Method, equipment and program product for predicting osteoporosis period based on content of methylglyoxal
Through the combination of CT imaging data and methylglyoxal content, the osteoporosis period is predicted, which solves the problem of early diagnosis of osteoporosis, reduces the risk of fractures, and improves the bone health management effect of diabetic patients.
Patent Information
- Application Number
- CN202410981662.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
AI Technical Summary
Osteoporosis, especially diabetic osteoporosis, has difficulty in early diagnosis, leading to a high risk of fracture and low compliance with treatment, affecting the treatment effect.
By obtaining CT image data at different time points, extracting and comparing image features such as cortical bone thickness, total cortical bone area, and trabecular bone resolution, combining methylglyoxal content, using neural network models to predict the period of osteoporosis. Realize early diagnosis and staging.
Effectively detecting early osteoporosis, reducing the risk of fractures, and improving bone health management in diabetic patients is of great clinical significance.
Smart Images

Figure CN118942644A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent medical care, and specifically to a method, device, program product and computer-readable storage medium for predicting the period of osteoporosis based on methylglyoxal content. Background Art
[0002] Osteoporosis is a systemic skeletal disease characterized by decreased bone mass and deterioration of bone microarchitecture, leading to bone fragility and increased risk of fractures. It usually occurs in middle-aged and elderly people, especially postmenopausal women, but can also affect men and children. Osteoporosis is one of the global public health issues because it increases the risk of fractures, especially fractures of the spine, hip, and wrist, which can lead to chronic pain, decreased physical function, disability, and even shortened life. In recent years, research on osteoporosis has made considerable progress, including a deeper understanding of the disease mechanism, the development of new treatments, and the optimization of prevention strategies. However, the pathogenesis of osteoporosis involves multiple systems and molecular pathways, and its complexity increases the difficulty of research. The manifestation and progression rate of osteoporosis vary greatly in different individuals, which affects the universal applicability and effectiveness of the therapy. Osteoporosis often has no obvious symptoms in the early stages and is not discovered until a fracture occurs, which limits the opportunity for early intervention. The treatment of osteoporosis usually requires long-term persistence, but the patient's compliance is not high, which affects the treatment effect.
[0003] Diabetic osteoporosis (DOP) is a skeletal system change secondary to diabetes, characterized by decreased bone mineral density, increased bone brittleness, and increased risk of fractures. DOP can easily cause patients to experience local pain, deformity, paralysis and other symptoms, endangering their physical and mental health and reducing their quality of life. It is urgent to conduct in-depth research on the pathogenesis of DOP and seek methods to improve the bone quality of diabetic patients. Summary of the invention
[0004] In view of the above problems, the present invention proposes a method for predicting the osteoporosis period based on methylglyoxal content, which specifically includes:
[0005] S1: obtaining CT image data of a subject at different time points, where the subject is a subject with methylglyoxal accumulation;
[0006] S2: extracting the first node image features and the second node image features based on the CT image data at different time points;
[0007] S3: comparing the image feature of the first node with the image feature of the second node;
[0008] S4: The image features include cortical bone thickness. When the cortical bone thickness of the first node is less than that of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than that of the second node, it is determined to be the second stage of osteoporosis.
[0009] The image feature is replaced by the total cortical bone area. When the total cortical bone area of the first node is smaller than the total cortical bone area of the second node, it is determined to be the first stage of osteoporosis; when the total cortical bone area of the first node is larger than the total cortical bone area of the second node, it is determined to be the second stage of osteoporosis;
[0010] Optionally, the image feature is replaced by trabecular separation, when the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis;
[0011] Optionally, the image features are replaced by the total cortical bone area and the trabecular separation; when the total cortical bone area of the first node is smaller than the total cortical bone area of the second node, and the trabecular separation of the first node is larger than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the total cortical bone area of the first node is larger than the total cortical bone area of the second node, and the trabecular separation of the first node is smaller than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis;
[0012] Optionally, the first node image feature and the second node image feature are image features selected sequentially at different time points on the same time axis;
[0013] Optionally, the content of methylglyoxal increases with time.
[0014] The image features also include the total area of cortical bones. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the total area of the cortical bones of the first node is less than the total area of the cortical bones of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the total area of the cortical bones of the first node is greater than the total area of the cortical bones of the second node, it is determined to be the second stage of osteoporosis;
[0015] Optionally, the image feature further includes trabecular separation. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis;
[0016] Optionally, the image features further include total cortical bone area and trabecular separation; when the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, the total cortical bone area of the first node is less than the total cortical bone area of the second node, and the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, the total cortical bone area of the first node is greater than the total cortical bone area of the second node, and the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis;
[0017] Optionally, the cortical bone thickness is an average value of the cortical bone thickness in the image area.
[0018] The method further includes data processing, outlining and segmenting the target area of the CT image to obtain a target image, and extracting a first node image feature and a second node image feature based on the target image;
[0019] Optionally, the segmentation is performed by a threshold method to obtain a target image;
[0020] Optionally, the segmentation is performed by segmenting a segmentation model to obtain a target image;
[0021] Optionally, the segmentation model includes one or more of the following: FCN, U-Net, Mask R-CNN, DeepLab, PSPNet, RefineNet, EncNet, HRNet, OCRNet;
[0022] Optionally, the delineation includes cortical bone delineation and cancellous bone delineation;
[0023] Optionally, the cortical bone delineation selects the top layer of the cancellous bone to extend N layers toward the proximal end of the femur as the bottom layer, and continues to extend M layers toward the proximal end as the top layer, where N is a natural number greater than or equal to 500, and M is a natural number greater than or equal to 50;
[0024] Optionally, the cancellous bone delineation selects the layer where the distal femoral epiphyseal plate disappears as the bottom layer, and extends L layers toward the proximal femur as the top layer, where L is a natural number greater than or equal to 50.
[0025] Replace S3 with S3': input the first node image feature and the second node image feature into the trained model to predict the period of osteoporosis;
[0026] Optionally, the trained model is an animal network model, and the construction process of the animal network model is:
[0027] Obtaining animal CT images at different time points; the animals are animals injected with methylglyoxal for different cumulative days;
[0028] Performing region of interest segmentation on the animal CT images at different time points to obtain regions of interest;
[0029] Extracting node image features based on the region of interest;
[0030] Using the node image features as labels to label the CT images to obtain label data;
[0031] Inputting the labeled data into a neural network for training to obtain a trained model;
[0032] Optionally, the neural network is one or more of the following: Transformer, TransformerXL, CNN;
[0033] Optionally, the imaging features include: cortical bone thickness, total cortical bone area;
[0034] Optionally, the imaging features further include one or more of the following: bone volume fraction, trabecular thickness, trabecular number, trabecular separation, structural pattern index, and bone mineral density.
[0035] Replace S2 with S21: obtain the methylglyoxal content of the subject, input the CT image data and the methylglyoxal content into the neural network for feature learning and feature fusion to obtain the accumulated node image features, and predict the period of osteoporosis based on the accumulated node image features, wherein the accumulated node image features are node image features that contain methylglyoxal content information.
[0036] The S2 is replaced by S22: extracting the first node stiffness and the second node stiffness of the bone based on the CT image, and comparing the first node stiffness with the second node stiffness to obtain the period of osteoporosis;
[0037] Optionally, S22 is replaced by S23: obtaining the content and stiffness of methylglyoxal of the subject to be tested, inputting the stiffness and the CT image into a neural network for feature learning and feature fusion to obtain fusion features, and predicting the osteoporosis period based on the fusion features;
[0038] Optionally, the input to the neural network also includes one or more of the following: stress peak, strain peak, displacement peak.
[0039] The object of the present invention is to provide a computer program product having a computer program or instructions thereon, comprising:
[0040] The computer program or instructions are executed by a processor to implement the above-mentioned method for predicting the osteoporosis period based on the methylglyoxal content.
[0041] The object of the present invention is to provide a computer device, comprising a memory, a processor and a computer program or instruction stored in the memory, comprising:
[0042] The computer program or instructions are executed by a processor to implement the above-mentioned method for predicting the osteoporosis period based on the methylglyoxal content.
[0043] The object of the present invention is to provide a computer-readable storage medium having a computer program or instruction stored thereon, including:
[0044] The computer program or instructions are executed by a processor to implement the above-mentioned method for predicting the osteoporosis period based on the methylglyoxal content.
[0045] Advantages of the present invention:
[0046] The present invention stages osteoporosis by extracting the thickness of cortical bone or the area of cortical bone or the separation of trabeculae from methylglyoxal content and CT images, especially diabetic osteoporosis, and can detect early osteoporosis and reduce the risk of fracture, which has important clinical significance.
[0047] The present invention stages osteoporosis according to the methylglyoxal content and stiffness. As the methylglyoxal content increases, the stiffness of the bones increases. Based on this, osteoporosis can be effectively differentiated in stages, and fractures can be more effectively prevented. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0049] Figure 1 A schematic flow chart of a method for predicting osteoporosis period based on methylglyoxal content provided by an embodiment of the present invention;
[0050] Figure 2 A schematic diagram of a system for predicting osteoporosis period based on methylglyoxal content provided by an embodiment of the present invention;
[0051] Figure 3 A schematic diagram of a device for predicting the period of osteoporosis based on methylglyoxal content provided by an embodiment of the present invention;
[0052] Figure 4The micro-CT images of the MGO group and the control group provided in the embodiment of the present invention, a: CON1group; b: CON2group; c: CON3group; d: MGO1group; e: MGO2group; f: MGO3group;
[0053] Figure 5 A line graph of three-phase bone microstructure parameters of the MGO group and the control group provided in an embodiment of the present invention;
[0054] Figure 6 This is a line graph of the three-phase biomechanical parameters of the MGO group and the CON group provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0055] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.
[0056] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The sequence numbers of the operations, such as S101, S102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.
[0057] Figure 1 A schematic diagram of a method for predicting osteoporosis period based on methylglyoxal content provided by an embodiment of the present invention specifically includes:
[0058] S1: Acquire CT image data of a subject at different time points, where the subject is a subject with methylglyoxal accumulation;
[0059] In one embodiment, the content of methylglyoxal increases with time.
[0060] In a specific embodiment, this experiment uses a micro-CT scanner (SkyScan 1176, Bruker Micro-CT) to perform femoral microstructure imaging and obtain high-resolution images. After the scan is completed, the image is adjusted and segmented using DateViewer software (Bruker, 1.5.6.2). Then, CtAn software (Bruker, 1.17.7.2) is used to delineate the region of interest of the image and calculate various bone microstructure parameters.
[0061] Place the mouse femur steadily on the scanning frame, ensuring that the long axis of the femur is parallel to the long axis of the scanning frame. The scanning parameters are set as follows: voltage is 60kV, current is 417μA, and layer thickness is 8.3μm. The parameters are kept consistent for each scan. After the scan is completed, a high-resolution original image of the mouse femur is obtained.
[0062] S2: extracting the first node image features and the second node image features based on the CT image data at different time points;
[0063] In one embodiment, the first node image feature and the second node image feature are image features selected sequentially at different time points on the same time axis.
[0064] In one embodiment, the method further comprises data processing, delineating and segmenting the target area of the CT image to obtain a target image, and extracting a first node image feature and a second node image feature based on the target image;
[0065] Optionally, the segmentation is performed by a threshold method to obtain a target image;
[0066] Optionally, the segmentation is performed by segmenting a segmentation model to obtain a target image;
[0067] Optionally, the segmentation model includes one or more of the following: FCN, U-Net, Mask R-CNN, DeepLab, PSPNet, RefineNet, EncNet, HRNet, OCRNet;
[0068] Optionally, the delineation includes cortical bone delineation and cancellous bone delineation;
[0069] Optionally, the cortical bone delineation selects the top layer of the cancellous bone to extend N layers toward the proximal end of the femur as the bottom layer, and continues to extend M layers toward the proximal end as the top layer, where N is a natural number greater than or equal to 500, and M is a natural number greater than or equal to 50;
[0070] Optionally, the cancellous bone delineation selects the layer where the distal femoral epiphyseal plate disappears as the bottom layer, and extends L layers toward the proximal femur as the top layer, where L is a natural number greater than or equal to 50.
[0071] In one embodiment, S2 is replaced by S21: the methylglyoxal content of the subject is obtained, the CT image data and the methylglyoxal content are input into the neural network for feature learning and feature fusion to obtain accumulated node image features, and the osteoporosis period is predicted based on the accumulated node image features, wherein the accumulated node image features are node image features containing methylglyoxal content information.
[0072] Optionally, the neural network is one or more of the following: Transformer, TransformerXL, CNN.
[0073] In one embodiment, S2 is replaced by S22: extracting the first node stiffness and the second node stiffness of the bone based on the CT image, and comparing the first node stiffness with the second node stiffness to obtain the stage of osteoporosis.
[0074] In one embodiment, S22 is replaced by S23: obtaining the content and stiffness characteristics of methylglyoxal of the subject, inputting the stiffness characteristics and the CT images into a neural network for feature learning and feature fusion to obtain fusion characteristics, and predicting the osteoporosis period based on the fusion characteristics;
[0075] Optionally, the input to the neural network also includes one or more of the following: stress peak, strain peak, displacement peak.
[0076] In a specific embodiment, image segmentation and adjustment: the original image obtained by micro-CT scanning is imported into DateViewer software, and observed from three perspectives: cross-section, coronal and sagittal. The image position is adjusted by rotation, movement, etc., so that the image is in the best position of the positioning axis under the three perspectives, and the image is saved.
[0077] In a specific embodiment, the region of interest is delineated and the parameter calculation is performed: the image adjusted by the DateViewer software is imported into the CtAn software, and the regions of interest of the cortical bone and the cancellous bone are selected respectively. According to the existing research results, the cancellous bone microstructure is better displayed in the area near the epiphyseal plate, while the cortical bone microstructure is better displayed in the middle of the femur. In this experiment, the cancellous bone parameter calculation range selects the layer where the distal femoral epiphyseal plate disappears as the bottom layer, and extends 50 layers to the proximal femur as the top layer, and performs parameter calculation in the area between the two. When delineating the region of interest of the cancellous bone part, it is ensured that the cancellous bone part is delineated as completely as possible while avoiding the cortical bone. When selecting the cortical bone calculation range, in order to ensure the relative uniformity between the positions of each mid-femoral part, the top layer of the cancellous bone is selected to extend 500 layers toward the proximal femur as the bottom layer, and continues to extend 50 layers to the proximal end as the top layer, and performs parameter calculation in the area between the two. When delineating the region of interest of the cortical bone, try to be close to the outer layer of the cortical bone but do not touch the cortical bone. After the cancellous bone and cortical bone regions of interest were delineated, the CtAn software was used to calculate various parameters. Cancellous bone related parameters include: bone mineral density (BMD, g / cm3), bone volume fraction (BV / TV, %), trabecular thickness (Tb.Th, mm), trabecular separation (Tb.Sp, mm), trabecular number (Tb.N, 1 / mm), structural pattern index (SMI). Cortical bone related parameters include: cortical bone thickness (Ct.Th, mm), total cortical bone area (Tt.Ar, mm2). After the calculation is completed, the parameters are recorded and stored. The micro-CT images of the MGO group and the CON group are shown in the figure. Figure 4 shown.
[0078] S3: comparing the image feature of the first node with the image feature of the second node;
[0079] In one embodiment, S3 is replaced by S3': the first node image feature and the second node image feature are input into the trained model to predict the period of osteoporosis;
[0080] Optionally, the trained model is an animal network model, and the construction process of the animal network model is:
[0081] Obtaining animal CT images at different time points; the animals are animals injected with methylglyoxal for different cumulative days;
[0082] Performing region of interest segmentation on the animal CT images at different time points to obtain regions of interest;
[0083] Extracting node image features based on the region of interest;
[0084] Using the node image features as labels to label the CT images to obtain label data;
[0085] Input the labeled data into a neural network for training to obtain a trained model;
[0086] Optionally, the neural network is one or more of the following: Transformer, TransformerXL, CNN;
[0087] Optionally, the node image features include: cortical bone thickness, total cortical bone area;
[0088] Optionally, the node image features further include one or more of the following: bone volume fraction, trabecular bone thickness, trabecular bone number, trabecular bone separation, structure model index, bone mineral density.
[0089] In a specific embodiment, 72 six-week-old SPF-grade C57BL / 6N male mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and were raised and subsequent modeling operations were carried out in the Experimental Animal Center of the Third Hospital of Hebei Medical University (SYXK(Hebei)2017-003).
[0090] Animal feeding: The experimental mice had free access to food and water. Ordinary mouse food (Experimental Animal Center of Hebei Medical University) was sterilized by high temperature and high pressure and fed as needed. The indoor temperature was 22±2°C, the humidity was 60±5%, and the standard was a 12-hour light / 12-hour dark cycle.
[0091] Experimental reagents: The modeling reagent for the experimental group (MGO group) was a methylglyoxal solution with a concentration of 40%. The injection reagent for the control group (CON group) was a 0.9% sodium chloride aqueous solution.
[0092] After being raised to 20 weeks old, they were randomly divided into six groups: the experimental groups (MGO1 group, MGO2 group, MGO3 group, 14 mice in each group) and the control groups (CON1 group, CON2 group, CON3 group, 10 mice in each group). The MGO group was intraperitoneally injected with a methylglyoxal solution to simulate the abnormal accumulation state of MGO in the body. The CON group was intraperitoneally injected with an equal dose of sodium chloride aqueous solution at the same time. They were grouped according to the length of the cumulative injection time: The MGO1 group and the CON1 group were respectively cumulatively injected for 28 days, the MGO2 group and the CON2 group were respectively cumulatively injected for 42 days, and the MGO3 group and the CON3 group were respectively cumulatively injected for 56 days. The body weight of the mice was weighed and recorded every two weeks for calculating the injection solution dose. The dose of the methylglyoxal solution injected into the MGO group was 50 mg / kg / d, and the dose of the sodium chloride aqueous solution injected into the CON group was the same. The MGO group and the CON group started intraperitoneal injection at the same time, injected continuously for 4 days and rested for 1 day until the cumulative injection days of each group were reached and the modeling work of that group was ended.
[0093] Experimental sampling: After each group of modeling work was completed, the experimental sampling was carried out the next day. Before sampling, the mice were euthanized and killed by intraperitoneal injection of sodium pentobarbital at a dose of 180 mg / kg (overdose). After confirming the death of the mice, they were dissected and the bilateral femurs were removed to ensure the integrity of the femoral structure and to remove the surrounding soft tissue as much as possible. Then the femurs were fully immersed in 4% paraformaldehyde solution and removed for micro-CT scanning 48 hours later.
[0094] S4: The image features include cortical bone thickness. When the cortical bone thickness of the first node is less than that of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than that of the second node, it is determined to be the second stage of osteoporosis.
[0095] In one embodiment, the cortical bone thickness is an average value of the cortical bone thickness in the image area.
[0096] In one embodiment, the image feature is replaced by the total cortical bone area. When the total cortical bone area of the first node is smaller than the total cortical bone area of the second node, it is determined to be the first stage of osteoporosis; when the total cortical bone area of the first node is larger than the total cortical bone area of the second node, it is determined to be the second stage of osteoporosis.
[0097] In one embodiment, the image feature is replaced by trabecular separation. When the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis.
[0098] In one embodiment, the image features are replaced by the total cortical bone area and the trabecular separation; when the total cortical bone area of the first node is smaller than the total cortical bone area of the second node, and the trabecular separation of the first node is larger than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the total cortical bone area of the first node is larger than the total cortical bone area of the second node, and the trabecular separation of the first node is smaller than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis.
[0099] In one embodiment, the imaging features also include the total area of cortical bone. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the total area of the cortical bone of the first node is less than the total area of the cortical bone of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the total area of the cortical bone of the first node is greater than the total area of the cortical bone of the second node, it is determined to be the second stage of osteoporosis.
[0100] In one embodiment, the image features also include trabecular separation. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis.
[0101] In one embodiment, the imaging features also include the total cortical bone area and trabecular separation; when the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, the total cortical bone area of the first node is less than the total cortical bone area of the second node, and the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, the total cortical bone area of the first node is greater than the total cortical bone area of the second node, and the trabecular separation of the first node is less than the trabecular separation of the second node, it is determined to be the second stage of osteoporosis.
[0102] In a specific embodiment, the present invention observes the effect of MGO accumulation on bones over time, and sets up a time axis comparison, namely, MGO1 group, MGO2 group, and MGO3 group. The duration of intraperitoneal injection of MGO is different between the three groups in the MGO group, and the effect of the accumulation time of MGO in the body on bones can be observed. At the same time, a control group, namely, CON1 group, CON2 group, and CON3 group, is set up. The three groups in the CON group are intraperitoneally injected with the same dose of normal saline as the MGO group, and the injection and end time are matched with each group in the MGO group, so that the changes in bone accumulation over time under physiological conditions can be observed. After the modeling is completed, the MGO1 group, MGO2 group, and MGO3 group are compared to observe the differences and change patterns among the three. At the same time, the differences between the CON1 group, the CON2 group, and the CON3 group are compared to observe their change patterns.
[0103] In the MGO group, the parameters of cortical bone were significantly different. Ct.Th was significantly different between MGO1 and MGO2, and between MGO2 and MGO3 (P<0.05). Ct.Th in MGO2 group was larger than that in MGO1 and MGO3 groups. There were no significant statistical differences in the parameters of cancellous bone, as shown in Table 1. In the CON group, BV / TV, SMI, Tb.Th, Tb.N, Tb.Sp, BMD and Ct.Th were significantly different between CON1 and CON3, and between CON2 and CON3 (P<0.05). BV / TV, Tb.N, Tb.Th, BMD and Ct.Th in CON3 group were smaller than those in CON1 and CON2 groups, while SMI and Tb.Sp were larger than those in CON1 and CON2 groups, as shown in Table 2.
[0104] Table 1 Comparison of bone microstructure parameters in the MGO group at different stages
[0105]
[0106] Table 2 Comparison of bone microstructure parameters in the CON group at different stages
[0107]
[0108]
[0109] In a specific embodiment, the present invention not only observes the effect of MGO on bone accumulation over time, but also expects to observe the difference between bone microstructure under MGO intervention and physiological state in the same period. Therefore, the MGO1 group is compared with the CON1 group, the MGO2 group is compared with the CON2 group, and the MGO3 group is compared with the CON3 group.
[0110] The comparison between the MGO1 group and the CON1 group showed that there were statistical differences in all parameters (P<0.05). The BV / TV, Tb.Th, Tb.N, BMD, Ct.Th, and Tt.Ar in the MGO1 group were lower than those in the CON1 group, while the SMI and Tb.Sp were higher than those in the CON1 group, as shown in Table 3. The comparison between the MGO2 group and the CON2 group showed that there were statistical differences in BV / TV, SMI, Tb.Th, Tb.N, Tb.Sp, and BMD between the two groups (P<0.05). The BV / TV, Tb.Th, Tb.N, and BMD in the MGO2 group were lower than those in the CON2 group, while the SMI and Tb.Sp were higher than those in the CON2 group, as shown in Table 4. The comparison between the MGO3 group and the CON3 group showed that there were no statistical differences in all parameters.
[0111] Table 3 Comparison of bone microstructure parameters between MGO1 group and CON1 group
[0112]
[0113]
[0114] Table 4 Comparison of bone microstructure parameters between MGO2 group and CON2 group
[0115]
[0116] In a specific embodiment, the changes of each indicator in the MGO1 group, the MGO2 group, the MGO3 group, the CON1 group, the CON2 group, and the CON3 group are plotted as line graphs according to the time sequence and the group differences to observe the changing trends.
[0117] Over time, BV / TV, Tb.Th, and BMD showed a downward trend in the MGO group and the CON group, and SMI showed an upward trend; Tb.N showed a downward trend in the MGO group, and a slight upward trend and then a downward trend in the CON group; Tb.Sp showed a downward trend first and then an upward trend in the MGO group, and an upward trend in the CON group; Ct.Th showed an upward trend first and then a downward trend in the MGO group, and a downward trend in the CON group; Tt.Ar showed an upward trend first and then a downward trend in both the MGO group and the CON group. Figure 5 shown.
[0118] In a specific embodiment, the present invention is committed to observing the effect of MGO on bone microstructure by imaging methods through micro-CT scanning technology. Micro-CT can provide imaging information with a resolution of up to micrometer level, and has now become the imaging "gold standard" for evaluating bone morphology and bone microstructure. The cancellous bone parameters evaluated by the present invention include: BMD, BV / TV, SMI, Tb.Th, Tb.N, Tb.Sp, and the cortical bone parameters include: Ct.Th and Tt.Ar. The BMD of the cancellous bone part is the tissue bone density. Only the bone mineral density of the trabeculae is calculated, which can more truly reflect the density changes of the trabeculae. BV / TV is the ratio of the two indicators of bone volume (bone volume, BV) and tissue volume (tissue volume, TV). When the selected range of interest is fixed, the ratio of the two can reflect the amount of bone. SMI can reflect the structural information of trabeculae. It is the ratio of rod-shaped structure to lamellar structure in trabeculae. When trabeculae are mainly lamellar structures, SMI is close to 0, and when trabeculae are mainly rod-shaped structures, SMI is close to 3. In osteoporotic state, trabeculae change from lamellar structures to rod-shaped structures, and SMI value increases accordingly. Tb.Th reflects the thickness of trabeculae, Tb.N reflects the number of trabeculae, and Tb.Sp reflects the gaps between trabeculae. The three mainly reflect the spatial morphology and structure of trabeculae in terms of number, thickness and separation. CTAn software uses spherical fitting algorithm to calculate the above three indicators. For the calculation of Tb.Th, CTAn fits the trabeculae themselves into a sphere. When calculating Tb.Sp, CTAn fits the gaps between trabeculae. The calculation process applies the maximization principle, measures the maximum sphere diameter of each voxel, and calculates the average value after summarization. Finally, the average thickness and average separation of trabeculae in the delineated area of interest are obtained. Tt.Ar measures the total area of cortical bone in the cross section within the region of interest, and Ct.Th reflects the average thickness of cortical bone within the region of interest. Both are important parameters for evaluating the microstructure of cortical bone.
[0119] The present invention found that mice treated with intraperitoneal injection of MGO showed significant differences in femoral microstructures from the control group, and the differences existed at the initial stage of experimental intervention. The comparison between the MGO1 group and the CON1 group showed that MGO intervention reduced the mineral content of the femur, the number of trabeculae decreased, the thickness decreased, the separation of trabeculae increased, the proportion of rod-shaped trabeculae increased, the thickness of the cortical bone decreased, and the area of the cortical bone on the cross-section of the region of interest decreased. The above changes suggest that the mice in the MGO group have osteoporotic changes compared with the CON group mice in the same period, and this change occurs in the early stage of MGO accumulation in the body. As the injection time accumulates, the comparison between the MGO2 group and the CON2 group shows that the bone mineral content, bone volume fraction, the number and thickness of trabeculae are still on a downward trend, the separation of trabeculae is still increasing, the proportion of rod-shaped trabeculae is still increasing, and the differences between the above parameters of cancellous bone are still statistically significant.
[0120] In a specific embodiment, in order to more comprehensively evaluate the effect of MGO on the skeletal system, the present invention focuses on observing the changes in biomechanics, especially the effect on the risk of fracture, specifically including:
[0121] Mimics Medical 21.0 (Materialise) was used to build the model and assign values to the model. Geomagic Studio 2013 (Geomagic) was used to smooth the model. 3-Matic 13.0 (Materialise) was used to build the volume mesh. ABAQUS13.1 (Dassault) was used to simulate the model and apply force to obtain the mechanical parameters.
[0122] Convert the mouse femur micro-CT 2D image to DICOM format and import it into Mimics software. In micro-CT images, different tissue structures have different CT values and appear as different grayscale images. Through the threshold segmentation function, select the appropriate grayscale value range (700-5800HU) to remove the soft tissue image and retain only the bone structure image. Then select the new layer, region growth, edit mask and calculation functions in turn to convert the 2D image into a 3D model.
[0123] The generated 3D model was imported into Geomagic software in STL format for smoothing and repairing, removing sharp parts, filling holes and gaps, and ensuring that the femoral model was smooth and complete. The smoothed model was imported into 3-matic software in STP / STEP format for meshing. Meshing is crucial and largely determines the accuracy of subsequent analysis. The size and distribution of the mesh should be ensured to be uniform. This experiment used tetrahedral units with a mesh size of 50μm. Surface meshes and volume meshes were created in sequence and the solid model was exported in INP format.
[0124] The solid model is imported into Mimics software for value assignment. This experiment uses the grayscale value assignment method, that is, the material properties of each voxel are mapped based on the Hounsfield (HU) value of the pixel in the image.
[0125] Import the completed model into ABAQUS software to simulate the three-point bending test. First, establish a coordinate system, with the line connecting the medial and lateral condyles of the femur as the reference, the Y axis parallel to the line, the X axis along the convex surface of the femur and perpendicular to the Y axis, and the Z axis perpendicular to the XY plane. Then select the boundary conditions, take rectangular areas of equal size on the convex surface of the proximal and distal femur as fixed surfaces, and the distance between the center points of the fixed areas is 10mm. On the opposite side of the fixed surface, that is, the concave surface of the femur, apply a load of 5N along the Z axis. After completing the simulated force application, record the stress (Stress, MPa), strain (Strain,%) and displacement (Displacement, mm) values.
[0126] In a specific embodiment, with the combination of FEA and medical imaging, it has been widely used to monitor changes in bone strength and predict fracture risks. FEA can provide stress, strain and displacement parameters of the bone model after the load is applied. Stress is the force per unit area of the sample under test after the load is applied. Strain is the percentage of the change in length of the sample under stress to the original length. Stiffness, also known as stiffness, is the degree of stress that the sample under test bears when a unit strain occurs. The slope of the force-displacement curve can represent stiffness. After a fixed load is applied, the displacement of the model is inversely proportional to the stiffness.
[0127] In a specific embodiment, the changes of each indicator in the MGO1 group, the MGO2 group, the MGO3 group, the CON1 group, the CON2 group, and the CON3 group are plotted in line graphs according to the time sequence and the group differences to observe the change trend, such as Figure 6 As shown in the figure, over time, the strain peak value of the MGO group showed a trend of first decreasing and then increasing, while the strain peak value of the CON group showed a trend of first increasing and then decreasing; the stress peak value of the MGO group showed a trend of first decreasing and then increasing, while the CON group showed a decreasing trend; the displacement peak value of the MGO group showed a trend of first decreasing and then increasing, while the CON group showed a trend of first increasing and then decreasing.
[0128] In a specific embodiment, the present invention uses FEA to simulate the three-point bending mechanical test, mainly evaluating the mechanical properties of the middle part of the backbone, while the main mechanical material in the middle part of the backbone is the cortex, and the trabeculae in the medulla are scarce. The results show that in terms of displacement peak, the MGO2 group is significantly smaller than the CON2 group. Since the applied load value is fixed, the reduction in displacement peak can be understood as an increase in stiffness. Abnormal accumulation of MGO in the body can change the mechanical properties of bones and increase bone stiffness.
[0129] An embodiment of the present invention further provides a computer program product or system, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for predicting the osteoporosis period based on the methylglyoxal content.
[0130] Figure 2 A schematic diagram of a system for predicting osteoporosis period based on methylglyoxal content provided by an embodiment of the present invention specifically includes:
[0131] An acquisition unit is used to acquire CT image data of a subject at different time points, wherein the subject is a subject with methylglyoxal accumulation;
[0132] Feature unit: extracting the first node image feature and the second node image feature based on the CT image data at different time points;
[0133] Comparison unit: compares the first node image feature with the second node image feature;
[0134] Prediction unit: The image features include cortical bone thickness. When the first cortical bone thickness is less than the second cortical bone thickness, it is determined to be the first stage of osteoporosis; when the first cortical bone thickness is greater than the second cortical bone thickness, it is determined to be the second stage of osteoporosis.
[0135] Figure 3 A schematic diagram of a device for predicting osteoporosis period based on methylglyoxal content provided by an embodiment of the present invention specifically includes:
[0136] A memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, any one of the above-mentioned methods for predicting the osteoporosis period based on the methylglyoxal content is executed.
[0137] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, any one of the above-mentioned methods for predicting the osteoporosis stage based on the methylglyoxal content is performed.
[0138] The verification results of this verification example show that assigning inherent weights to indications can improve the performance of the method relative to the default settings. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. In several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways.
[0139] For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, each functional unit in each embodiment of the present invention may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.
[0140] A person skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be implemented by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned medium storage can be a read-only memory, a disk or an optical disk, etc.
[0141] The above is a detailed introduction to a computer device provided by the present invention. For a person skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A method for predicting the stage of osteoporosis based on methylglyoxal content, It is characterized in that include: S1: obtaining CT image data of a subject at different time points, where the subject is a subject with methylglyoxal accumulation; S2: extracting the first node image features and the second node image features based on the CT image data at different time points; S3: comparing the image feature of the first node with the image feature of the second node; S4: The image features include cortical bone thickness. When the cortical bone thickness of the first node is less than that of the second node, it is determined to be the first stage of osteoporosis; when the cortical bone thickness of the first node is greater than that of the second node, it is determined to be the second stage of osteoporosis.
2. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that The image feature is replaced by the total area of cortical bone. When the total area of cortical bone of the first node is less than the total area of cortical bone of the second node, it is determined to be the first stage of osteoporosis; when the total area of cortical bone of the first node is greater than the total area of cortical bone of the second node, it is determined to be the second stage of osteoporosis; optionally, the image feature is replaced by the trabecular separation degree. When the trabecular separation degree of the first node is greater than the trabecular separation degree of the second node, it is determined to be the first stage of osteoporosis; when the trabecular separation degree of the first node is less than the trabecular separation degree of the second node, it is determined to be the second stage of osteoporosis; Optionally, the image features are replaced by the total cortical bone area and the trabecular separation; when the total cortical bone area of the first node is smaller than the total cortical bone area of the second node, and the trabecular separation of the first node is larger than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis; When the total cortical bone area of the first node is greater than the total cortical bone area of the second node, and the trabecular separation degree of the first node is less than the trabecular separation degree of the second node, it is determined to be the second stage of osteoporosis; Optionally, the first node image feature and the second node image feature are image features selected sequentially at different time points on the same time axis; Optionally, the content of methylglyoxal increases with time.
3. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that The image features also include the total area of cortical bones. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the total area of the cortical bones of the first node is less than the total area of the cortical bones of the second node, it is determined to be the first stage of osteoporosis. When the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the total cortical bone area of the first node is greater than the total cortical bone area of the second node, it is determined to be the second stage of osteoporosis; Optionally, the image feature further includes trabecular separation. When the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, and the trabecular separation of the first node is greater than the trabecular separation of the second node, it is determined to be the first stage of osteoporosis. When the cortical bone thickness of the first node is greater than the cortical bone thickness of the second node, and the trabecular separation degree of the first node is less than the trabecular separation degree of the second node, it is determined to be the second stage of osteoporosis; Optionally, the image features further include the total area of cortical bone and the separation of trabeculae; when the cortical bone thickness of the first node is less than the cortical bone thickness of the second node, the total area of the cortical bone of the first node is less than the total area of the cortical bone of the second node, and the separation of trabeculae of the first node is greater than the separation of trabeculae of the second node, it is determined to be the first stage of osteoporosis; When the cortical bone thickness of the first node is greater than that of the second node, the total cortical bone area of the first node is greater than that of the second node, and the trabecular separation of the first node is less than that of the second node, it is determined to be the second stage of osteoporosis; Optionally, the cortical bone thickness is an average value of the cortical bone thickness in the image area.
4. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that The method further includes data processing, outlining and segmenting the target area of the CT image to obtain a target image, and extracting a first node image feature and a second node image feature based on the target image; Optionally, the segmentation is performed by a threshold method to obtain a target image; Optionally, the segmentation is performed by segmenting a segmentation model to obtain a target image; Optionally, the segmentation model includes one or more of the following: FCN, U-Net, Mask R-CNN, DeepLab, PSPNet, RefineNet, EncNet, HRNet, OCRNet; Optionally, the delineation includes cortical bone delineation and cancellous bone delineation; Optionally, the cortical bone delineation selects the top layer of the cancellous bone to extend N layers toward the proximal end of the femur as the bottom layer, and continues to extend M layers toward the proximal end as the top layer, where N is a natural number greater than or equal to 500, and M is a natural number greater than or equal to 50; Optionally, the cancellous bone delineation selects the layer where the distal femoral epiphyseal plate disappears as the bottom layer, and extends L layers toward the proximal femur as the top layer, where L is a natural number greater than or equal to 50.
5. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that Replace S3 with S3': input the first node image feature and the second node image feature into the trained model to predict the period of osteoporosis; Optionally, the trained model is an animal network model, and the construction process of the animal network model is: Obtaining animal CT images at different time points; the animals are animals injected with methylglyoxal for different cumulative days; Performing region of interest segmentation on the animal CT images at different time points to obtain regions of interest; Extracting node image features based on the region of interest; Using the node image features as labels to label the CT images to obtain label data; Inputting the labeled data into a neural network for training to obtain a trained model; Optionally, the neural network is one or more of the following: Transformer, Transformer XL, CNN; Optionally, the imaging features include: cortical bone thickness, total cortical bone area; Optionally, the imaging features further include one or more of the following: bone volume fraction, trabecular thickness, trabecular number, trabecular separation, structural pattern index, and bone mineral density.
6. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that Replace S2 with S21: obtain the methylglyoxal content of the subject, input the CT image data and the methylglyoxal content into the neural network for feature learning and feature fusion to obtain the accumulated node image features, and predict the period of osteoporosis based on the accumulated node image features, wherein the accumulated node image features are node image features that contain methylglyoxal content information.
7. The method for predicting osteoporosis stage based on methylglyoxal content according to claim 1, It is characterized in that The S2 is replaced by S22: extracting the first node stiffness and the second node stiffness of the bone based on the CT image, and comparing the first node stiffness with the second node stiffness to obtain the period of osteoporosis; optionally, the S22 is replaced by S23: obtaining the content and stiffness of methylglyoxal of the subject to be tested, inputting the stiffness and the CT image into the neural network for feature learning and feature fusion to obtain fusion features, and predicting the period of osteoporosis based on the fusion features; Optionally, the input to the neural network also includes one or more of the following: stress peak, strain peak, displacement peak.
8. A computer program product having a computer program or instructions thereon, It is characterized in that include: The computer program or instructions are executed by a processor to implement the method for predicting the osteoporosis stage based on methylglyoxal content as described in any one of claims 1 to 7.
9. A computer device comprising a memory, a processor and a computer program or instructions stored on the memory, It is characterized in that include: The computer program or instructions are executed by a processor to implement the method for predicting the osteoporosis stage based on methylglyoxal content as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program or instructions stored thereon, It is characterized in that include: The computer program or instructions are executed by a processor to implement the method for predicting the osteoporosis stage based on methylglyoxal content as described in any one of claims 1 to 7.
