Ultrasonic auxiliary system for automatic recognition and diagnosis of emergency fracture

Abstract

The invention discloses an ultrasonic auxiliary system for automatic recognition and diagnosis of emergency fracture, and relates to the technical field of acoustic fracture detection. The auxiliary system comprises an ultrasonic image acquisition module used for acquiring ultrasonic image data of a fracture position of a patient; the processing module is used for processing the ultrasonic image data and outputting identification data; and the diagnosis decision module is used for outputting diagnosis data based on the identification data of the processing module. Through the arrangement of the region focusing analysis unit, resource allocation is optimized and calculated by gradually reducing the analysis range, the primary detection layer quickly eliminates intact regions and outputs primary fracture region marks, the intermediate classification layer performs soft tissue reaction quantitative analysis on the primary fracture regions and generates marked regions, and the advanced positioning layer processes the marked regions. The resource consumption is reduced in combination with sub-pixel-level analysis, and the architecture enables system resources to be quickly concentrated in a fracture area, so that the processing time efficiency of fracture recognition is reduced.

Description

Technical Field

[0001] This invention relates to the field of ultrasound fracture detection technology, specifically to an ultrasound-assisted system for automatic identification and diagnosis of emergency fractures. Background Technology

[0002] Emergency fractures are a common traumatic disease in clinical practice, and rapid and accurate diagnosis is crucial for treatment decisions and prognostic assessment.

[0003] For example, application number "CN115272176A", entitled "A Fracture Recognition Model and Method Based on Hip Joint DR Images", captures global contextual information through multi-head self-attention layer and multi-head attention layer in the fracture recognition model, thereby establishing long-distance dependence on the target and extracting more powerful features; and fully combines the information of the fracture area. The algorithm network proposed in the above patent has a significant improvement in the recall rate of the fracture area and a significant reduction in the number of false positives for fractures, thereby improving the accuracy of fracture recognition.

[0004] When identifying fractures at the hip joint, the aforementioned patent describes an ultrasound image of the hip joint containing an intact anatomical region. Since there is no fracture in this region, over-analysis of this region significantly increases the processing time for fracture identification. Therefore, an ultrasound-assisted system for automatic identification and diagnosis of emergency fractures has been invented. Summary of the Invention

[0005] The purpose of this invention is to provide an ultrasound-assisted system for automatic identification and diagnosis of emergency fractures, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an ultrasound-assisted system for automatic identification and diagnosis of emergency fractures, the system comprising:

[0007] An ultrasound image acquisition module used to obtain ultrasound image data of the patient's fracture location;

[0008] A processing module for processing ultrasound image data and outputting recognition data;

[0009] A diagnostic decision module for outputting diagnostic data based on the identification data of the processing module;

[0010] The processing module includes:

[0011] Skeletal topology unit; Ultrasound image data is processed using skeletal analysis methods to establish a mapping relationship between the ultrasound coordinate system and the anatomical coordinate system, and a topology map containing the mapping relationship is obtained;

[0012] Region Focused Analysis Unit: It is equipped with a progressive analysis architecture for gradually narrowing down and focusing the processing, and the progressive analysis architecture is equipped with a feedback mechanism to realize information interaction between each layer;

[0013] The progressive analysis architecture includes:

[0014] Primary detection layer: The topology map is analyzed by recognition methods to obtain data on the continuity of the bone cortex, the strong echo data of bone fragments, and the fragmentation edge feature data of the fracture, thereby obtaining a fracture sign dataset. Based on the location and mapping relationship of the continuity interruption, the strong echo area of ​​bone fragments, and the fragmentation edge feature of the fracture, the preliminary fracture area is obtained from the ultrasound image data.

[0015] Intermediate classification layer: The intensity of periosteal reaction in the initial fracture area and the extent of soft tissue injury are quantified by a quantification method to obtain a soft tissue reaction dataset. The soft tissue injury extent and the initial fracture area are combined to obtain the labeled area.

[0016] Advanced localization layer: The labeled area is analyzed using a sub-pixel level method. The sub-pixel level method includes analyzing the spatial arrangement structure of bone trabeculae and microfracture lines, calculating the angular deviation value of the normal vector of the bone trabeculae fracture surface, labeling the positions of bone trabeculae whose angular deviation value exceeds the deviation threshold, and obtaining the micro-damage area and micro-damage dataset.

[0017] The identification data is obtained by combining topology maps, fracture sign datasets, soft tissue reaction datasets, and micro-injury datasets.

[0018] Furthermore, the skeletal analysis method includes:

[0019] The skeletal topology unit stores a library of full-skeletal anatomical segmentation models and a skeletal anatomical reference database for realizing image data structure parsing;

[0020] Geometric feature vectors of the skeletal region in ultrasound image data are extracted using a whole-skeletal anatomical segmentation model, and skeletal anatomical landmarks in ultrasound image data are labeled.

[0021] The whole skeletal anatomy reference database includes a set of reference anatomical feature points, and the skeletal anatomical landmarks and reference anatomical feature points in the set of reference anatomical feature points are mapped to each other. Based on the mapping relationship, the corresponding skeletal anatomy reference data is obtained from the whole skeletal anatomy reference database.

[0022] A topological graph is constructed based on geometric feature vectors, mapping relationships, and ultrasound image data.

[0023] Furthermore, the feedback mechanism includes:

[0024] The micro-damage areas detected by the advanced localization layer are fed back to the primary detection layer. The primary detection layer performs a second scan of the micro-damage areas at the sub-pixel level to obtain a sub-pixel-level fracture sign dataset of the micro-damage areas. The sub-pixel-level fracture sign dataset replaces the fracture sign dataset.

[0025] The size of the initial fracture region in the primary detection layer is adjusted based on the periosteal thickening value quantified by the intermediate classification layer.

[0026] Furthermore, the diagnostic decision module includes processing the identified data using diagnostic methods to obtain a diagnostic report;

[0027] The diagnostic method includes: linking the AO classification database with the trauma scoring unit, identifying data input into the AO classification database and the trauma scoring unit and linking them, and generating a diagnostic report containing fracture type and displacement risk level.

[0028] Furthermore, the identification method includes performing cortical continuity analysis on the topological map, scanning the geometric features of the cortical surface, detecting linear fracture zones with lengths exceeding a length threshold, identifying the fragmented edge features of comminuted fractures based on curvature abrupt change points, and obtaining cortical continuity data, strong echo data of bone fragments, and fragmented edge feature data of fractures.

[0029] The establishment of the mapping relationship between the ultrasound coordinate system and the anatomical coordinate system includes:

[0030] A three-dimensional view of the overlapping bone structure in ultrasound image data is unfolded using a surface reconstruction algorithm, and an ultrasound coordinate system is established.

[0031] Based on skeletal analysis methods, anatomical reference data correlated with ultrasound image data is obtained. A surface reconstruction algorithm is used to unfold a three-dimensional view of the anatomical reference data and establish an anatomical coordinate system.

[0032] The mapping relationship between the ultrasound coordinate system and the anatomical coordinate system is established based on skeletal anatomical landmarks and reference anatomical feature points.

[0033] Furthermore, the geometric feature vector of the skeletal region includes the radius of curvature R and the cortical thickness gradient ΔT, and the anatomical landmarks include the epiphyseal plate growth line L and the articular surface boundary B.

[0034] Furthermore, the quantization method includes:

[0035] Measure the continuous value ΔH of the periosteal elevation height, quantify the hematoma area A, and mark the area where the soft tissue swelling is located;

[0036] Generate a tissue response feature vector with continuous values ​​ΔH and hematoma area A.

[0037] When ΔH is greater than the continuous value threshold, the periosteal region where ΔH is greater than the continuous value threshold is marked;

[0038] The extent of soft tissue injury is determined by combining the periosteal region and the area of ​​soft tissue swelling.

[0039] The extent of soft tissue injury, combined with Doppler blood flow signals, allows for the identification of indirect periosteal signs and reactions.

[0040] A soft tissue response dataset is obtained by combining the types of indirect periosteal signs and the extent of soft tissue injury.

[0041] Furthermore, the sub-pixel-level analysis includes:

[0042] Analyze the spatial arrangement of trabecular bone structure, and detect and calculate the deviation of the normal vector angle of microfracture lines and fracture surfaces;

[0043] Output micro-damage maps, establish spatial association between the coordinate set of the micro-damage core area and anatomical landmarks, and obtain micro-damage datasets by combining micro-fracture lines, fracture surface normal vector angle differences and spatial associations.

[0044] By setting up the region-focused analysis unit, which adopts a three-layer architecture of "primary detection → intermediate classification → advanced localization", the system optimizes the allocation of computing resources by gradually narrowing the analysis scope. The primary detection layer quickly excludes intact areas and outputs preliminary fracture area labels. The intermediate classification layer performs soft tissue reaction quantification analysis on the preliminary fracture area and generates labeled areas. The advanced localization layer processes the labeled areas and reduces resource consumption by combining sub-pixel-level analysis. This architecture enables the system resources to be quickly concentrated on the fracture area, reducing the processing time of fracture identification.

[0045] By establishing a mapping relationship between the ultrasound coordinate system and the anatomical coordinate system, and combining the automatic annotation of anatomical landmarks such as the epiphyseal plate and articular surfaces, the bone structure can be accurately located in complex images, reducing the obscuring of fracture signs by the acoustic shadowing area. The intermediate classification layer uses gray-level co-occurrence matrix to quantify the texture features of the periosteum region, and at the same time quantifies the contrast difference between the hematoma area and the surrounding tissue, effectively distinguishing fracture features from artifacts.

[0046] By using a surface reconstruction algorithm to unfold a three-dimensional view of overlapping skeletal structures, the system automatically identifies bone types and calls up standard anatomical models, dynamically annotating anatomical landmarks such as epiphyseal plates and articular surfaces. This technology solves the diagnostic blind spot problem caused by overlapping skeletal structures in traditional two-dimensional ultrasound, reduces the obscuring of fracture signs by the acoustic shadowing zone, and allows for observation of the spatial arrangement of fractures and the distribution of micro-damage, thereby improving the ability to identify complex fractures. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the system of the present invention;

[0048] Figure 2This is a schematic diagram of the processing module of the present invention;

[0049] Figure 3 This is a schematic diagram of the progressive analysis architecture of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] like Figure 1 - Figure 3 As shown, the present invention provides a technical solution: an ultrasound-assisted system for automatic identification and diagnosis of emergency fractures, the system comprising:

[0052] An ultrasound image acquisition module used to obtain ultrasound image data of the patient's fracture location;

[0053] A processing module for processing ultrasound image data and outputting recognition data;

[0054] A diagnostic decision module for outputting diagnostic data based on the identification data of the processing module;

[0055] The processing module includes:

[0056] Skeletal topology unit; Ultrasound image data is processed using skeletal analysis methods to establish a mapping relationship between the ultrasound coordinate system and the anatomical coordinate system, and a topology map containing the mapping relationship is obtained;

[0057] Region Focused Analysis Unit: It is equipped with a progressive analysis architecture for gradually narrowing down and focusing the processing. The progressive analysis architecture is equipped with a feedback mechanism to realize information interaction between each layer.

[0058] The progressive analysis architecture includes:

[0059] Primary detection layer: The topology map is analyzed by recognition methods to obtain data on the continuity of the bone cortex, the strong echo data of bone fragments, and the fragmentation edge feature data of the fracture, thereby obtaining a fracture sign dataset. Based on the location and mapping relationship of the continuity interruption, the strong echo area of ​​bone fragments, and the fragmentation edge feature of the fracture, the preliminary fracture area is obtained from the ultrasound image data.

[0060] Intermediate classification layer: The intensity of periosteal reaction in the initial fracture area and the extent of soft tissue injury are quantified by a quantification method to obtain a soft tissue reaction dataset. The soft tissue injury extent and the initial fracture area are combined to obtain the labeled area.

[0061] Advanced localization layer: The labeled area is analyzed using a sub-pixel level method. The sub-pixel level method includes analyzing the spatial arrangement structure of bone trabeculae and microfracture lines, calculating the angular deviation value of the normal vector of the bone trabeculae fracture surface, labeling the positions of bone trabeculae whose angular deviation value exceeds the deviation threshold, and obtaining the micro-damage area and micro-damage dataset.

[0062] The identification data is obtained by combining topology maps, fracture sign datasets, soft tissue reaction datasets, and micro-injury datasets.

[0063] Skeletal analysis methods include:

[0064] The skeletal topology unit stores a library of full-skeletal anatomical segmentation models and a skeletal anatomical reference database for realizing image data structure parsing;

[0065] Geometric feature vectors of the skeletal region in ultrasound image data are extracted using a whole-skeletal anatomical segmentation model, and skeletal anatomical landmarks in ultrasound image data are labeled.

[0066] The whole skeletal anatomy reference database includes a set of reference anatomical feature points, skeletal anatomical landmarks, and reference anatomical feature points in the set of reference anatomical feature points. Based on the correspondence, the corresponding skeletal anatomical reference data is obtained from the whole skeletal anatomy reference database.

[0067] A topological graph is constructed based on geometric feature vectors, mapping relationships, and ultrasound image data.

[0068] Feedback mechanisms include:

[0069] The micro-damage areas detected by the advanced localization layer are fed back to the primary detection layer. The primary detection layer performs a second scan of the micro-damage areas at the sub-pixel level to obtain a sub-pixel-level fracture sign dataset of the micro-damage areas. The sub-pixel-level fracture sign dataset replaces the fracture sign dataset.

[0070] The size of the initial fracture region in the primary detection layer is adjusted based on the periosteal thickening value quantified by the intermediate classification layer.

[0071] The diagnostic decision module includes processing the identified data using diagnostic methods to obtain a diagnostic report;

[0072] The diagnostic method includes: linking the AO classification database with the trauma scoring unit, identifying data input into the AO classification database and the trauma scoring unit and linking them, and generating a diagnostic report that includes fracture type and displacement risk level.

[0073] The identification method includes performing cortical continuity analysis on the topology map, scanning the geometric features of the cortical surface, detecting linear fracture zones with lengths exceeding a length threshold, identifying the fragmented edge features of comminuted fractures based on curvature abrupt change points, and obtaining cortical continuity data, strong echo data of bone fragments, and fragmented edge feature data of fractures.

[0074] Establishing the mapping relationship between the ultrasound coordinate system and the anatomical coordinate system includes:

[0075] A three-dimensional view of the overlapping bone structure in ultrasound image data is unfolded using a surface reconstruction algorithm, and an ultrasound coordinate system is established.

[0076] Based on skeletal analysis methods, anatomical reference data correlated with ultrasound image data is obtained. A surface reconstruction algorithm is used to unfold a three-dimensional view of the anatomical reference data and establish an anatomical coordinate system.

[0077] The mapping relationship between the ultrasound coordinate system and the anatomical coordinate system is established based on skeletal anatomical landmarks and reference anatomical feature points.

[0078] The geometric feature vectors of the skeletal region include the radius of curvature R and the cortical thickness gradient ΔT. The anatomical landmarks include the epiphyseal plate growth line L and the articular surface boundary B.

[0079] Quantification methods include:

[0080] Measure the continuous value ΔH of the periosteal elevation height, quantify the hematoma area A, and mark the area where the soft tissue swelling is located;

[0081] Generate a tissue response feature vector with continuous values ​​ΔH and hematoma area A.

[0082] When ΔH is greater than the continuous value threshold, the periosteal region where ΔH is greater than the continuous value threshold is marked;

[0083] The extent of soft tissue injury is determined by combining the periosteal region and the area of ​​soft tissue swelling.

[0084] The extent of soft tissue injury, combined with Doppler blood flow signals, allows for the identification of indirect periosteal signs and reactions.

[0085] A soft tissue response dataset is obtained by combining the types of indirect periosteal signs and the extent of soft tissue injury.

[0086] Subpixel-level analysis includes:

[0087] Analyze the spatial arrangement of trabecular bone structure, and detect and calculate the deviation of the normal vector angle of microfracture lines and fracture surfaces;

[0088] Output micro-damage maps, establish spatial association between the coordinate set of the micro-damage core area and anatomical landmarks, and obtain micro-damage datasets by combining micro-fracture lines, fracture surface normal vector angle differences and spatial associations.

[0089] The AO classification database is a fracture classification system established by the International Society for the Study of Internal Fixation. It uses an alphanumeric coding system to standardize the description of fracture location, morphology, and severity. The AO classification database is an existing database, and its specific content will not be described here. When writing a diagnostic report based on the AO classification database, the alphanumeric coding system in the AO classification database can be used to represent the fracture. At the same time, the diagnostic module uses natural language processing technology to automatically convert quantitative parameters into structured text, which includes diagnostic conclusions and damage mechanism analysis.

[0090] The inner cortex and both ends of the bone are composed of many irregular sheet-like or linear bone structures called trabeculae. Trabeculae are abundant in the metaphysis and, although they are continuous with the diaphysis in the inner cortex, they are relatively less abundant in the diaphysis. The trabeculae are arranged in accordance with the maximum stress and tension, and are interconnected in a loose, spongy manner, which is called cancellous bone.

[0091] The primary detection layer quickly excludes intact areas based on the fracture characteristics of the bone cortex; the intermediate classification layer performs secondary filtering through soft tissue reactions; and the advanced localization layer only processes the area where the fracture is located. By setting up the primary detection layer, intermediate classification layer, and advanced localization layer, the waste of computational resources is reduced.

[0092] The angle deviation of the fracture surface normal vector is calculated by spatial geometric analysis to quantitatively assess the degree of abnormal trabecular bone arrangement. The angle deviation of the trabecular bone fracture surface normal vector is analyzed. In ultrasound image data, the signs of fracture include direct and indirect signs. Direct signs of fracture include the breakage or displacement of the originally continuous, high-echoic cortical bone line in the ultrasound image, forming obvious interrupted gaps. Irregular strong echo or hypoechoic areas can be seen at the fracture ends, indicating bone separation. Indirect signs of fracture include thickening of the periosteum and enhanced echo at the fracture site, possibly accompanied by increased blood flow signal (visible on Doppler imaging). The appearance of anechoic or hypoechoic areas in the soft tissues around the fracture, with clear or blurred boundaries, indicates hemorrhage or effusion. The surrounding tissues show enhanced echo and blurred structure, accompanied by pain when the probe is pressed. When there is a fracture near the joint, anechoic or hypoechoic areas can be seen in the joint cavity, indicating increased joint fluid or hemorrhage.

[0093] When performing ultrasound detection on a fracture site, the obtained ultrasound image data may contain intact areas without signs of fracture. When automatic recognition is enabled, there may be a situation where this area needs to be identified and analyzed. Since there is no fracture, the processing and analysis will increase the computer's processing time.

[0094] Region focusing analysis narrows down the entire area, enabling focus on multiple fracture regions. This reduces the scope of subsequent identification and diagnosis from complete ultrasound image data, decreasing computer processing time. The module comprises a primary detection layer, an intermediate classification layer, and a high-level localization layer. The primary detection layer identifies the fracture location. These three layers are core components of the region focusing analysis module, working through a hierarchical detection and feedback mechanism to achieve closed-loop verification from coarse screening to precise diagnosis. Specific functions include: the primary detection layer, based on a topology map generated by the skeletal topology mapping subsystem, utilizes pre-annotated solutions... Preliminary screening is performed using anatomical landmarks (such as epiphyseal plates and articular surfaces) and direct signs (cortical bone interruption, bone fragments). An edge detection algorithm is used to scan the cortical bone surface to identify linear fracture bands longer than 1 mm and curvature abrupt change points (characteristics of comminuted fractures). The fracture probability map and coordinates of high-risk areas are output to provide an initial localization framework for subsequent analysis. In the primary detection layer, the geometric features of the cortical bone surface are scanned using an edge detection algorithm to identify linear fracture bands longer than a set threshold, such as 1 mm. Based on curvature abrupt change points, the fragmented edge features of comminuted fractures are analyzed, and areas with abnormally increased curvature are marked as suspicious fracture areas.

[0095] The intermediate classification layer performs clipping and focusing on high-risk areas output by the primary detection layer. It quantifies the periosteal elevation height ΔH and hematoma area A through threshold segmentation, and determines the periosteal reaction type (acute inflammation or chronic repair) by combining Doppler blood flow signals (resistance index RI). It generates a tissue reaction heatmap and transmits the ΔH value to the advanced localization layer to dynamically adjust the analysis intensity. Based on the preliminary fracture area marked by the primary detection layer, the intermediate classification layer analyzes the periosteal texture features of the preliminary fracture area through the gray-level co-occurrence matrix (GLCM), calculates parameters such as contrast and energy, and uses Otsu's automatic threshold segmentation technique to quantify the contrast difference between the hematoma area and the surrounding normal tissue in the preliminary fracture area.

[0096] Within the area labeled by the intermediate classification layer, the advanced localization layer uses a sub-pixel-level analysis algorithm to detect the deviation of the normal vector angle of the trabecular fracture surface (θ>15°), generates a micro-damage cloud map, and labels the coordinates of the micro-damage core area.

[0097] Simultaneously, a feedback mechanism triggers the primary detection layer to perform a secondary scan of the micro-damage coordinate area at the sub-pixel level using the detected micro-damage coordinates. This generates a fracture sign dataset that replaces the original fracture sign dataset, forming a closed loop of "preliminary screening → precise localization → secondary verification." This ensures that the fracture localization error is less than 1mm. The region focusing analysis unit employs a three-layer architecture: "primary detection → intermediate classification → advanced localization." By progressively narrowing the analysis scope, computational resource allocation is optimized. The primary detection layer quickly eliminates intact areas and outputs preliminary fracture area labels. The intermediate classification layer performs soft tissue reaction quantification analysis on the preliminary fracture area, generating labeled areas. The advanced localization layer processes the labeled areas. Combined with sub-pixel-level analysis, resource consumption is reduced. This architecture allows system resources to be quickly concentrated on the fracture area, reducing the processing time for fracture identification.

[0098] The Gray-Level Co-occurrence Matrix (GLCM) is a core tool for image texture analysis. By statistically analyzing the co-occurrence frequency of pixel gray-level pairs at specific distances and directions, it quantifies features such as roughness and directionality of image texture. In fracture detection, GLCM identifies abnormalities by analyzing changes in periosteal texture. In the intermediate classification layer, Otsu's method is used to quantify the contrast difference between hematoma area and normal tissue. Through automatic thresholding, it accurately identifies the boundary of the hematoma area. When the hematoma area is >5k square centimeters, the system automatically increases the classification priority of that area, triggering more refined analysis. k is affected by the size of human bones. Since bone size varies at different ages, the threshold for hematoma area is affected by the size of human bones. Significant differences in bone size mean that the same hematoma area may represent different degrees of severity in people of different ages. Therefore, k is set.

[0099] The regions marked in the intermediate classification layer include: periosteal reaction zone: periosteal elevation height ΔH: when ΔH > 2.5mm, the periosteal tissue in this area will be included in the soft tissue injury range; soft tissue injury zone: the hematoma area A is obtained by detection, the area where the soft tissue swelling is located, the swelling area is marked by the edge detection algorithm to help assess the degree of injury, the swelling area is included in the soft tissue injury range, the soft tissue injury range is corrected for the suspected fracture area, and the marked area is obtained.

[0100] The advanced localization layer employs a phase-based subpixel localization algorithm, performing fine scanning within the labeled regions of the intermediate classification layer. It combines Histogram of Oriented Gradients (HOG) with a subpixel edge detection algorithm to improve fracture line localization accuracy to the 0.1 mm level. Micro-damage cloud mapping visualizes the trabecular fracture surface distribution, assisting physicians in assessing fracture stability. A secondary scan of the micro-damage area at the subpixel level allows for a re-run of the primary detection layer at a high-pixel level. Histogram of Oriented Gradients (HOG) is a statistical method for image feature description; its core function is to capture the edge and shape features of objects by quantifying the gradient direction distribution of local image regions. Principal Component Analysis (PCA) is applied to calculate... The normal vector of the trabecular fracture surface is used to annotate the fracture surface when the angle deviation of the fracture surface is greater than the deviation threshold. In ultrasound fracture detection, PCA quantifies the angle deviation of the normal vector of the trabecular fracture surface to help determine the fracture type (such as oblique or spiral) and stability. It is a commonly used feature analysis tool in existing technologies. By establishing the mapping relationship between the ultrasound coordinate system and the anatomical coordinate system, and combining the automatic annotation of anatomical landmarks such as the epiphyseal plate and articular surfaces, the bone structure can be accurately located in complex images. The intermediate classification layer uses gray-level co-occurrence matrix to quantify the texture features of the periosteal region. Combined with Otsu's automatic threshold segmentation technology, the contrast difference between the hematoma area and the surrounding tissue is quantified, effectively distinguishing fracture features from artifacts.

[0101] Based on the quantification of periosteal elevation height in the intermediate classification layer, the size of the preliminary fracture area in the primary detection layer is adjusted to determine whether the periosteal membrane exists in the preliminary fracture area in the primary detection layer. If the periosteal membrane cannot be included, the parameters for generating the size of the preliminary fracture area in the primary detection layer will be adjusted so that it can include both the normal periosteal membrane and the periosteal membrane that has been elevated due to the fracture.

[0102] The ultrasound image acquisition module, as the input of the system, is responsible for acquiring the raw ultrasound image data of the patient's fracture site. The diagnostic decision module, as the output of the system, integrates the analysis results of the processing module and generates the final diagnostic report based on established medical standards (such as the AO classification database).

[0103] The coordinate system of an ultrasound image is a local coordinate system relative to the probe position, while the anatomical structure of the human body exists in a standardized three-dimensional anatomical coordinate system. Therefore, accurate mapping between the two is crucial. This system incorporates a full-skeletal anatomical segmentation model library and a full-skeletal anatomical reference database. The database stores a large amount of standardized skeletal anatomical data, including key anatomical feature points. Image feature points are automatically extracted from real-time acquired ultrasound images. Subsequently, by matching these feature points with the anatomical feature point set in the database, the system can identify the currently scanned bone type (e.g., femur, tibia) and its approximate location. Once the bone location is identified, the system calls the corresponding skeletal anatomical segmentation model. This model further extracts more detailed geometric feature vectors from the ultrasound image, establishing the mapping relationship between the ultrasound local coordinate system and the standard anatomical coordinate system.

[0104] The skeletal topology unit integrates all information into a topology map containing the distribution of signs. The topology map not only includes the type, orientation, and precise coordinate mapping of bones, but is then transmitted to the region focusing analysis unit, providing a key anatomical benchmark for its subsequent progressive analysis.

[0105] The primary detection layer outputs one or more preliminary fracture areas, thus narrowing the focus of analysis from the entire scan range to a few fracture areas (preliminary fracture areas). After locating the preliminary fracture areas, the intermediate classification layer intervenes, focusing on the surrounding tissue response caused by the fracture, i.e., the quantitative analysis of indirect signs. Soft tissue response is an indirect sign. This intermediate layer generates a tissue response feature vector containing parameters ΔH and A, and further precisely labels the areas based on these data, providing a more focused target for advanced analysis and obtaining the labeled areas of the intermediate classification layer.

[0106] The advanced localization layer performs subpixel-level analysis on the labeled areas of the intermediate classification layer. When the deviation of the normal vector angle of the fracture surface of the trabecular bone exceeds the deviation threshold, the area is labeled as a micro-damage area. The deviation threshold is 13°-15°. The advanced localization layer finally outputs a micro-damage map and establishes a spatial relationship between the coordinate set of the micro-damage core area and the anatomical landmarks labeled by the bone topology unit. This allows micro-damage to be accurately located on the anatomical structure.

[0107] After the advanced localization layer detects the micro-damage area, its coordinate set (micro-damage coordinate feedback) is fed back to the primary detection layer for secondary scanning and processing of this specific micro-region, enabling a review of the micro-damage area to ensure accuracy and obtain a dataset of fracture signs in the micro-damage area. The periosteal thickening value ΔH quantified by the intermediate classification layer is used to expand the preliminary fracture area range defined by the primary detection layer. Periosteal reaction is a physiological response of the bone to stimuli such as injury, inflammation, or tumors. Its types can be classified according to pathological mechanisms, morphological characteristics, and clinical significance. Periosteal reaction is a current technology.

[0108] The processing module comprises a skeletal topology unit for anatomical localization and a region-focusing analysis unit for lesion analysis. The latter employs a three-layer progressive analysis architecture (primary detection, intermediate classification, and advanced localization) and introduces a closed-loop feedback mechanism to achieve step-by-step and precise identification from macroscopic cortical fractures to microscopic trabecular bone damage. Combined with sign-anatomical correlation analysis (generating a sign distribution topology map through geometric structural feature vectors of skeletal regions, such as radius of curvature R and cortical thickness gradient ΔT), the system realizes the correlation between direct signs (such as cortical bone interruption) and indirect signs (such as periosteal thickening). This integrated diagnostic approach significantly improves the ability to identify complex fractures (such as comminuted fractures and intra-articular fractures). By using a surface reconstruction algorithm to unfold a three-dimensional view of overlapping skeletal structures, it automatically identifies bone types and calls up standard anatomical models, dynamically annotating anatomical landmarks such as epiphyseal plates and articular surfaces. This technology solves the diagnostic blind spot problem caused by overlapping skeletal structures in traditional two-dimensional ultrasound, reduces the obscuring of fracture signs by the acoustic shadowing zone, and facilitates the observation of the spatial arrangement of fractures and the distribution of micro-damage. Combined with sign-anatomical correlation analysis, the system realizes the correlation diagnosis of direct and indirect signs, significantly improving the ability to identify complex fractures.

[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. An ultrasound-assisted system for automatic identification and diagnosis of emergency fractures, characterized in that, The auxiliary system includes: An ultrasound image acquisition module used to obtain ultrasound image data of the patient's fracture location; A processing module for processing ultrasound image data and outputting recognition data; A diagnostic decision module for outputting diagnostic data based on the identification data of the processing module; The processing module includes: Skeletal topology unit; Ultrasound image data is processed using skeletal analysis methods to establish a mapping relationship between the ultrasound coordinate system and the anatomical coordinate system, and a topology map containing the mapping relationship is obtained; Region Focused Analysis Unit: It is equipped with a progressive analysis architecture for gradually narrowing down and focusing the processing, and the progressive analysis architecture is equipped with a feedback mechanism to realize information interaction between each layer; The progressive analysis architecture includes: Primary detection layer: The topology map is analyzed by recognition methods to obtain data on the continuity of the bone cortex, the strong echo data of bone fragments, and the fragmentation edge feature data of the fracture, thereby obtaining a fracture sign dataset. Based on the location and mapping relationship of the continuity interruption, the strong echo area of ​​bone fragments, and the fragmentation edge feature of the fracture, the preliminary fracture area is obtained from the ultrasound image data. Intermediate classification layer: The intensity of periosteal reaction in the initial fracture area and the extent of soft tissue injury are quantified by a quantification method to obtain a soft tissue reaction dataset. The soft tissue injury extent and the initial fracture area are combined to obtain the labeled area. Advanced localization layer: The labeled area is analyzed using a sub-pixel level method. The sub-pixel level method includes analyzing the spatial arrangement structure of bone trabeculae and microfracture lines, calculating the angular deviation value of the normal vector of the bone trabeculae fracture surface, labeling the positions of bone trabeculae whose angular deviation value exceeds the deviation threshold, and obtaining the micro-damage area and micro-damage dataset. The identification data is obtained by combining topology maps, fracture sign datasets, soft tissue reaction datasets, and micro-injury datasets.

2. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The skeletal analysis method includes: The skeletal topology unit stores a library of full-skeletal anatomical segmentation models and a skeletal anatomical reference database for realizing image data structure parsing; Geometric feature vectors of the skeletal region in ultrasound image data are extracted using a whole-skeletal anatomical segmentation model, and skeletal anatomical landmarks in ultrasound image data are labeled. The whole skeletal anatomy reference database includes a set of reference anatomical feature points, and the skeletal anatomical landmarks and reference anatomical feature points in the set of reference anatomical feature points are mapped to each other. Based on the mapping relationship, the corresponding skeletal anatomy reference data is obtained from the whole skeletal anatomy reference database. A topological graph is constructed based on geometric feature vectors, mapping relationships, and ultrasound image data.

3. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The feedback mechanism includes: The micro-damage areas detected by the advanced localization layer are fed back to the primary detection layer. The primary detection layer performs a second scan of the micro-damage areas at the sub-pixel level to obtain a sub-pixel-level fracture sign dataset of the micro-damage areas. The sub-pixel-level fracture sign dataset replaces the fracture sign dataset. The size of the initial fracture region in the primary detection layer is adjusted based on the quantified periosteal elevation height in the intermediate classification layer.

4. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The diagnostic decision module includes processing the identified data using diagnostic methods to obtain a diagnostic report; The diagnostic method includes: linking the AO classification database with the trauma scoring unit, identifying data input into the AO classification database and the trauma scoring unit and linking them, and generating a diagnostic report containing fracture type and displacement risk level.

5. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The identification method includes performing cortical continuity analysis on the topology map, scanning the geometric features of the cortical surface, detecting linear fracture zones with lengths exceeding a length threshold, identifying the fragmented edge features of comminuted fractures based on curvature mutation points, and obtaining cortical continuity data, strong echo data of bone fragments, and fragmented edge feature data of fractures. The establishment of the mapping relationship between the ultrasound coordinate system and the anatomical coordinate system includes: A three-dimensional view of the overlapping bone structure in ultrasound image data is unfolded using a surface reconstruction algorithm, and an ultrasound coordinate system is established. Based on skeletal analysis methods, anatomical reference data correlated with ultrasound image data is obtained. A surface reconstruction algorithm is used to unfold a three-dimensional view of the anatomical reference data and establish an anatomical coordinate system. The mapping relationship between the ultrasound coordinate system and the anatomical coordinate system is established based on skeletal anatomical landmarks and reference anatomical feature points.

6. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The geometric feature vector of the skeletal region includes the radius of curvature R and the cortical thickness gradient ΔT, and the anatomical landmarks include the epiphyseal plate growth line L and the articular surface boundary B.

7. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The quantization method includes: Measure the continuous value ΔH of the periosteal elevation height, quantify the hematoma area A, and mark the area where the soft tissue swelling is located; Generate a tissue response feature vector with continuous values ​​ΔH and hematoma area A. When ΔH is greater than the continuous value threshold, the periosteal region where ΔH is greater than the continuous value threshold is marked. The extent of soft tissue injury is determined by combining the periosteal region and the area of ​​soft tissue swelling. The extent of soft tissue injury, combined with Doppler blood flow signals, allows for the identification of indirect periosteal signs and reactions. A soft tissue response dataset is obtained by combining the types of indirect periosteal signs and the extent of soft tissue injury.

8. The ultrasound-assisted system for automatic identification and diagnosis of emergency fractures according to claim 1, characterized in that: The sub-pixel-level analysis includes: Analyze the spatial arrangement of trabecular bone structure, and detect and calculate the deviation of the normal vector angle of microfracture lines and fracture surfaces; Output micro-damage maps, establish spatial association between the coordinate set of the micro-damage core area and anatomical landmarks, and obtain micro-damage datasets by combining micro-fracture lines, fracture surface normal vector angle differences and spatial associations.

Images