region.

[0108] Among them, the greater the skeleton smoothness, the smoother the shape of the corresponding regional skeleton.

[0109] For example, the skeleton smoothness corresponding to the skeleton of the kth highlight region (the minimum value of k is 1, and the maximum value corresponds to the total number of highlight regions included in the first image) among the highlight regions is It can be expressed as:

[0110]

[0111] in, Indicates the number of multiple skeleton points of the corresponding regional skeleton, Represents the sliding vector from the u-th skeleton point to the u+1-th skeleton point among the multiple skeleton points of the skeleton in the region, Represents the sliding vector from the u+1th skeleton point to the u+2th skeleton point among the multiple skeleton points of the skeleton of the region, Indicates calculating the cosine similarity between two vectors (that is, the aforementioned vector similarity).

[0112] According to the skeleton smoothness and regional pixel mean corresponding to each highlight region included in each first image, the pleural line similarity value of each highlight region included in each first image is obtained, specifically:

[0113] The product of the skeleton smoothness corresponding to each highlight region included in each first image and the regional pixel mean is calculated to obtain a pleural line similarity value of each highlight region included in each first image.

[0114] For example, if the pleural line similarity value of the kth highlighted area is set to , the corresponding regional pixel mean is , then the pleural line similarity value of the kth highlighted area can be expressed as:

[0115]

[0116] Based on the above process, analyzing the similarity between the highlighted area and the area where the pleural line is located from the two aspects of regional morphology and regional pixel concentration trend can avoid the errors existing in single-dimensional evaluation and improve the accuracy of the determined pleural line similarity value. Specifically, since the pleural line is usually a smooth line with a certain width, the smoother the regional skeleton of the highlighted area, the more similar the highlighted area and the area where the pleural line are located are in morphology, and vice versa. Similarly, since the pleural line is usually highlighted in ultrasound images, the higher the pixel mean value of the highlighted area (manifested as a brighter area), the more similar the pixel concentration trend of the highlighted area and the area where the pleural line are located is, and vice versa.

[0117] Furthermore, the analyzing the regional depth and pixel discrete trend of each highlight region included in each first image to obtain the pleural line difference value of each highlight region included in each first image includes:

[0118] Analyzing a difference between a region depth of each highlighted region included in each first image and a reference depth to obtain a depth difference coefficient corresponding to each highlighted region included in each first image, wherein the reference depth is a rib depth of the target patient;

[0119] Obtaining a regional dispersion coefficient of each highlight area included in each first image, wherein the regional dispersion coefficient is a dispersion coefficient of a plurality of pixel values ​​corresponding to a plurality of pixel points in the highlight area;

[0120] A pleural line difference value of each highlight area included in each first image is obtained according to the depth difference coefficient and the area dispersion coefficient corresponding to each highlight area included in each first image.

[0121] The reference depth may be acquired through a plurality of first images by feature matching. Specifically, the reference depth acquisition process may be:

[0122] Based on a standard human chest image (obtained from a database, and having a number of key points corresponding to the ribs and a number of key points corresponding to the surface skin annotated in the standard human chest image), feature matching is performed on each of the plurality of first images to determine a skin region and a rib region corresponding to each of the first images;

[0123] Calculating the shortest distance between the skin area and the rib area corresponding to each first image to obtain an image distance corresponding to each first image;

[0124] An average of a plurality of image distances corresponding to the plurality of first images is calculated to obtain a reference depth.

[0125] The depth of the highlighted area is specifically the shortest distance between the corresponding skin area and the highlighted area.

[0126] The process of analyzing the difference between the regional depth of each highlight region included in each first image and the reference depth to obtain the depth difference coefficient corresponding to each highlight region included in each first image may be:

[0127] Calculating a difference between a region depth of each highlight region included in each first image and a reference depth to obtain a depth difference value of each highlight region included in each first image;

[0128] Calculating the absolute value of the depth difference of each highlight area included in each first image to obtain the absolute value of the depth difference of each highlight area included in each first image;

[0129] The ratio of the absolute value of the depth difference of each highlight area included in each first image to the reference depth is calculated to obtain a depth difference coefficient corresponding to each highlight area included in each first image.

[0130] For example, if the depth of the kth highlight area is set to , the reference depth is , then the depth difference coefficient corresponding to the highlighted area is It can be expressed as:

[0131]

[0132] Correspondingly, the pleural line probability value of the kth highlighted area is It can be expressed as:

[0133]

[0134] in, represents the inverse proportional function, Represents the regional discrete coefficient of the kth highlighted area; exp is an exponential function with a natural constant as the base.

[0135] The pleural line probability value is used to indicate the probability that the corresponding highlighted area is the area where the pleural line is located.

[0136] It should be noted that a depth difference threshold (such as 0.5) can be set, and when the depth difference coefficient corresponding to the highlighted area is greater than or equal to the depth difference threshold, the calculation of the pleural line probability value of the highlighted area can be skipped, and the highlighted area can be directly determined as an area where the non-pleural line is located.

[0137] In the above settings, based on the calculation of depth difference and the degree of pixel dispersion within the region, the difference between the highlighted area and the area where the pleural line is located is comprehensively evaluated from two aspects: image depth and the degree of uniform distribution of pixels within the region to ensure the accuracy of the calculated pleural line difference value.

[0138] Step S2: distinguishing a plurality of second images from the plurality of first images.

[0139] The multiple second images are multiple first images corresponding to the same ultrasonic detection position.

[0140] When performing a chest ultrasound examination on a target patient, in order to obtain ultrasound images of various positions of the target patient's chest cavity, an image will be collected for each position of the target patient's chest cavity (i.e., several original ultrasound images corresponding to different detection positions). This is the dynamic acquisition stage of the ultrasound examination. In order to accurately identify the movement status of the target patient's chest cavity, the ultrasound examination will continue to stay at a certain position of the target patient's chest cavity to collect multiple images at a certain position of the chest cavity (i.e., several original ultrasound images corresponding to the same detection position). This is the static acquisition stage of the ultrasound examination.

[0141] The aforementioned multiple second images can be understood as multiple first images corresponding to the static acquisition phase of ultrasonic testing.

[0142] In one example, the similarity between image features of two adjacent first images may be analyzed, and a plurality of consecutive first images having image feature similarity greater than a feature similarity threshold (eg, 0.7) may be determined as the plurality of second images.

[0143] Step S3: Analyze pixel changes between the key area of ​​each second image and the key areas of adjacent second images in the plurality of second images to obtain a sliding parameter value of each second image.

[0144] The sliding sign parameter value is used to indicate the degree of the pleural sliding sign reflected by the corresponding second image, that is, the strength of the dynamic change of the pleural line reflected by the corresponding second image.

[0145] The pleural line represents the contact interface between the visceral pleura and the parietal pleura. The visceral pleura covers the surface of the lungs, while the parietal pleura covers the inner wall of the chest cavity. When the target patient does not have chest or lung disease, the visceral pleura will slide relative to the parietal pleura as the target patient breathes, resulting in dynamic changes in the pleural line, which is the pleural sliding sign. When the target patient develops chest or lung disease, the above-mentioned pleural sliding sign will disappear or be very weak. Therefore, by analyzing the pixel changes in the key area indicating the pleural line between two adjacent ultrasound images, the strength of the pleural sliding sign reflected in the corresponding ultrasound image can be accurately identified.

[0146] Furthermore, in the plurality of second images, analyzing pixel changes between a key area of ​​each second image and key areas of adjacent second images to obtain a sliding sign parameter value of each second image includes:

[0147] Analyzing the difference in pleural line probability values ​​between each key area of ​​the second image and the key areas of adjacent second images to obtain a probability difference coefficient corresponding to each second image;

[0148] Clustering a plurality of pixel points included in a key area of ​​each second image to obtain a plurality of clusters corresponding to each second image;

[0149] Analyzing position differences between the multiple clusters corresponding to each second image and the multiple clusters corresponding to adjacent second images to obtain a position difference coefficient corresponding to each second image;

[0150] According to the probability difference coefficient and the position difference coefficient corresponding to each second image, the sliding characteristic parameter value of each second image is obtained.

[0151] In the above process, the distribution characteristics of each pixel point in the corresponding key area are extracted based on the clustering method, and the clusters obtained by clustering are used as the analysis objects of pixel changes, rather than the pixel points. This can reduce the calculation complexity of the sliding sign parameter value while ensuring the calculation accuracy of the sliding sign parameter value; among them, the introduction of the probability difference coefficient to participate in the calculation of the sliding sign parameter value can further analyze the differences between different key areas in adjacent ultrasound images from the probability dimension, so that the final calculated sliding sign parameter value is more accurate.

[0152] The above clustering operation can be completed based on the use of the K-means clustering algorithm, and the K value required for clustering can be obtained through the elbow method.

[0153] Specifically, analyzing the position differences between the multiple clusters corresponding to each second image and the multiple clusters corresponding to adjacent second images to obtain the position difference coefficient corresponding to each second image includes:

[0154] Acquire multiple cluster critical distances corresponding to each second image, wherein the cluster critical distance is the distance between the center point of the corresponding cluster and the region boundary line of the key region where the cluster is located;

[0155] Based on the multiple cluster critical distances corresponding to each second image, a multiple cluster offset distance corresponding to each second image is obtained, where the cluster offset distance is the difference between the cluster critical distance of the corresponding cluster and the cluster critical distance of its associated cluster, and two clusters indicating the same area in two adjacent second images are associated with each other;

[0156] The average of the multiple cluster offset distances corresponding to each second image is determined as the position difference coefficient corresponding to each second image.

[0157] For example, whether two clusters existing in two adjacent second images are associated may be determined based on the pixel mean of multiple pixel values ​​of multiple pixel points included in the cluster. The specific process is as follows:

[0158] If any two adjacent second images are defined as the first second image and the second second image, and the first second image is clustered to obtain multiple first clusters, and the second second image is clustered to obtain multiple second clusters;

[0159] Then the second cluster associated with each first cluster is: the second cluster among the plurality of second clusters, the second cluster having the smallest difference between the corresponding pixel mean and the pixel mean corresponding to the first cluster.

[0160] In one example, if the sliding parameter value of the i-th second image (the minimum value of i is 1 and the maximum value is the total number of multiple second images) is set to ,but It can be expressed as:

[0161]

[0162] in, is the pleural line probability value of the key area of ​​the i-th second image, is the pleural line probability value of the key area of ​​the i+1th second image, It can be understood as the probability difference coefficient corresponding to the i-th second image, is the offset distance of the qth cluster corresponding to the ith second image (specifically, the difference in cluster critical distance between the qth cluster corresponding to the ith second image and its associated cluster in the i+1th second image), is the total number of clusters obtained by clustering in the key interval of the i-th second image, It can be understood as the position difference coefficient corresponding to the i-th second image.

[0163] Step S4: Analyze the sliding sign parameter value of each second image to obtain the chest cavity characteristic value corresponding to each second image.

[0164] Among them, the multiple chest cavity feature values ​​corresponding to the multiple second images are key data for forming the disease identification result of the target patient.

[0165] The plurality of chest cavity characteristic values corresponding to the plurality of second images are key data for forming the identification result of the condition of the target patient, which can be understood as follows: the plurality of chest cavity characteristic values corresponding to the plurality of second images will affect the formation of the identification result of the condition of the target patient.

[0166] In the application, the plurality of chest cavity characteristic values can be used as an auxiliary index for a doctor to diagnose the chest disease of the target patient, so that the doctor can more accurately determine the chest disease of the target patient.

[0167] Further, a difference between the sliding sign parameter value of each second image and the sliding sign parameter value of the adjacent second image is analyzed to obtain a sliding sign difference coefficient corresponding to each second image.

[0168] A difference between the anechoic area of each second image and the anechoic area of the adjacent second image is analyzed to obtain an area difference coefficient corresponding to each second image.

[0169] According to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image, a chest cavity characteristic value corresponding to each second image is obtained.

[0170] In the ultrasonic detection process, due to the obstruction of the liquid, when the ultrasonic energy passes through the liquid (usually pleural effusion), an anechoic area (represented as a black area in the ultrasonic image) is formed, so the anechoic area can be understood as an area corresponding to the pleural effusion in the second image.

[0171] It should be understood that the anechoic area of each second image is specifically the total area of one or more anechoic areas existing below the key area (i.e., the pleural line) of the second image.

[0172] In the application, it is found that when the patient has a condition such as pneumothorax, although theoretically the patient's lung will not show a sliding sign, the sliding sign may still be detected in the actual detection process (such as in the early stage of pneumothorax), so the sliding sign parameter value of each second image cannot effectively represent the difference between different chest diseases.

[0173] To solve the above problem, the application calculates the difference between the sliding sign parameter values of adjacent second images to obtain a sliding sign difference coefficient corresponding to each second image, so as to extract the strong and weak change trend of the sliding sign in the corresponding detection period; and the area difference of the anechoic area between adjacent second images is calculated, so as to extract the change trend of the pleural effusion of the target patient in the corresponding detection period, so as to comprehensively form the chest cavity characteristic value which can comprehensively and accurately indicate the chest cavity condition of the target patient by combining the sliding sign and the pleural effusion.

[0174] Exemplarily, the process of analyzing the difference between the sliding sign parameter value of each second image and the sliding sign parameter value of the adjacent second image of each second image to obtain the sliding sign difference coefficient corresponding to each second image can be:

[0175] calculating the ratio between the sliding sign parameter value of each second image and the sliding sign parameter value of the adjacent second image of each second image to obtain the sliding sign difference coefficient corresponding to each second image.

[0176] Further, the process of analyzing the difference between the anechoic area of each second image and the anechoic area of the adjacent second image of each second image to obtain the area difference coefficient corresponding to each second image comprises:

[0177] obtaining the anechoic area of each second image, wherein the anechoic area is a region below the corresponding key region and the mean value of the pixel values of a plurality of pixel points in the region is less than or equal to a set threshold value;

[0178] counting the number of pixel points included in the anechoic area of each second image to obtain the anechoic area of each second image;

[0179] calculating the difference between the anechoic area of each second image and the anechoic area of the adjacent second image of each second image to obtain the area difference value corresponding to each second image;

[0180] determining the absolute value of the area difference value corresponding to each second image as the area difference coefficient corresponding to each second image.

[0181] Exemplarily, the set threshold value can be 5 or 10, and the value range of the pixel value is 0 (pure black) ~ 255 (pure white).

[0182] In one example, the area difference coefficient corresponding to the i-th second image can be represented as:

[0183]

[0184] wherein, represents the total number of anechoic regions included in the i-th second image, represents the total number of anechoic regions included in the i+1-th second image, represents the area of the j-th anechoic region in the i-th second image, represents the area of the j-th anechoic region in the i+1-th second image.

[0185] ​​Further, the chest feature value corresponding to each second image is obtained according to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image, and the chest feature value corresponding to each second image is obtained according to the sliding sign difference coefficient corresponding to each second image.

[0186] The sliding sign similarity coefficient corresponding to each second image is obtained according to the sliding sign difference coefficient corresponding to each second image, and the sum of the sliding sign similarity coefficient and the corresponding sliding sign difference coefficient is 1.

[0187] The product of the sliding sign similarity coefficient corresponding to each second image and the corresponding area difference coefficient is calculated to obtain the chest feature value corresponding to each second image.

[0188] The chest feature value corresponding to the i th second image is The chest feature value corresponding to the i th second image can be expressed as:

[0189]

[0190] Wherein, The sliding sign parameter value of the i th second image is represented as, The sliding sign parameter value of the i+1 th second image is represented as, The sliding sign difference coefficient corresponding to the i th second image is represented as, The sliding sign similarity coefficient corresponding to the i th second image is represented as.

[0191] In summary, the present application first performs image analysis on a plurality of first images to determine the key region indicating the pleural line in each first image, so as to avoid image interference caused by subcutaneous fat by accurately positioning the position of the pleural line, and then identifies a plurality of second images corresponding to the same ultrasonic detection position from the plurality of first images, and analyzes the pixel change between the key region of each second image and the key region of the adjacent second image to identify whether the target patient shows the pleural sliding sign in the detection period corresponding to the plurality of second images, and accordingly analyzes the chest feature value of each second image, that is, obtains the key data that can accurately reflect the physiological characteristics of the chest of the target patient, so as to assist in identifying the chest condition of the target patient according to the obtained key data. Compared with conventional measures such as image noise reduction, the present application extracts key data from a plurality of ultrasonic images to reflect the physiological characteristics of the chest of the target patient through a plurality of chest feature values instead of ultrasonic images, which can effectively avoid various image noises existing in ultrasonic images under complex working conditions and improve the image processing effect obtained by ultrasonic images.

[0192] The present application provides an emergency patient multiple injury classification system based on ultrasonic images, please refer to Figure 2 which shows the structure schematic diagram of an emergency patient multiple injury classification system 200 based on ultrasonic images provided by an embodiment of the present application, and the system comprises:

[0193] a region detection module 201 configured to perform image analysis on each of the plurality of first images to determine a key region included in each of the first images, wherein the plurality of first images are ultrasound images acquired during ultrasound examination of a target patient, and the key region is configured to indicate a pleural line of the target patient;

[0194] An image distinguishing module 202 is configured to distinguish a plurality of second images from the plurality of first images, wherein the plurality of second images are a plurality of first images corresponding to the same ultrasonic detection position;

[0195] a sliding sign detection module 203 for analyzing pixel variations between a key region of each second image and key regions of adjacent second images in the plurality of second images to obtain a sliding sign parameter value for each second image, wherein the sliding sign parameter value is used to indicate the degree of pleural sliding sign reflected by the corresponding second image;

[0196] The sliding sign analysis module 204 is used to analyze the sliding sign parameter value of each second image to obtain the chest cavity feature value corresponding to each second image, wherein the multiple chest cavity feature values ​​corresponding to the multiple second images are key data for forming the disease identification result of the target patient.

[0197] It should be noted that the system provided in the above embodiment is merely exemplified by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the above embodiment provides an ultrasound image-based emergency patient polytrauma classification system and an ultrasound image-based emergency patient polytrauma identification method embodiment, which are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0198] The embodiment of the present invention also provides an electronic device. Figure 3 , the electronic device may include a processor 301, a memory 302, and a program 3021 stored in the memory 302 and executable on the processor 301.

[0199] When the program 3021 is executed by the processor 301, it can achieve Figure 1 Any steps in the corresponding method embodiments and achieving the same beneficial effects will not be repeated here.

[0200] Those skilled in the art will appreciate that all or part of the steps of implementing the above-described embodiment method may be accomplished through hardware associated with program instructions, and the program may be stored in a readable medium.

[0201] The embodiment of the present invention further provides a readable storage medium, wherein the readable storage medium stores a computer program, and when the computer program is executed by a processor, the above Figure 1 Any steps in the corresponding method embodiments can achieve the same technical effects and will not be described again here to avoid repetition.

[0202] The computer-readable storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component.

[0203] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0204] The program code contained on the storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0205] Computer program code for performing the operations of the present invention may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or terminal. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0206] An embodiment of the present invention further provides a computer program product. When the computer program product is run on a computer, the computer is caused to execute the above-mentioned related steps to implement the method for identifying multiple injuries of emergency patients based on ultrasound images provided in the above embodiment.

[0207] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0208] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for identifying multiple injuries in emergency patients based on ultrasound images, characterized in that: The method comprises: performing image analysis on each of the plurality of first images to determine a key area included in each of the first images, wherein the plurality of first images are a plurality of ultrasound images acquired when a target patient undergoes ultrasound examination, and the key area is used to indicate a pleural line of the target patient; distinguishing a plurality of second images from the plurality of first images, wherein the plurality of second images are a plurality of first images corresponding to a same ultrasonic detection position; Analyzing pixel changes between a key area of ​​each second image and key areas of adjacent second images in the plurality of second images to obtain a sliding sign parameter value for each second image, the sliding sign parameter value being used to indicate a degree of pleural sliding sign reflected by the corresponding second image; Analyzing the sliding sign parameter value of each second image to obtain a chest cavity feature value corresponding to each second image, wherein the multiple chest cavity feature values ​​corresponding to the multiple second images are key data for forming a disease identification result of the target patient; The step of analyzing pixel changes between a key area of ​​each second image and key areas of adjacent second images in the plurality of second images to obtain a sliding sign parameter value of each second image includes: Analyzing the difference in pleural line probability values ​​between each key area of ​​the second image and the key areas of adjacent second images to obtain a probability difference coefficient corresponding to each second image; Clustering a plurality of pixel points included in a key area of ​​each second image to obtain a plurality of clusters corresponding to each second image; Analyzing position differences between the multiple clusters corresponding to each second image and the multiple clusters corresponding to adjacent second images to obtain a position difference coefficient corresponding to each second image; Obtaining a sliding sign parameter value of each second image according to a probability difference coefficient and a position difference coefficient corresponding to each second image; The analyzing the position difference between the multiple clusters corresponding to each second image and the multiple clusters corresponding to the adjacent second images to obtain the position difference coefficient corresponding to each second image includes: Acquire multiple cluster critical distances corresponding to each second image, wherein the cluster critical distance is the distance between the center point of the corresponding cluster and the region boundary line of the key region where the cluster is located; Based on the multiple cluster critical distances corresponding to each second image, a multiple cluster offset distance corresponding to each second image is obtained, where the cluster offset distance is the difference between the cluster critical distance of the corresponding cluster and the cluster critical distance of its associated cluster, and two clusters indicating the same area in two adjacent second images are associated with each other; Determine the mean of the multiple cluster offset distances corresponding to each second image as the position difference coefficient corresponding to each second image; Among them, the sliding parameter value of the i-th second image is The calculation formula is: Wherein, i is a positive integer less than or equal to the total number of the plurality of second images, is the pleural line probability value of the key area of ​​the i-th second image, is the pleural line probability value of the key area of ​​the i+1th second image, is the probability difference coefficient corresponding to the i-th second image, is the qth cluster offset distance corresponding to the i-th second image, is the total number of clusters obtained by clustering in the key interval of the i-th second image, is the position difference coefficient corresponding to the i-th second image; Analyzing the sliding sign parameter value of each second image to obtain the chest cavity characteristic value corresponding to each second image includes: Analyzing the difference between the sliding sign parameter value of each second image and the sliding sign parameter value of its adjacent second image to obtain a sliding sign difference coefficient corresponding to each second image; Analyzing the difference between the area of ​​the anechoic region of each second image and the area of ​​the anechoic region of its adjacent second image to obtain an area difference coefficient corresponding to each second image; Obtaining a chest cavity characteristic value corresponding to each second image according to a sliding sign difference coefficient and an area difference coefficient corresponding to each second image; Obtaining the chest cavity characteristic value corresponding to each second image according to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image includes: Obtaining a sliding sign similarity coefficient corresponding to each second image according to a sliding sign difference coefficient corresponding to each second image, wherein the sum of the sliding sign similarity coefficient and the corresponding sliding sign difference coefficient is 1; The product of the sliding sign similarity coefficient and the corresponding area difference coefficient corresponding to each second image is calculated to obtain the chest cavity characteristic value corresponding to each second image.

2. The method for identifying multiple injuries in emergency patients based on ultrasound images according to claim 1, characterized in that: The performing image analysis on each of the plurality of first images to determine a key area included in each of the first images includes: Performing image segmentation on each of the plurality of first images to obtain a plurality of highlight regions included in each of the first images; Analyzing the skeleton morphology and pixel concentration trend of each highlight region included in each first image to obtain a pleural line similarity value of each highlight region included in each first image; analyzing a regional depth and a pixel discrete trend of each highlight region included in each first image to obtain a pleural line difference value of each highlight region included in each first image; Obtaining a pleural line probability value for each highlighted area included in each first image according to the pleural line similarity value and the pleural line difference value for each highlighted area included in each first image; Among the multiple highlight regions included in each first image, a highlight region with the largest pleural line probability value is determined as a key region included in each first image.

3. The method for identifying multiple injuries in emergency patients based on ultrasound images according to claim 2, characterized in that: The analyzing the skeleton morphology and pixel concentration trend of each highlight region included in each first image to obtain the pleural line similarity value of each highlight region included in each first image includes: Performing skeleton processing on each of the multiple highlight regions included in each first image to obtain multiple region skeletons corresponding to each first image; Performing smoothness analysis on the multiple regional skeletons corresponding to each first image to obtain the smoothness of the multiple skeletons corresponding to each first image; Calculating pixel means for each of the multiple highlight regions included in each of the first images to obtain pixel means for each of the multiple regions corresponding to the first image, wherein the pixel means for each region is an average of multiple pixel values ​​of multiple pixel points in the corresponding highlight region; Obtaining a pleural line similarity value for each highlighted region included in each first image according to a skeleton smoothness and a regional pixel mean corresponding to each highlighted region included in each first image; Among them, the pleural line similarity value of the kth highlighted area is The calculation formula is: In the above formula, k is a positive integer less than or equal to the total number of the plurality of highlight regions included in the corresponding first image. Indicates the skeleton smoothness corresponding to the k-th highlighted area, Indicates the mean value of the pixels in the region corresponding to the k-th highlighted region.

4. The method for identifying multiple injuries in emergency patients based on ultrasound images according to claim 2, characterized in that: The analyzing the regional depth and pixel discrete trend of each highlight region included in each first image to obtain the pleural line difference value of each highlight region included in each first image includes: Analyzing a difference between a region depth of each highlighted region included in each first image and a reference depth to obtain a depth difference coefficient corresponding to each highlighted region included in each first image, wherein the reference depth is a rib depth of the target patient; Obtaining a regional dispersion coefficient of each highlight area included in each first image, wherein the regional dispersion coefficient is a dispersion coefficient of a plurality of pixel values ​​corresponding to a plurality of pixel points in the highlight area; Obtaining a pleural line difference value for each highlight area included in each first image according to a depth difference coefficient and an area dispersion coefficient corresponding to each highlight area included in each first image; Among them, the depth difference coefficient of the kth highlight area The calculation formula is: In the above formula, k is a positive integer less than or equal to the total number of the plurality of highlight regions included in the corresponding first image. represents the region depth of the kth highlighted region, Indicates the reference depth; The probability value of the pleural line of the kth highlighted area The calculation formula is: in, represents the regional discrete coefficient of the kth highlighted area, exp represents the exponential function with a natural constant as the base, represents the pleural line similarity value of the kth highlighted area, represents the pleural line difference value of the kth highlighted area.

5. The method for identifying multiple injuries in emergency patients based on ultrasound images according to claim 1, characterized in that: The analyzing the difference between the area of ​​the anechoic region of each second image and the area of ​​the anechoic region of its adjacent second image to obtain the area difference coefficient corresponding to each second image includes: Acquire an anechoic region of each second image, wherein the anechoic region is located below the corresponding key region and an average of pixel values ​​of a plurality of pixels in the region is less than or equal to a set threshold; Counting the number of pixels included in the anechoic region of each second image to obtain the area of ​​the anechoic region of each second image; Calculating the difference between the area of ​​the anechoic region of each second image and the area of ​​the anechoic region of its adjacent second image to obtain the area difference corresponding to each second image; determining an absolute value of the area difference corresponding to each second image as an area difference coefficient corresponding to each second image; Among them, the area difference coefficient corresponding to the i-th second image is Expressed as: Wherein, i is a positive integer less than or equal to the total number of the plurality of second images, represents the total number of anechoic regions included in the i-th second image, represents the total number of anechoic regions included in the (i+1)th second image, represents the first The area of ​​the anechoic region, represents the first image in the i+1th second image The area of ​​the anechoic region.

6. A system for classifying multiple injuries in emergency patients based on ultrasound images, characterized in that: The system comprises: a region detection module, configured to perform image analysis on each of the plurality of first images to determine a key region included in each of the first images, wherein the plurality of first images are a plurality of ultrasound images acquired when an ultrasound examination of a target patient is performed, and the key region is used to indicate a pleural line of the target patient; an image distinguishing module, configured to distinguish a plurality of second images from the plurality of first images, wherein the plurality of second images are a plurality of first images corresponding to a same ultrasonic detection position; a sliding sign detection module, configured to analyze pixel variations between a key area of ​​each second image and key areas of adjacent second images in the plurality of second images, to obtain a sliding sign parameter value for each second image, the sliding sign parameter value being used to indicate the degree of pleural sliding sign reflected by the corresponding second image; a sliding sign analysis module, configured to analyze the sliding sign parameter value of each second image to obtain a chest cavity feature value corresponding to each second image, wherein the multiple chest cavity feature values ​​corresponding to the multiple second images are key data for forming a disease identification result of the target patient; The step of analyzing pixel changes between a key area of ​​each second image and key areas of adjacent second images in the plurality of second images to obtain a sliding sign parameter value of each second image includes: Analyzing the difference in pleural line probability values ​​between each key area of ​​the second image and the key areas of adjacent second images to obtain a probability difference coefficient corresponding to each second image; Clustering a plurality of pixel points included in a key area of ​​each second image to obtain a plurality of clusters corresponding to each second image; Analyzing position differences between the multiple clusters corresponding to each second image and the multiple clusters corresponding to adjacent second images to obtain a position difference coefficient corresponding to each second image; Obtaining a sliding sign parameter value of each second image according to a probability difference coefficient and a position difference coefficient corresponding to each second image; The step of analyzing the position differences between the multiple clusters corresponding to each second image and the multiple clusters corresponding to adjacent second images to obtain the position difference coefficient corresponding to each second image includes: Acquire multiple cluster critical distances corresponding to each second image, wherein the cluster critical distance is the distance between the center point of the corresponding cluster and the region boundary line of the key region where the cluster is located; Based on the multiple cluster critical distances corresponding to each second image, a multiple cluster offset distance corresponding to each second image is obtained, where the cluster offset distance is the difference between the cluster critical distance of the corresponding cluster and the cluster critical distance of its associated cluster, and two clusters indicating the same area in two adjacent second images are associated with each other; Determine the mean of the multiple cluster offset distances corresponding to each second image as the position difference coefficient corresponding to each second image; Among them, the sliding parameter value of the i-th second image is The calculation formula is: Wherein, i is a positive integer less than or equal to the total number of the plurality of second images, is the pleural line probability value of the key area of ​​the i-th second image, is the pleural line probability value of the key area of ​​the i+1th second image, is the probability difference coefficient corresponding to the i-th second image, is the qth cluster offset distance corresponding to the i-th second image, is the total number of clusters obtained by clustering in the key interval of the i-th second image, is the position difference coefficient corresponding to the i-th second image; The step of analyzing the sliding sign parameter value of each second image to obtain the chest cavity characteristic value corresponding to each second image includes: Analyzing the difference between the sliding sign parameter value of each second image and the sliding sign parameter value of its adjacent second image to obtain a sliding sign difference coefficient corresponding to each second image; Analyzing the difference between the area of ​​the anechoic region of each second image and the area of ​​the anechoic region of its adjacent second image to obtain an area difference coefficient corresponding to each second image; Obtaining a chest cavity characteristic value corresponding to each second image according to a sliding sign difference coefficient and an area difference coefficient corresponding to each second image; The step of obtaining the chest cavity characteristic value corresponding to each second image according to the sliding sign difference coefficient and the area difference coefficient corresponding to each second image includes: Obtaining a sliding sign similarity coefficient corresponding to each second image according to a sliding sign difference coefficient corresponding to each second image, wherein the sum of the sliding sign similarity coefficient and the corresponding sliding sign difference coefficient is 1; The product of the sliding sign similarity coefficient and the corresponding area difference coefficient corresponding to each second image is calculated to obtain the chest cavity characteristic value corresponding to each second image.

Images