Bone development capability evaluation method and device

Abstract

The invention discloses a bone development capability evaluation method and device, and relates to the technical field of image processing, and the method comprises the steps: receiving a preset bone X-ray image of a target user, and extracting key features of bone growth; calling a pre-trained bone development capability evaluation model to analyze the bone growth key features, and generating an initial bone development evaluation result; and acquiring preset bone use information of the target user, comparing the preset bone use information with the model sample parameters, correcting an evaluation result by configuring a feedback coefficient, and generating a personalized target bone development evaluation report. The technical problems that an existing bone development assessment method is high in subjectivity, insufficient in individual difference consideration and lack of a dynamic correction mechanism, and consequently the assessment accuracy and the individuation degree are low are solved, and the purpose that the assessment accuracy and the individuation degree are improved by combining automatic image analysis, individualized bone use information and the correction mechanism is achieved. And the accuracy, individuation and reliability of bone development evaluation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a method and device for evaluating bone development ability. Background Art

[0002] Bone development assessment is an important means in medicine to judge the bone development level of children and adolescents. Traditional methods mainly rely on X-ray images and manual interpretation to evaluate bone age and infer future development potential. However, traditional methods have certain limitations, mainly reflected in large manual intervention, the evaluation accuracy being greatly affected by the experience of operators, and the individual differences of each person not being fully considered. Although some automated evaluation technologies have emerged, most of them have not effectively integrated factors such as the living habits, exercise frequency, and medical history of individuals, resulting in a low degree of personalization of evaluation results when facing different users. In addition, interference factors such as metal artifacts in X-ray images often affect the image quality and thus the accuracy of the evaluation. Summary of the Invention

[0003] This application provides a method and device for evaluating bone development ability, which are used to solve the technical problems that the existing bone development evaluation methods are highly subjective, do not fully consider individual differences, and lack a dynamic correction mechanism, resulting in low evaluation accuracy and personalization degree.

[0004] In the first aspect of this application, a method for evaluating bone development ability is provided. The method includes: receiving a preset bone X-ray image of a target user; extracting key features of bone growth from the preset bone X-ray image; calling a pre-trained bone development ability evaluation model to analyze the key features of bone growth to generate an initial bone development evaluation result; obtaining preset bone usage information of the target user, comparing it with the sample parameters of the bone development ability evaluation model, and configuring a feedback coefficient to correct the initial bone development evaluation result to generate a target bone development evaluation report.

[0005] The second aspect of the present application provides a device for evaluating bone development ability, which includes: a skeletal X-ray image receiving module for receiving a preset skeletal X-ray image of a target user; a growth key feature extraction module for extracting key features of bone growth from the preset skeletal X-ray image; an initial bone development evaluation module for calling a pre-trained bone development ability evaluation model to analyze the key features of bone growth and generate an initial bone development evaluation result. The bone development ability evaluation model includes multiple evaluation units corresponding to multiple bones, and any evaluation unit has an evaluation contribution weight, which is the precision ratio of the corresponding bone for bone development ability evaluation; an evaluation result correction module for obtaining the preset bone usage information of the target user, comparing it with the sample parameters of the bone development ability evaluation model, and configuring a feedback coefficient to correct the initial bone development evaluation result to generate a target bone development evaluation report, specifically including: the preset bone usage information corresponds to the sample parameters, including the usage frequency of the preset bone part and the bone intervention characteristics; comparing the differences between the preset bone usage information and the sample parameters to generate usage differences; configuring the feedback coefficient with the usage differences, where through the configuration of the difference influence layer, the difference influence layer includes a pre-trained influence curve of the usage difference on the bone development evaluation result.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages: A method and device for evaluating bone development ability provided by the present application relate to the technical field of image processing. By receiving the left wrist X-ray image of a target user and extracting key features of bone growth, calling a pre-trained bone development evaluation model to generate a preliminary bone development evaluation result, obtaining the bone usage information of the user, comparing it with the model sample parameters, and correcting the evaluation result by configuring a feedback coefficient to generate a personalized target bone development evaluation report, it solves the technical problems that the existing bone development evaluation methods are highly subjective, do not fully consider individual differences, and lack a dynamic correction mechanism, resulting in low evaluation accuracy and personalization. It realizes the technical effect of improving the accuracy, personalization, and reliability of bone development evaluation by combining automated image analysis, personalized bone usage information, and a correction mechanism. Description of the Drawings

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0008] Figure 1 Schematic flow chart of a method for evaluating bone development ability provided by an embodiment of the present application; Figure 2 Schematic structural diagram of a device for evaluating bone development ability provided by an embodiment of the present application.

[0009] Explanation of reference numerals: Bone X-ray image receiving module 11, key growth feature extraction module 12, initial bone development evaluation module 13, evaluation result correction module 14. Detailed implementation manners

[0010] The present application provides a method and a device for evaluating bone development ability, which are used to solve the technical problems that the existing bone development evaluation methods are highly subjective, do not fully consider individual differences, and lack a dynamic correction mechanism, resulting in low evaluation accuracy and personalization degree.

[0011] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0012] It should be noted that the terms "first", "second", etc. in the specification and the above accompanying drawings of the present application are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, methods, products, or devices.

[0013] Embodiment 1, as Figure 1 shown, the present application provides a method for evaluating bone development ability, and the method includes: P10: Receive a preset bone X-ray image of a target user. Wherein, the preset bone X-ray image is an X-ray image of the left wrist of the target user.

[0014] It should be understood that a preset skeletal X-ray image provided by the target user is received first. This image is an X-ray image of the target user's left wrist. The reason for choosing the left wrist is that the bone structure of the left wrist is significantly representative during the individual's development process. Characteristics such as the development sequence of the epiphyseal line, changes in bone density, and the maturity of bone morphology can clearly reflect the individual's bone development level. At the same time, the left wrist has a relatively low usage frequency, and the possibility of the bone being damaged by the outside world is small. Therefore, its image can more stably reflect the natural development state of the bone. However, for left-handed users or users who have had surgery, the usage frequency and bone state of the left hand may be different, and the system should have the ability to automatically judge and adjust the evaluation. For example, if the user is left-handed, it is necessary to automatically identify the situation where the left hand has a higher usage frequency, or if the left hand has received surgical treatment, compensation adjustment should be made based on historical data to ensure the accuracy of the evaluation results.

[0015] The process of image acquisition needs to ensure high-quality X-ray images, usually carried out in a professional institution. The user should place the left wrist correctly according to the guidance of the technical personnel and ensure that it is completely exposed in the imaging area of the X-ray machine to avoid image blurring caused by improper hand posture or angle problems. The X-ray equipment should be set to high-resolution mode to ensure clear images, especially in the display of key areas such as bone details, epiphyseal lines, and bone density. At the same time, the equipment needs to be calibrated before image acquisition to ensure that parameters such as exposure and contrast meet the technical standards, so as to capture delicate bone features.

[0016] After the image acquisition is completed, the image data will be uploaded to the evaluation system through a digital imaging transmission system (such as DICOM format). During the transmission process, the system will ensure data integrity and conduct a preliminary verification of the image file to check for transmission errors or data loss. After the image data is successfully transmitted, the system will perform a preliminary processing of the image to determine whether the image meets the evaluation criteria. For example, the system will check for metal artifacts or obvious blurred areas. If the image quality does not meet the requirements, the system will prompt the technical personnel to re-acquire the image.

[0017] When the image data passes the preliminary verification and is confirmed to meet the standards, preprocessing of the image is carried out. This includes using artifact removal techniques to identify and process metal artifacts in the image to ensure that the bone area is clearly visible. Using image segmentation techniques to separate the artifact area from the bone area and repairing or reconstructing the artifact area through algorithms to ensure the authenticity and reliability of the final image. And performing image enhancement processing on the image to enhance the detail performance of the bone area by means such as increasing contrast and sharpening edges, which helps subsequent feature extraction and analysis and improves the evaluation accuracy. For complex cases, more advanced processing of the image can also be carried out, such as local magnification and enhancement of bone details, to ensure that each user's image can provide sufficient data information.

[0018] Finally, all the imaging data and the user's skeletal information will be saved by the system and archived as the basic data for subsequent evaluations. The system will generate detailed imaging processing records and data logs for each user to ensure that the processes of image acquisition, processing, and adjustment are traceable. If the user needs to view historical data or conduct further evaluations, the system provides a data access interface through which the user can view the detailed reports of their bone development assessments at any time. The core of the whole process lies in ensuring the high-quality acquisition, precise processing, and personalized adjustment of the imaging data to provide reliable data support for bone development assessments. This not only guarantees the accuracy of the assessment results but also ensures adaptability and fairness across different user backgrounds.

[0019] Furthermore, after receiving the preset skeletal X-ray image of the target user, step P10 of the embodiment of the present application further includes: P11a: Conduct metal artifact analysis based on the preset skeletal X-ray image to determine whether there are metal artifacts; P12a: If so, perform image reconstruction of the metal artifact area to reconstruct the image for bone development ability assessment.

[0020] Optionally, after receiving the preset skeletal X-ray image of the target user, artifact removal techniques can be further used to identify and process the metal artifacts in the image to ensure that the final assessment results are not interfered with by the artifacts.

[0021] First, metal artifact analysis is performed on the preset skeletal X-ray image. Metal artifacts occur when X-rays penetrate metal objects and scatter or are absorbed, resulting in distorted areas in the image that cannot accurately reflect the true structure of the bones. For example, some users' wrists may contain metal prostheses or surgical implants, and these metal objects usually produce obvious artifacts in the image, which can affect the accuracy of subsequent bone feature extraction and assessment results. Therefore, it is necessary to detect whether there are metal artifacts in the image and determine their positions and ranges. This analysis process can be achieved through image processing algorithms. For example, threshold segmentation techniques can be used to identify high-density areas, or edge detection algorithms can be used to identify the boundaries of metal artifacts. In addition, machine learning algorithms can be combined to achieve automatic detection and classification of metal artifacts through training on a large number of image samples with metal artifacts. The system will check each part of the image and identify the distorted areas that may be caused by the interference of metal objects.

[0022] If the presence of metal artifacts is detected, image reconstruction of the metal artifact area needs to be performed. The purpose of image reconstruction is to remove or reduce the interference of metal artifacts on bone feature analysis through algorithm processing, so as to generate high-quality images that can be used for bone development ability assessment. The metal artifact area usually appears as black or white areas in the image, or as unclear bone parts. Image reconstruction can be achieved through various techniques. A common method is based on iterative reconstruction algorithms, which optimize the image reconstruction process through multiple iterations and gradually reduce the impact of metal artifacts. Another method is to utilize deep learning techniques, by training a convolutional neural network (CNN) model to identify and repair metal artifact areas. These models can be trained on a large amount of image data with metal artifacts to learn how to recover a clear bone structure from the artifact-containing images.

[0023] After performing image reconstruction, the generated reconstructed image will be used for subsequent bone development ability assessment. The quality of the reconstructed image needs to meet the following criteria: First, the bone features after reconstruction should be clearly visible and accurately reflect the shape and structure of the bone; Second, the metal artifacts in the reconstructed image should be effectively removed or significantly reduced to avoid interference with subsequent analysis; Finally, the reconstructed image should maintain the same anatomical features and proportional relationships as the original image to ensure the accuracy of the assessment results.

[0024] Through the above steps, this application can effectively process X-ray images containing metal artifacts, ensuring the accuracy and reliability of subsequent bone development ability assessment. The introduction of this step greatly improves the robustness and adaptability of the solution of the present invention. Especially when dealing with complex cases, it can effectively avoid errors caused by metal artifacts.

[0025] Furthermore, for the image reconstruction of the metal artifact area, step P12a of the embodiment of this application further includes: P12-1a: Determine a set of healthy pixel points on the artifact boundary of the metal artifact area, and the set of healthy pixel points is evenly distributed on the artifact boundary; P12-2a: Based on the set of healthy pixel points, perform radial basis function interpolation on each pixel point in the metal artifact area to complete artifact removal, and then perform transition smoothing processing on the edge of the repaired area to generate the reconstructed image.

[0026] In a possible embodiment of this application, the image reconstruction process of the metal artifact area can be further refined to ensure the quality and usability of the reconstructed image.

[0027] First, determine the set of healthy pixel points at the artifact boundary of the metal artifact region. The artifact boundary is usually the intersection between the artifact region and the normal bone region, where there is a transitional phenomenon, usually manifested as distortion or blurring of the bone image. To restore the artifact region, through image processing algorithms, such as edge detection algorithms, the boundary of the metal artifact can be accurately identified. On this basis, a certain number of healthy pixel points are evenly selected around the artifact boundary. The selection of these pixel points should be based on their gray values, texture features, etc., to ensure that they are not affected by the artifact and can be evenly distributed on the artifact boundary, providing an accurate reference for the subsequent interpolation process. The distribution of healthy pixel points will ensure that at the edge of the artifact region, the interpolation algorithm can more accurately capture the true shape of the bone.

[0028] Next, after determining the set of healthy pixel points, perform radial basis function (RBF) interpolation on each pixel point within the metal artifact region based on these healthy pixel points. The radial basis function is a commonly used interpolation method that can effectively fill the gaps between data points and maintain the smoothness and accuracy of the interpolation results. Here, the RBF interpolation method calculates each pixel point in the artifact region according to the distribution and characteristics of the healthy pixel points, ensuring that the pixel values in the artifact region are consistent with those in the surrounding healthy regions. Through this interpolation method, the details in the artifact region can be restored, avoiding any distortion within the artifact region.

[0029] After the interpolation is completed, there may be discontinuities in the gray values between the restored region and the surrounding normal image region, resulting in unnatural edge transitions. Therefore, it is necessary to perform smoothing processing on the edges of the restored region. Image filtering techniques, such as Gaussian filtering or median filtering, can be used to smooth the edges of the restored region. By selecting appropriate filters and parameters, the sudden change in gray values can be effectively reduced, making the transition between the restored region and the surrounding regions more natural and continuous.

[0030] Through the above steps, this application can effectively remove the influence of metal artifacts on the skeletal X-ray image and generate a high-quality reconstructed image. This process not only ensures the integrity and accuracy of the image but also provides a reliable basis for the subsequent assessment of bone development ability.

[0031] P20: Extract the key features of bone growth from the preset skeletal X-ray image.

[0032] Specifically, the preset skeletal X-ray images are processed to extract the key features of bone growth. To ensure that the extracted features can accurately reflect the bone growth status, it is necessary to preprocess the X-ray images first to optimize the image quality and reduce interference factors. Specifically, the preprocessing includes converting the images into grayscale images to simplify the subsequent analysis process, adjusting the grayscale value range through normalization operations to make it conform to a unified standard, thereby reducing the differences caused by acquisition with different devices. In addition, a filtering algorithm is used to remove noise to enhance the clarity of the images, and contrast enhancement techniques are used to highlight bone details for subsequent analysis.

[0033] After the preprocessing is completed, it enters the key feature extraction stage. Exemplarily, the key features of bone growth mainly include epiphyseal line features, bone density features, bone morphology features, and bone maturity features. Among them, the epiphyseal line is a key marker of bone growth, and its morphology, width, and closure degree can reflect the bone development stage. Through image segmentation technology, the position of the epiphyseal line can be accurately identified, and its length, width, and morphological features can be extracted. For example, the width of the epiphyseal line can be quantified by calculating its pixel width on the image, and the morphological features can be evaluated by analyzing the smoothness or irregularity of its edge.

[0034] Bone density is an important indicator reflecting bone health status and can be indirectly reflected by analyzing the grayscale distribution of X-ray images. Specifically, calculating statistical quantities such as the average grayscale value and standard deviation of the bone region can provide quantitative information on bone density. In addition, texture analysis methods are used to extract texture features of the bone region, such as texture uniformity and contrast, and these features are also closely related to bone density and can further enrich the evaluation dimensions of bone density.

[0035] The morphological features of the bone include the length, width, curvature, etc. of the bone. Through image measurement technology, the length and width of the bone can be accurately measured, and its curvature can be calculated. For example, for long bones, the curvature can be calculated by fitting the center line of the bone to evaluate the morphological development of the bone. These morphological features can not only reflect the current state of the bone but also provide a basis for predicting the future development trend of the bone.

[0036] Finally, the bone maturity features can be evaluated by analyzing the development of the ossification centers of the bone. For example, for the wrist bones, the maturity of the bone can be judged by identifying and measuring the number, size, and position of the ossification centers. These features are important bases for evaluating bone development ability and can provide key information for clinical diagnosis and treatment.

[0037] In the process of extracting the above key features, various image processing and analysis techniques can be adopted. For example, edge detection algorithms can be used to identify the contours of bones, region growing algorithms can be used to segment bone regions, and convolutional neural networks (CNNs) can be used to automatically extract complex features. As a powerful deep learning model, CNN can automatically learn the feature representations in images, thereby improving the accuracy and efficiency of feature extraction. In addition, to ensure the accuracy and reliability of feature extraction, the extracted features also need to be verified and calibrated. It can be verified by comparing with known bone development standards or using data annotated by experts to ensure that the extracted features can truly reflect the growth status of bones.

[0038] P30: Call a pre-trained bone development ability evaluation model to analyze the key features of bone growth and generate an initial bone development evaluation result.

[0039] Among them, the bone development ability evaluation model includes multiple evaluation units corresponding to multiple bones, and any evaluation unit has an evaluation contribution weight, and the evaluation contribution weight is the precision ratio of the corresponding bone to the bone development ability evaluation.

[0040] It should be understood that a pre-trained bone development ability evaluation model is called to analyze the key features of bone growth extracted and generate a preliminary bone development evaluation result. First of all, the bone development ability evaluation model is a deep learning model trained with a large amount of bone development data. This model can be processed based on multiple evaluation units corresponding to multiple bones. Each evaluation unit is responsible for analyzing the growth characteristics of a specific bone part and making a comprehensive evaluation according to the development characteristics of different bone parts.

[0041] In the construction of the model, each evaluation unit is associated with a specific bone region (such as the radius, ulna, metacarpal, etc.). Each bone corresponding to an evaluation unit has a specific evaluation contribution weight, and this weight value represents the contribution ratio of the bone to the bone development evaluation. The development of different bone parts has different impacts on bone development. The development of some parts may have higher accuracy for the overall bone development ability evaluation, so the evaluation units of these parts will be given higher weights. While the evaluation units of other bone parts may be given relatively lower weights according to their smaller influence on the evaluation results.

[0042] The system generates a preliminary bone development assessment result by matching and calculating the key features of bone growth extracted with the models of each evaluation unit. In this process, the evaluation model takes into account the growth characteristics of each bone region, combines their corresponding weights, accurately predicts the development status of this region, and obtains the overall bone development assessment result through the comprehensive analysis of these units. For example, if the development of the epiphyseal line in the wrist region is typical and clear, and the evaluation contribution weight of this part is high, then the evaluation result of this part will have a greater impact on the overall evaluation result.

[0043] Finally, the evaluation results of these individual bone regions are comprehensively calculated according to the set weight ratio to generate a preliminary bone development assessment result. This result will include the estimated value of bone age, the development level of the bones, and the risk of possible abnormalities or developmental lags. This preliminary assessment result provides data support for the subsequent bone development ability assessment and a basis for individualized assessment.

[0044] Furthermore, for the construction of the bone development ability assessment model, Embodiment of the present application further includes step P30a, and step P30a further includes: P31a: Collect a key feature sample set of the multiple bones and a corresponding bone age evaluation sample set for the multiple bones; P32a: Train the multiple evaluation units with the key feature sample set of the multiple bones and the bone age evaluation sample set combined according to the corresponding relationship; P33a: Construct a fusion layer with the evaluation contribution weights, connect it with the multiple evaluation units, and generate the bone development ability assessment model.

[0045] Optionally, the construction process of the bone development ability assessment model can be further detailed.

[0046] First, collect a key feature sample set of the multiple bones and a corresponding bone age evaluation sample set. These sample sets are obtained by analyzing a large amount of bone imaging data and clinical case data, and contain key feature data of multiple bone regions (such as the radius, ulna, metacarpal, etc.), and each feature data corresponds to a professionally evaluated bone age value. The key feature sample set of the bones includes indicators such as the bone morphology in the image, the width of the epiphyseal line, and bone density, while the bone age evaluation sample set is the result of judging the bone development stage by professional standards. The system uses these data as inputs to ensure that the evaluation model can be trained based on diverse bone features.

[0047] Next, the multiple collected key bone feature sample sets and bone age assessment sample sets are combined according to the corresponding relationship, and multiple evaluation units are trained in combination with deep learning technology. Each evaluation unit is specifically used to analyze and process the feature data of a specific bone region. By training on these data sets, the relationship between different bone regions and bone age development is learned, providing each evaluation unit with strong prediction ability. Exemplarily, machine learning or deep learning algorithms, such as convolutional neural networks (CNNs), train the model through a large amount of sample data so that it can automatically learn the mapping relationship between features and bone age. During the training process, the model continuously adjusts its internal parameters to minimize the error between the prediction result and the actual bone age assessment, thereby improving the prediction accuracy of each evaluation unit. By training the evaluation units, each unit can accurately analyze a specific bone region and extract the most representative bone growth features.

[0048] Finally, in order to integrate multiple evaluation units into a complete bone development ability evaluation model, a fusion layer needs to be constructed. The role of the fusion layer is to perform weighted fusion on the output results of each evaluation unit according to the evaluation contribution weights of each evaluation unit. The evaluation contribution weights reflect the importance of each bone in the overall bone development ability evaluation, and these weights can be set or optimized according to clinical experience and experimental data. Through the fusion layer, the prediction results of multiple evaluation units are comprehensively processed to generate the final bone development ability evaluation model. This model can comprehensively consider the development characteristics of multiple bone parts and their contributions to the overall bone development ability, thereby providing a comprehensive and accurate bone development evaluation result.

[0049] Through these steps, the system can construct a comprehensive and accurate bone development evaluation model. This model can not only independently analyze the growth characteristics of each bone region but also combine the comprehensive information of these regions to generate a personalized bone development ability evaluation report.

[0050] P40: Obtain the preset bone usage information of the target user, compare it with the sample parameters of the bone development ability evaluation model, configure a feedback coefficient to correct the initial bone development evaluation result, and generate a target bone development evaluation report.

[0051] Further, step P40 of the embodiment of the present application further includes: P41: The preset bone usage information corresponds to the sample parameters and includes the preset usage frequency of the bone part and the bone intervention characteristics; P42: Compare the difference between the preset bone usage information and the sample parameters to generate a usage difference; P43: Configure the feedback coefficient with the usage difference, where the configuration is performed through a difference impact layer, and the difference impact layer includes a pre-trained curve of the impact of the usage difference on the bone development evaluation result.

[0052] It should be understood that the personalized correction process for bone development ability assessment can be further refined to ensure that the assessment results can more accurately reflect the actual bone development status of the target user.

[0053] Specifically, first, the preset bone usage information of the target user needs to be obtained. This information includes the usage frequency of the preset bone parts and the bone intervention characteristics (such as whether the user has ever undergone surgery, had a fracture, etc.). These usage information are compared with the sample parameters in the bone development ability assessment model. The sample parameters are obtained through training based on a large amount of sample data and reflect the relationship between bone development and usage under normal circumstances. Through comparison, the system can identify the differences between the bone usage information of the target user and the sample parameters, that is, the usage differences.

[0054] Subsequently, a feedback coefficient is configured according to the identified usage differences. This process is achieved through the difference impact layer, which includes pre-trained impact curves of usage differences on bone development assessment results. These impact curves are obtained through training based on a large amount of sample data and can reflect the specific impact degree of different usage differences on bone development assessment results. The difference impact layer will look up the corresponding correction coefficient on the pre-trained impact curve according to the identified usage differences. For example, if the usage frequency of the preset bone part of the target user is much higher than the average level in the sample parameters and the part has undergone surgery, the difference impact layer will find the corresponding correction coefficient on the impact curve according to these differences. These correction coefficients will be used to adjust the initial bone development assessment results to more accurately reflect the actual bone development status of the target user.

[0055] The configuration and application of the feedback coefficient can be achieved through a mathematical model. The specific formula can be expressed as: the corrected bone development assessment result = the initial bone development assessment result × the feedback coefficient. Among them, the feedback coefficient is found in the difference impact layer according to the usage differences and can quantify the impact of usage differences on bone development assessment results.

[0056] After the above correction process, the system will generate a target bone development assessment report. This report not only includes the initial bone development assessment results but also reflects the personalized correction based on the preset bone usage information of the target user. The report will detail the reasons and bases for the correction. For example, the high usage frequency of the target user's bones or the fact that the user has undergone surgery, etc., and the specific impact of these factors on bone development assessment results. In this way, the target bone development assessment report can provide more accurate and targeted diagnostic bases for clinicians, thus better serving the treatment and rehabilitation of patients.

[0057] Through the above steps, the present application realizes personalized correction of bone development ability evaluation. It not only considers the individual differences of the target user, but also ensures the scientificity and accuracy of the correction process through the pre-trained differential influence layer and feedback coefficient configuration. The finally generated target bone development evaluation report can more truly reflect the skeletal development status of the target user, providing strong support for clinical diagnosis and treatment.

[0058] Further, step P40 of the embodiment of the present application further includes: P44: Extract the actual bone age evaluation result according to the corrected bone development evaluation result; P45: Read the user attribute information of the target user, including age, gender, height, and the height of the parents; P46: Call the residual height growth potential analysis model to predict the adult height growth potential in combination with the actual bone age evaluation result and the user attribute information, and generate a target growth potential prediction result. The residual height growth potential analysis model is trained based on neural network using historical data samples; P47: Generate the target bone development evaluation report with the actual bone age evaluation result and the target growth potential prediction result.

[0059] Specifically, the process of bone development ability evaluation can be further refined. Not only is the initial bone development evaluation result corrected in a personalized manner, but the adult height growth potential can also be predicted in combination with the detailed information of the target user, and finally a comprehensive target bone development evaluation report is generated.

[0060] First, extract the actual bone age evaluation result according to the corrected bone development evaluation result. After the above-mentioned correction steps, the system has optimized and adjusted the initial bone development evaluation result, making the evaluation result more in line with the individual situation of the user. At this time, the system will extract the actual bone age from the corrected evaluation result. This bone age value reflects the true development stage of the user's bones and is used as the basis for subsequent growth potential prediction.

[0061] Next, read the user attribute information of the target user, including age, gender, height, and the height of the parents. The personal attribute information of the user is crucial for predicting the growth potential. For example, gender and the height of the parents have a significant impact on predicting the adult height. By reading this attribute information, the system can combine the user's growth background with other data to provide the necessary input for growth potential prediction.

[0062] Further, call the residual height growth potential analysis model, and combine the actual bone age assessment result and the user attribute information to predict the adult height growth potential. The residual height growth potential analysis model is trained based on a neural network. This model is obtained through learning a large number of historical data samples and can analyze the influence of the user's bone age and growth background (such as age, gender, parental height, etc.) on the final adult height. The system inputs the actual bone age and the user attribute information into this model, and the model predicts the user's future height growth potential based on this information and obtains the predicted value of the height after adulthood. Specifically, the model will judge the current bone development stage of the target user according to the actual bone age assessment result, and combine factors such as his age, gender, current height, and parental height, and predict the height growth potential at adulthood through the neural network algorithm.

[0063] Finally, combine the actual bone age assessment result and the target growth potential prediction result to generate a target bone development assessment report. This report not only includes the actual level of the target user's current bone development, but also provides a prediction of his future height growth potential. The report will detail the difference between the target user's bone age and the actual age, the current bone development status, and the adult height prediction based on genetics and the current growth status. In addition, the report will also provide relevant explanations and suggestions. For example, if the prediction result shows that the target user has insufficient height growth potential, the report may recommend further medical examinations or intervention measures.

[0064] In summary, this application not only realizes the personalized assessment of bone development ability, but also further expands the scope of assessment and provides a prediction of future growth potential. This report can provide users with a comprehensive assessment of the current situation of bone development and future growth potential, ensuring the scientificity, personalization, and accuracy of the assessment results.

[0065] Further, after extracting the actual bone age assessment result, step P40 of the embodiment of this application further includes: P45a: Compare the actual bone age assessment result with the current age of the target user to determine the bone age deviation, and the bone age deviation carries a deviation direction identifier; P46a: Generate a bone age abnormality reminder message based on the bone age deviation.

[0066] Optionally, after extracting the actual bone age assessment result, the assessment process can be further expanded to add the function of analyzing the bone age deviation and generating a bone age abnormality reminder message.

[0067] First, after completing the personalized correction of the initial bone development assessment results and extracting the actual bone age assessment results, the system compares the actual bone age assessment results with the current age of the target user to determine the bone age deviation. The bone age deviation refers to the difference between the actual bone age and the current actual age of the target user, which can reflect whether the skeletal development of the target user is consistent with that of their peers. Specifically, the bone age deviation can be a positive or negative value, indicating that the bone age is higher or lower than the actual age, respectively. To more clearly represent this difference, the bone age deviation also carries a deviation direction identifier, which is used to clearly indicate whether the bone age is advanced or delayed. For example, if the actual bone age is 12 years old and the current age of the target user is 10 years old, the bone age deviation is +2 years old, and the deviation direction identifier is "advanced"; conversely, if the actual bone age is 8 years old and the current age of the target user is 10 years old, the bone age deviation is -2 years old, and the deviation direction identifier is "delayed".

[0068] Next, based on the determined bone age deviation, a bone age abnormality reminder message is further generated. The purpose of this reminder message is to promptly notify the target user that there is a significant difference between their bone age and actual age. The bone age abnormality reminder message will include the specific value of the bone age deviation, the deviation direction identifier, and relevant explanations. For example, if the bone age deviation is large and the direction is advanced, the reminder message may indicate that the target user may be at risk of precocious puberty and recommend further medical examinations; if the bone age deviation is large and the direction is delayed, the reminder message may prompt that the target user may have growth and development retardation and recommend nutritional assessment or endocrine examinations. If the user's bone age deviation exceeds the normal growth range, the system will generate relevant abnormality reminder messages through this deviation. The reminder message will specifically describe the direction and degree of the bone age deviation, and based on information such as the user's age and gender, provide a judgment on whether there is an abnormality.

[0069] Through this process, users can obtain detailed feedback on their own bone development status. The calculation of the bone age deviation and the generation of abnormality reminders enable the system to provide personalized and timely health guidance to help users identify potential developmental problems at an early stage.

[0070] In summary, the embodiments of the present application have at least the following technical effects: This application improves the accuracy of bone development assessment through automated image analysis and deep learning models, reducing errors caused by manual intervention. At the same time, by combining the bone usage information of the target user (such as usage frequency, exercise habits, medical history, etc.), personalized assessment results are provided. Through metal artifact analysis and image reconstruction technology, artifact interference is eliminated, ensuring clear images and reliable assessment results. In addition, the system is configured with a feedback coefficient to correct the initial assessment results according to the individual characteristics of the user, ensuring that the assessment results are more in line with the actual bone development status. Finally, a comprehensive bone development assessment report is automatically generated to provide scientific and accurate health management support for users.

[0071] It achieves the technical effect of improving the accuracy, personalization, and reliability of bone development assessment by combining automated image analysis, personalized bone usage information, and a correction mechanism.

[0072] Embodiment 2, based on the same inventive concept as the method for evaluating bone development ability in the foregoing embodiment, as Figure 2 shown, this application provides a device for evaluating bone development ability. The device in the embodiment of this application and the method embodiment are based on the same inventive concept. Among them, the device includes: A skeletal X-ray image receiving module 11, which is used to receive a preset skeletal X-ray image of the target user. Among them, the preset skeletal X-ray image is an X-ray image of the left wrist of the target user.

[0073] A growth key feature extraction module 12, which is used to extract the key features of bone growth from the preset skeletal X-ray image.

[0074] An initial bone development assessment module 13, which is used to call a pre-trained bone development ability assessment model to analyze the key features of bone growth and generate an initial bone development assessment result. Among them, the bone development ability assessment model includes multiple evaluation units corresponding to multiple bones, and any evaluation unit has an evaluation contribution weight, and the evaluation contribution weight is the precision ratio of the corresponding bone for bone development ability assessment.

[0075] An assessment result correction module 14, which is used to obtain the preset bone usage information of the target user, compare it with the sample parameters of the bone development ability assessment model, configure a feedback coefficient to correct the initial bone development assessment result, and generate a target bone development assessment report.

[0076] Furthermore, the skeletal X-ray image receiving module 11 is also used to perform the following steps: Perform metal artifact analysis based on the preset skeletal X-ray image to determine whether there are metal artifacts; if so, perform image reconstruction of the metal artifact area and evaluate the bone development ability using the reconstructed image.

[0077] Furthermore, the skeletal X-ray image receiving module 11 is further configured to perform the following steps: Determine a set of healthy pixel points on the artifact boundary of the metal artifact area, and the set of healthy pixel points is evenly distributed on the artifact boundary; based on the set of healthy pixel points, perform radial basis function interpolation on each pixel point within the metal artifact area to complete artifact removal, and then perform transitional smoothing on the edge of the repaired area to generate the reconstructed image.

[0078] Furthermore, the initial bone development evaluation module 13 is further configured to perform the following steps: Collect a set of key feature samples and corresponding sets of bone age evaluation samples for the multiple bones; train the multiple evaluation units by combining the set of key feature samples and the sets of bone age evaluation samples for the multiple bones according to the corresponding relationship; construct a fusion layer with the evaluation contribution weights and connect it to the multiple evaluation units to generate the bone development ability evaluation model.

[0079] Furthermore, the evaluation result correction module 14 is further configured to perform the following steps: The preset skeletal usage information corresponds to the sample parameters, including the usage frequency of the preset skeletal part and the skeletal intervention characteristics; compare the difference between the preset skeletal usage information and the sample parameters to generate a usage difference; configure the feedback coefficient with the usage difference, where the configuration is through a difference impact layer, and the difference impact layer includes a pre-trained curve of the impact of the usage difference on the bone development evaluation result.

[0080] Furthermore, the evaluation result correction module 14 is further configured to perform the following steps: Extract the actual bone age evaluation result according to the corrected bone development evaluation result; read the user attribute information of the target user, including age, gender, height, and parental height; call the residual height growth potential analysis model to predict the adult height growth potential by combining the actual bone age evaluation result and the user attribute information to generate a target growth potential prediction result, and the residual height growth potential analysis model is trained based on historical data samples using a neural network; generate the target bone development evaluation report with the actual bone age evaluation result and the target growth potential prediction result.

[0081] Furthermore, the evaluation result correction module 14 is further configured to perform the following steps: Compare the actual bone age assessment result with the current age of the target user to determine a bone age deviation, where the bone age deviation carries a deviation direction identifier; generate a bone age abnormality reminder message based on the bone age deviation.

[0082] It should be noted that the above order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above specific embodiments of the present specification have been described. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0084] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A method for evaluating bone development ability, characterized in that, Comprising: Receiving a preset bone X-ray image of a target user; Extracting key features of bone growth from the preset bone X-ray image; Invoking a pre-trained bone development ability evaluation model to analyze the key features of bone growth, generating an initial bone development evaluation result. The bone development ability evaluation model includes multiple evaluation units corresponding to multiple bones, and any evaluation unit has an evaluation contribution weight, and the evaluation contribution weight is the precision ratio of the corresponding bone for bone development ability evaluation; Obtaining the preset bone usage information of the target user, comparing it with the sample parameters of the bone development ability evaluation model, configuring a feedback coefficient to correct the initial bone development evaluation result, and generating a target bone development evaluation report, specifically including: The preset bone usage information corresponds to the sample parameters, including the usage frequency of the preset bone part and the bone intervention characteristics; Comparing the difference between the preset bone usage information and the sample parameters to generate a usage difference; Configuring the feedback coefficient with the usage difference, wherein, through the configuration of the difference influence layer, the difference influence layer includes a pre-trained influence curve of the usage difference on the bone development evaluation result.

2. The method for evaluating bone development ability according to claim 1, wherein The preset bone X-ray image is an X-ray image of the left wrist of the target user.

3. The method for evaluating bone development ability according to claim 1, wherein The construction steps of the bone development ability evaluation model include: For the multiple bones, collecting multiple key feature sample sets of the bones and corresponding multiple bone age evaluation sample sets; Training the multiple evaluation units with the multiple key feature sample sets of the bones and the multiple bone age evaluation sample sets combined according to the corresponding relationship; Constructing a fusion layer with the evaluation contribution weight, connecting it with the multiple evaluation units, and generating the bone development ability evaluation model.

4. The method for evaluating bone development ability according to claim 1, wherein After receiving the preset bone X-ray image of the target user, it further includes: Performing metal artifact analysis based on the preset bone X-ray image to determine whether there are metal artifacts; If so, performing image reconstruction on the metal artifact area and evaluating the bone development ability with the reconstructed image.

5. The method for evaluating bone development ability according to claim 4, wherein, Performing image reconstruction on the metal artifact area includes: Determining a set of healthy pixel points at the artifact boundary of the metal artifact area, and the set of healthy pixel points is evenly distributed at the artifact boundary; Based on the set of healthy pixel points, performing radial basis function interpolation on each pixel point in the metal artifact area to complete artifact removal, and then performing transitional smoothing processing on the edge of the repaired area to generate the reconstructed image.

6. The method for evaluating bone development ability according to claim 1, characterized in that Configuring a feedback coefficient to correct the initial bone development evaluation result and generating a target bone development evaluation report includes: Extracting the actual bone age evaluation result according to the corrected bone development evaluation result; Reading the user attribute information of the target user, including age, gender, height, and parents' height; Invoking a residual height growth potential analysis model to predict the adult height growth potential by combining the actual bone age evaluation result and the user attribute information, generating a target growth potential prediction result. The residual height growth potential analysis model is trained based on historical data samples using a neural network; Generating the target bone development evaluation report with the actual bone age evaluation result and the target growth potential prediction result.

7. The method for evaluating bone development ability according to claim 6, wherein After extracting the actual bone age evaluation result, it further includes: Compare the actual bone age assessment result with the current age of the target user to determine a bone age deviation, and the bone age deviation carries a deviation direction identifier; Generate a bone age abnormality reminder message based on the bone age deviation.

8. An apparatus for evaluating bone development ability, characterized in that, For implementing the steps of the method for evaluating bone development ability according to any one of claims 1 to 7, the device includes: A skeletal X-ray image receiving module for receiving a preset skeletal X-ray image of a target user; A growth key feature extraction module for extracting key skeletal growth features from the preset skeletal X-ray image; An initial bone development assessment module for calling a pre-trained bone development ability assessment model to analyze the key skeletal growth features and generate an initial bone development assessment result. The bone development ability assessment model includes multiple evaluation units corresponding to multiple bones, and any evaluation unit has an evaluation contribution weight, and the evaluation contribution weight is the accuracy ratio of the corresponding bone for bone development ability assessment; An evaluation result correction module for obtaining the preset skeletal usage information of the target user, comparing it with the sample parameters of the bone development ability assessment model, and configuring a feedback coefficient to correct the initial bone development assessment result to generate a target bone development assessment report, specifically including: The preset skeletal usage information corresponds to the sample parameters and includes the usage frequency of the preset skeletal part and the skeletal intervention characteristics; Compare the difference between the preset skeletal usage information and the sample parameters to generate a usage difference; Configure the feedback coefficient with the usage difference, wherein, through the difference influence layer configuration, the difference influence layer includes a pre-trained influence curve of the usage difference on the bone development assessment result.

Images