Method for preparing and driving knee joint orthopedic brace based on 4D printing and brace structure

Abstract

The invention relates to a preparation and driving method of a knee joint orthopedic brace based on 4D printing and a brace structure, the preparation method and the driving method establish a biomechanical model with knee joint deformity characteristics by combining knee joint three-dimensional scanning body surface data and medical images, provide a model basis for brace printing, and improve the knee joint orthopedic brace. Prediction parameters are provided for driving the brace, and the practicability of the brace is improved; the brace structure comprises a first orthopedic plate and a second orthopedic plate which are connected through a first connecting plate and a second connecting plate, assembling is convenient, and walking is not affected after the brace structure is worn; the driving method comprises the steps that a biomechanical model with knee joint deformity characteristics is utilized, then a deep learning algorithm is combined to construct a knee joint orthopedic target angle prediction model, a reference is provided for various morphological changes of the orthopedic brace, the model can be adjusted according to specific stages to continue to be used, efficiency is improved, and the resource utilization rate is increased.

Description

Technical Field

[0001] The present invention belongs to the technical field of orthopedic braces, and specifically relates to a preparation and driving method of a knee joint orthopedic brace based on 4D printing, a brace structure and a driving method. Background Art

[0002] Traditional knee joint orthopedic braces usually adopt a standardized design, which is difficult to fully match the individual differences of patients, resulting in limited orthopedic effects and even possible discomfort or complications. In addition, traditional braces lack intelligence and dynamic adaptability in the manufacturing process and cannot be adjusted according to the biomechanical changes during the patient's rehabilitation process. With the continuous development of 3D printing technology, especially the emergence of 4D printing technology, it provides new possibilities for the personalized customization and dynamic adaptation of orthopedic braces. 4D printing technology combines three-dimensional space printing and shape changes in the time dimension, enabling more complex structures and functions, bringing revolutionary changes to the design and manufacturing of orthopedic braces. Summary of the Invention

[0003] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0004] In view of the following technical problems in the prior art: existing braces are used by replacing different braces according to the patient's conditions at each stage, which is too costly; and the manufacturing steps of braces at each stage are also cumbersome.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a preparation and preparation method of a knee joint orthopedic brace based on 4D printing, including,

[0006] S1, collecting data of the patient, including three-dimensional scanned body surface data of the knee joint and medical image CT or MRI data;

[0007] S2, performing three-dimensional reconstruction on the medical image data to obtain a biomechanical model of the knee joint deformity characteristics;

[0008] S3, obtaining the finite element analysis results of the initial orthopedic brace structure model, and designing the structure of the orthopedic brace according to the analysis results to obtain a 4D printing model;

[0009] S4, using shape memory polymer intelligent materials, directly forming the obtained model by fused deposition modeling technology and assembling the orthopedic brace.

[0010] As a preferred technical solution for the preparation and driving method of a knee orthotic brace based on 4D printing, the biomechanical model for obtaining the characteristics of knee deformity includes:

[0011] Import medical imaging data;

[0012] The target knee joint bones are retained through threshold segmentation, and most soft tissues and noise data are eliminated;

[0013] Three-dimensional reconstruction generates a skeleton model B;

[0014] The optimized three-dimensional skeleton model B is obtained by filtering out noise data and repairing the model integrity. * ;

[0015] Combined with the patient's known deformity characteristic parameters, a biomechanical model M with knee deformity characteristics was established. bio , expressed as:

[0016] M bio =(B * , P, Λ)

[0017] Among them: B * is the optimized bone geometry model mentioned above; P is the physical and mechanical properties of bones and soft tissues; Λ is a set of deformity angles, joint force lines or other deformity characteristic parameters.

[0018] As a preferred technical solution for the preparation and driving method of a knee orthosis brace based on 4D printing, the threshold segmentation includes setting a preliminary threshold interval [I min , I max ];

[0019] For an image voxel position x, if I min ≤I(x)≤I max , then the voxel is considered to belong to bone tissue, and its mask label M(x) = 1; otherwise M(x) = 0; as follows:

[0020]

[0021] According to whether the masked 3D preview image fully displays the patient's lower limb bones and joint structures, if there is a lot of soft tissue or noise residue, adjust [I min , I max ]; if the bone structure is broken or missing, the upper and lower limits of the threshold will be modified accordingly.

[0022] As a preferred technical solution of a preparation and driving method for a knee orthosis based on 4D printing, preprocessing of the finite element analysis of the initial orthosis structural model is carried out, including checking the size of the model and making necessary adjustments:

[0023] According to the actual use situation of the orthosis and biomechanical analysis, corresponding loads and boundary conditions are applied to the outer surface of the orthosis to simulate the actual stress conditions such as gravity, extrusion, and correction moment on the orthosis;

[0024] Based on the finite element analysis results, judge the mechanical strength and stability of the model. Determine the displacement distribution of the model through the displacement nephogram, read the position and magnitude of the maximum displacement, and evaluate the impact of the deformation degree on the use function;

[0025] Optimize the structure of the orthosis for problems such as stress concentration areas and areas with excessive displacement, adjust the shape, position, size, or change the overall structural layout of the orthosis, etc. After the design is completed, a 4D printing model is obtained, which is used for subsequent 4D printing to prepare the knee orthosis.

[0026] As a preferred technical solution of a preparation and driving method for a knee orthosis based on 4D printing, the printing material used for the orthosis is a shape memory polymer intelligent material.

[0027] As a preferred technical solution of a preparation and driving method for a knee orthosis based on 4D printing,

[0028] In step S4,

[0029] Set the nozzle temperature in the printing parameters within the range of the melting point of the selected intelligent polymer material +5°C to ensure that the material can be fully melted without excessive degradation, so as to retain the best molecular chain state for subsequent shape memory characteristics.

[0030] Set the bottom plate temperature in the printing parameters within the range of the glass transition temperature of the selected intelligent polymer material -15°C to -20°C. When printing at this temperature slightly lower than the glass transition temperature, the activity of the intelligent polymer material molecules can be reasonably regulated.

[0031] As a preferred technical solution of a preparation and driving method for a knee orthosis based on 4D printing, it includes calculating the internal / valgus angle based on the knee joint biomechanical model, constructing a knee joint orthopedic target angle prediction model using a deep learning algorithm, and compensating the orthopedic angle to obtain the orthosis structural model;

[0032] Drive the orthosis to deform and regulate its stiffness through external stimulus conditions, and combine the orthosis structural model to change the correction angle of the orthosis to obtain an orthosis with orthopedic ability;

[0033] During the second correction, by changing the malformation data and compensating the correction angle again, the orthosis structural model for the next stage is obtained.

[0034] As a preferred technical solution of a method for preparing and driving a knee orthosis based on 4D printing, the construction of a knee orthosis target angle prediction model using a deep learning algorithm includes:

[0035] Obtain the three-dimensional data of the patient's knee joint and preprocess it, and divide the data set;

[0036] Define the network model and network structure;

[0037] Train the model and construct a loss function;

[0038] Perform cross-validation and optimize the model;

[0039] Output of the model.

[0040] As a preferred technical solution of a method for preparing and driving a knee orthosis based on 4D printing, the obtaining of the three-dimensional data of the patient's knee joint and preprocessing, and dividing the data set includes,

[0041] Based on the obtained three-dimensional data of the patient's knee joint and the clinically known target internal / valgus angle after orthosis, a training data set and a validation data set are formed;

[0042] Take the original malformation angle of the patient's corresponding knee joint and the finally clinically confirmed corrected angle as training labels, and denote the corrected angle as Y i ;

[0043] Perform unified formatting processing on multi-modal data and perform structured encoding on relevant mechanical parameters for input into the network.

[0044] As a preferred technical solution of a method for preparing and driving a knee orthosis based on 4D printing, define the input layer as:

[0045] Denote the preprocessed data as X, where:

[0046] X = {x1, x2,..., x N}

[0047] Each piece of training data x i includes the three-dimensional morphological characteristics and mechanical parameters of the patient's knee joint, and is input into the network in the form of a vector or tensor;

[0048] Define the output layer as:

[0049] That is, output the predicted value of the patient's target internal / valgus angle for regression prediction.

[0050] As a preferred technical solution of a preparation and driving method for a knee orthosis based on 4D printing,

[0051] The loss function L is defined as:

[0052]

[0053] Where: is the predicted angle of the model for the i-th sample; Y i is the true orthopedic angle (label value) of the i-th sample; Ω(W) is the regularization term, and W represents all trainable parameters in the network; α is the regularization coefficient, which is used to balance the data fitting error and the model complexity

[0054] The orthosis is changed in the correction angle again, and the orthosis is driven to deform and regulate the stiffness through external stimulation conditions. Specifically, one or more of thermal excitation, electrical excitation, optical excitation, magnetic excitation, and microwave excitation are used to drive the orthosis to change to the target angle. To drive the orthosis to reach the required correction angle.

[0055] A brace structure prepared according to the foregoing preparation and driving method for a knee orthosis based on 4D printing, including,

[0056] A first orthopedic plate and a second orthopedic plate. One end of the first orthopedic plate is provided with a first connecting plate, and the second orthopedic plate is provided with a second connecting plate. The first connecting plate and the second connecting plate are connected by a splint,

[0057] The first connecting plate is provided with a first deformation section, and the second connecting plate is provided with a second deformation section;

[0058] The end of the first connecting plate is provided with a first gear tooth, the end of the second connecting plate is provided with a second gear tooth, the first gear tooth and the second gear tooth are meshed, the first connecting plate is provided with a first through hole, the second connecting plate is provided with a second through hole, a first pin shaft is inserted through the first through hole and connected to the splint, and a second pin shaft is inserted through the second through hole and connected to the splint;

[0059] The first orthopedic plate is provided with a first strap hole, and the second orthopedic plate is provided with a second strap hole.

[0060] The beneficial effects of the present invention:

[0061] By conducting detailed data collection and three-dimensional reconstruction on patients, a biomechanical model with specific knee joint deformity characteristics of the patient is generated, thereby designing an orthosis highly matched to the individual situation of the patient; a deep learning algorithm is used to construct a knee joint orthosis target angle prediction model, which not only improves the prediction accuracy of the orthosis angle, but also compensates the first orthosis angle according to the prediction result, ensuring that the initial design of the orthosis structure is close to the optimal state, and the 4D printed orthosis can be repeatedly adjusted to adapt to the orthosis stage without re-making, reducing expenses. Brief Description of the Drawings

[0062] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts. Among them:

[0063] Figure 1 It is a schematic diagram of the overall method flow structure in the present invention;

[0064] Figure 2 It is a schematic diagram of the software interface after importing data in the present invention;

[0065] Figure 3 It is a schematic diagram of the threshold interval adjustment interface in the present invention;

[0066] Figure 4 It is a schematic diagram of the masked 3D model after threshold interval adjustment in the present invention;

[0067] Figure 5 It is a schematic diagram of the export parameter setting interface in the present invention;

[0068] Figure 6 It is a schematic diagram of the model optimization and knee joint varus / valgus angle adjustment interface in the present invention;

[0069] Figure 7 It is a schematic diagram of the overall structure of the orthosis;

[0070] Figure 8 It is a schematic diagram of the structure at the connection of the orthosis.

[0071] Reference Numerals: First Round Tooth 102, Second Round Tooth 201a, First Connecting Plate 101, Second Connecting Plate 201, First Perforation 101b, First Pin 101c, Second Perforation 201b, Second Pin 201c, Splint 300, First Orthopedic Plate 100, First Strapping Hole 101d, Second Orthopedic Plate 200, Second Strapping Hole 201d. Detailed Embodiments

[0072] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification.

[0073] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0074] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that are mutually exclusive with other embodiments.

[0075] Furthermore, the present invention is described in detail in conjunction with schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views showing the device structure are enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0076] Embodiment 1

[0077] Referring to Figures 1 to 8 , this embodiment provides a preferred technical solution for the preparation and driving method of a knee joint orthosis based on 4D printing, including,

[0078] Step 1: Collect data from the patient, including three-dimensional scanned body surface data of the knee joint and medical imaging CT or MRI data;

[0079] Step 2: Perform three-dimensional reconstruction on the medical imaging data to obtain a biomechanical model of the knee joint deformity characteristics; calculate the internal / valgus angle based on the knee joint biomechanical model, and compensate for the initial orthosis angle to obtain the initial orthosis structure model;

[0080] Step 3: Analyze the results of the initial orthosis structure model by finite element analysis, and design the structure of the orthosis according to the analysis results to obtain a 4D printing model;

[0081] Step 4: Use shape memory polymer intelligent materials to 4D print the obtained model by fused deposition modeling technology, and assemble the orthosis, including a hinge structure, an angle indicator, and a length adjustment device;

[0082] Step 5: Drive the orthosis to deform and regulate its stiffness through external stimulus conditions, and change the correction angle of the orthosis to meet the continuously changing biomechanical needs of the patient during the rehabilitation process.

[0083] Further, steps 1 and 2 specifically include:

[0084] Collect medical image data of the orthopedic part of the patient. The medical image data is CT or MRI scan data that conforms to the DICOM image file standard format. Reconstruct the orthopedic part of the patient, optimize the model by denoising, etc., and obtain a biomechanical model with knee joint deformity characteristics.

[0085] Use the angle measurement function to measure the varus / valgus angles of the knee joint in the biomechanical model with knee joint deformity characteristics, and obtain the patient's initial varus / valgus angle.

[0086] Adopt deep learning algorithms, such as convolutional neural network (CNN) or recurrent neural network (RNN), to construct a knee joint orthopedic target angle prediction model. Through learning from the training data, automatically extract features related to the orthopedic target angle, and establish a complex mapping relationship between the input parameters and the target angle. Optimize the model using cross-validation and regularization techniques, and output the target varus / valgus angle after orthopedic treatment for this patient. Compensate the initial orthopedic brace structure model. Obtain the initial orthopedic brace structure model.

[0087] Among them, obtaining the biomechanical model with knee joint deformity characteristics includes:

[0088] (1) Data reading and importing

[0089] Obtain medical image data of the orthopedic part (knee joint) of the patient through CT or MRI scan, and store it as a file that conforms to the DICOM standard format.

[0090] For each patient, the medical image data can be denoted as D, where D = {d1, d2,..., d n}, d i is the i-th tomographic scan image; n is the total number of images in this group of CT / MRI scans.

[0091] In the main interface of Mimics software, select "File" → "New projects" in sequence. After selecting the DICOM data folder in the pop-up file browser, click "Next".

[0092] The software automatically reads and displays the basic information of the DICOM sequence images, including the number of image layers, slice thickness, pixel pitch, etc.; if multiple groups of DICOM data need to be added, repeat the above operation.

[0093] Click "Open" and wait for the software to finish processing. After that, Mimics will present a preview of the basic image after the CT or MRI data reconstruction is completed. The integrity and accuracy of the data can be verified through this preview interface.

[0094] (2) Threshold segmentation and mask creation

[0095] In the project management menu on the right side of the Mimics interface, click the Mask option, right-click in a blank space and select "New Mask" to open the "Thresholding" threshold setting interface.

[0096] According to the difference in gray value (or signal intensity) between bone tissue and soft tissue in CT / MRI scans, the initial threshold interval [I min , I max ].

[0097] The corresponding mathematical expression can be defined as: For the image voxel position x, if I min ≤I(x)≤I max , then the voxel is considered to belong to bone tissue, and its mask label M(x) = 1; otherwise M(x) = 0, as follows:

[0098]

[0099] After clicking the confirmation button on the threshold setting interface, the software will display the preliminary three-dimensional bone model obtained according to the threshold interval in real time in the mask 3D preview area on the right.

[0100] According to whether the masked 3D preview image fully displays the patient's lower limb bones and joint structures, if there is a lot of soft tissue or noise residue, it can be adjusted appropriately [I min , I max ]; if the bone structure is broken or missing, the upper and lower limits of the threshold will be modified accordingly.

[0101] Repeat the above adjustment process until the target knee joint bones can be completely and clearly displayed in the preview model and most soft tissues and noise data can be removed as much as possible.

[0102] (3) 3D reconstruction to generate a skeleton model

[0103] After the threshold segmentation effect is satisfactory, you can select the "Calculate 3D" or "3D Preview / Update" option in the mask management interface, and the software will automatically generate a three-dimensional surface model based on the established mask M(x). The model is stored in the form of triangular facets or polygonal meshes, denoted as B.

[0104] B is a three-dimensional geometric representation of the patient's lower limb bones (especially the knee joint area), including the position and morphology information of the bone structure.

[0105] The three-dimensional skeleton model B is exported in common formats such as STL, OBJ or PLY for subsequent denoising and optimization processing.

[0106] In subsequent algorithms or software modules, it is necessary to associate B with the patient's deformity feature information (such as known deformity angles or force line offsets) to obtain a biomechanical model with knee joint deformity characteristics.

[0107] (4) Optimization such as denoising

[0108] Perform denoising or filtering on the three-dimensional model B to remove isolated floating meshes, spikes, or non-bony fine structures.

[0109] Common methods include:

[0110] Mesh Smoothing: Use algorithms such as Laplacian smoothing or bilateral filtering to iteratively update the vertex normal vectors or coordinates to remove local burrs on the model surface;

[0111] Small Component Removal: Through connected domain analysis, if the number of faces or voxels of a certain part of the connected mesh is less than the preset threshold δ, it is regarded as noise or artifact and deleted.

[0112] For holes or faults on the model surface caused by threshold segmentation or scanning defects, topological repair (HoleFilling) or interpolation compensation (Interpolation) can be performed; key regions (such as the distal femur, proximal tibia, etc.) can also be cut or selected according to clinical needs.

[0113] The finally obtained optimized three-dimensional bone model is denoted as B * , which is improved in terms of geometric continuity, surface smoothness, and deformity feature presentation, and can be used for subsequent mechanical simulation and orthopedic parameter calculation.

[0114] Next, add deformity feature information and establish a biomechanical model: Combine the model B * with the patient's known deformity feature parameters (such as the original varus / valgus angle, knee joint force line offset distance, etc.), and perform data annotation or attribute assignment according to the mechanical boundary conditions (such as the mechanical properties of muscles, ligaments, or cartilage).

[0115] In Mimics or subsequent mechanical analysis software (such as ANSYS, Abaqus, etc.), establish a biomechanical model M with knee joint deformity characteristics according to the bone material properties, joint contact surfaces, soft tissue support, and force line directions bio , which can be expressed as: M bio =(B * , P, Λ)

[0116] Where:

[0117] B * is the optimized bone geometry model described above;

[0118] Let \(P\) be the physical and mechanical properties of bones and soft tissues (such as elastic modulus, Poisson's ratio, tendon tension, etc.);

[0119] Let \(\Lambda\) be the set of deformity angles, joint force lines, or other deformity characteristic parameters.

[0120] Through the specific implementation of the above step S1, the following technical effects can be achieved:

[0121] Efficient data import: Using multiple CT / MRI tomographic images in DICOM standard format ensures the integrity and accuracy of the original data of the patient's knee joint;

[0122] Threshold segmentation and mask management: Using threshold strategies in Mimics software to accurately segment bone tissues and quickly generate a preliminary three-dimensional model;

[0123] Denoising and model optimization: Through operations such as smoothing, small component deletion, and hole repair, the geometric continuity and reliability of the model are improved;

[0124] Integration of biomechanical information: Combining the patient's deformity angle and mechanical boundary conditions lays an accurate biomechanical foundation for subsequent orthopedic parameter calculation and 4D printing brace design.

[0125] In summary, step S1, through the reading, threshold segmentation, mask generation, and subsequent denoising and optimization of the patient's knee joint CT / MRI images, can not only obtain a complete and clear three-dimensional bone model, but also construct a biomechanical model \(M\) bio by combining deformity characteristic parameters, providing scientific and accurate data support for the orthopedic brace design and 4D printing process of the subsequent steps (steps S2 - S5). This process is operable, accurate, and innovative, and can be applied to the technical solution of the "Preparation and Driving Method of Knee Orthopedic Brace Based on 4D Printing" involved in the present invention. Specific Embodiment

[0127] I. Data Import

[0128] (I) Prepare medical image data

[0129] Collect the medical image data (CT or MRI scan data) of the part of the patient to be orthopedically treated, ensure that the data format is the DICOM image file standard format, stored in the same folder with complete sequence, clear images, no artifacts and motion interference, and the data quality directly affects the reconstruction accuracy and effect.

[0130] (II) Start Mimics software and import data

[0131] After opening Mimics software, on the main interface, successively select "File" - "New projects". In the popped-up file browser, select the DICOM data folder. After selection, click "Next". The software automatically reads the data and displays the image browsing interface to show the image sequence. You can view the basic information of the images (such as the number of layers, slice thickness, pixel spacing), verify the integrity and accuracy of the data, and can add multiple groups of DICOM data files. Finally, click "Open" and wait for the software to finish processing. At this time, the software will present the image after the CT or MRI scan data reconstruction is completed.

[0132] Refer to Figure 2 It is a schematic diagram of the software interface after importing the data.

[0133] II. Image Processing

[0134] (I) Threshold Adjustment

[0135] Click the mask option under the project management menu on the right. Right-click in the blank area to open the menu bar and select New Mask. Initially adjust the "Thresholding" threshold range and click OK.

[0136] Figure 3 It is a schematic diagram of the threshold range adjustment interface.

[0137] At this time, the view area in the lower right corner is still blank. Click the mask 3D preview icon on the right, and a mask 3D model will appear in the blank view area in the lower right corner. Continuously adjust the threshold range according to the situation of the mask 3D preview. The gray value of the bone tissue is higher than that of the surrounding soft tissues. Initially screen the bone and joint tissues according to the difference in tissue gray values. Move the slider to set the threshold range until the 3D model clearly and completely shows the lower limb bone and joint model of the patient, and at the same time minimize the appearance of other soft tissues as much as possible. Obtain the required 3D model of the patient's bones. However, due to image noise, partial volume effect, etc., subsequent optimization is required.

[0138] Figure 4 In the middle and lower right is a schematic diagram of the mask 3D model after the threshold range adjustment.

[0139] (II) STL File Export

[0140] On the main interface, select "File" - "Export" - "STL". To facilitate subsequent reading and processing in 3D modeling software, here you can choose an ASCII STL format file or a binary STL format file. There are differences between ASCII STL and binary STL files. ASCII is text-based, readable and editable, with an identifier in the header, and the facet data is recorded line by line. The format is regular but redundant, resulting in a large file size. Binary is stored in a compact binary encoding, not directly readable, with reserved information in the header and the data stored sequentially after the facet count. The file is small and fast to read and write. Each is suitable for specific 3D model processing requirement scenarios and can be selected according to the computer configuration and processing needs. After selecting the file format, select the masked 3D model to be exported, click Next to set the export parameters, and click Finish after the settings are completed. The software will convert the 3D model to an STL format file and save it to the specified path according to the settings.

[0141] Figure 5 Schematic diagram of the export parameter setting interface

[0142] III. Model angle calculation

[0143] (I) Model optimization

[0144] Open the STL file of the masked 3D model with ZBrush software, optimize the model, delete the redundant background and soft tissues, and optimize using the software editing tools. The "Smoothing" tool smooths the model surface according to the set iteration times and parameters, reducing jagged edges; the "Remeshing" tool adjusts the model mesh layout and density to improve the topology quality to meet the subsequent analysis and manufacturing requirements.

[0145] Figure 6 Schematic diagram of the model optimization and knee varus / valgus angle adjustment interface

[0146] Among them, the construction of the knee orthopedic target angle prediction model using deep learning algorithms includes:

[0147] 1. Data acquisition and preprocessing

[0148] (1) Obtain training data:

[0149] Based on the three-dimensional data of the patient's knee joint obtained in step S1 (including the body surface scan data and the knee joint biomechanical model information reconstructed from CT / MRI medical images) and the known orthopedic target internal / valgus angles after orthosis (or the true correction angle values verified by surgical and rehabilitation programs), a training data set and a validation data set are formed.

[0150] (2) Data annotation:

[0151] Use the original knee joint deformity angle (such as varus / valgus angle) corresponding to the patient and the finally clinically confirmed corrected angle as training labels. To perform supervised learning in the deep learning model, the corrected angle needs to be denoted as Y i 。

[0152] (3) Data preprocessing:

[0153] Uniformly format multimodal data (such as 3D surface meshes, volume data reconstructed from CT / MRI, patient physiological mechanics parameters, etc.), including size normalization, coordinate alignment, noise filtering, etc., so that the input data meets the input requirements of the subsequent deep learning model. At the same time, perform structured encoding on relevant mechanics parameters (such as force line distribution, soft tissue data) for input into the network.

[0154] 2. Model construction and network structure design

[0155] (1) Deep learning network type:

[0156] According to the actual data type and feature extraction requirements, select deep network structures such as convolutional neural network (CNN) or recurrent neural network (RNN); for cases with more 3D morphology and image data, the CNN structure can be preferentially used to extract spatial or image features.

[0157] (2) Definition of the input layer:

[0158] Denote the preprocessed data as X, where:

[0159] X = {x1, x2,..., x N}

[0160] Each piece of training data x i includes the 3D morphological features of the patient's knee joint (such as point cloud or voxel data) and mechanics parameters (such as joint space, force line distribution, bone density information, etc.), and is input into the network in the form of a vector or tensor.

[0161] (3) Main structure of the network:

[0162] For CNN, it can be composed of multiple convolutional layers, pooling layers, and fully connected layers; for RNN, it can be composed of multiple recurrent units (LSTM or GRU) and fully connected layers. The main role of the network is to automatically learn the deep features highly relevant to the correction angle in the multimodal knee joint data.

[0163] (4) Definition of the output layer:

[0164] The output layer is used for regression prediction, that is, to output the predicted value of the patient's target varus / valgus angle In this technical solution, denoted as It means "predicted target angle after knee joint orthosis of the patient (varus / valgus direction)".

[0165] 3. Model Training and Loss Function Construction

[0166] (1) Training Dataset and Validation Dataset:

[0167] The above dataset is divided into a training set and a validation set according to a certain ratio for model training and performance verification.

[0168] (2) Loss Function:

[0169] The mean squared error (MSE) in regression form is adopted as the basic loss function, and at the same time, a regularization term is combined to suppress overfitting. Suppose there are N sample data in the training set, and the loss function L can be defined as:

[0170] Where:

[0171] is the predicted angle of the i-th sample by the model;

[0172] Y i is the true angle after orthosis (label value) of the i-th sample;

[0173] Ω(W) is the regularization term, and W represents all trainable parameters in the network;

[0174] α is the regularization coefficient, which is used to balance the data fitting error and the model complexity.

[0175] (3) Parameter Update and Optimization:

[0176] The stochastic gradient descent (SGD), Adam or other optimization algorithms are adopted to update the network weights W and biases b by minimizing the loss function L. The update process is as follows:

[0177]

[0178] where η is the learning rate.

[0179] Furthermore, K-fold cross-validation or leave-one-out cross-validation (LOOCV) is adopted to evaluate the training process. The training dataset is divided into K folds. Each time, one fold is taken as the validation set, and the remaining K - 1 folds are taken as the training set. The training is cycled and the validation error is calculated. Finally, the stability and generalization ability of the model are evaluated by synthesizing the results of K validations.

[0180] (2) Regularization and Hyperparameter Selection:

[0181] According to the validation error and the magnitude of the regularization term, select appropriate regularization coefficient α and network structure hyperparameters (such as convolutional kernel size, number of network layers, number of hidden units, learning rate η, etc.) to ensure that the model avoids overfitting or underfitting while maintaining prediction accuracy.

[0182] (3) Early stopping strategy:

[0183] During the training process, monitor the loss on the validation set. When the loss on the validation set does not decrease or shows an upward trend for several consecutive iterations, terminate the training and retain the model parameter combination W * ,b * .

[0184] After the training is completed, the finally obtained "knee joint orthopedic target angle prediction model" can be expressed as:

[0185]

[0186] Where:

[0187] f(·) is the non-linear mapping function represented by the deep neural network;

[0188] W * ,b * are the network parameters optimized through training;

[0189] X is the multi-modal feature (knee joint biomechanical model, three-dimensional data, etc.) corresponding to the current patient (or newly input sample).

[0190] Model output That is, the predicted optimal correction internal / valgus angle for this patient, comprehensively considering the original deformity characteristics, bone geometric structure and soft tissue biomechanical factors obtained by the patient in step S1.

[0191] Take as the target correction angle required for the subsequent orthosis design (including compensation for the initial orthosis angle). In steps S3 - S5, the structure of the 4D printed orthosis can be designed, optimized and dynamically regulated in combination with this prediction result to meet the personalized needs of the patient during the rehabilitation process.

[0192] Through the specific implementation manner of step S2 above, the following technical effects can be achieved:

[0193] Automatic feature extraction: Extract effective features from the three-dimensional data of the knee joint and its biomechanical information through deep learning networks such as CNN or RNN;

[0194] Personalized complex mapping: Map the multi-modal data of the patient to the personalized orthopedic target angle to achieve more accurate internal / valgus angle prediction;

[0195] Model generalization ability: By using techniques such as cross-validation and regularization, overfitting of the model is prevented, and the robustness and accuracy of prediction are improved.

[0196] Linkage with subsequent steps: The output result of this prediction model can be directly used in subsequent steps S3 - S5 to guide the design of the orthosis and the 4D printing forming process.

[0197] Through the implementation of the above deep learning method, the knee orthosis based on 4D printing is more personalized and accurate in angle setting, can fully meet the changing biomechanical needs of patients during the rehabilitation process, and has significant innovation and feasibility.

[0198] Open the STL file of the mask 3D model with 3D modeling software, use the angle measurement function to measure the varus and valgus angles of the knee joint. The initial valgus angle of the patient is X1. Adjust it to the X2 angle according to the treatment plan of the clinician. The skeletal model and body surface data at this time are used as the basis for the orthosis design. Obtain the initial orthosis structural model according to the recalculated skeletal model and body surface data.

[0199] For step S3, it specifically includes:

[0200] Perform preprocessing on the initial orthosis structural model, including checking the model size and making necessary adjustments.

[0201] According to the actual use situation of the orthosis and biomechanical analysis, apply corresponding loads and boundary conditions to the outer surface of the orthosis to simulate the actual stress conditions such as gravity, extrusion, and correction moment on the orthosis.

[0202] Judge the mechanical strength and stability of the model according to the finite element analysis results, determine the displacement distribution of the model through the displacement nephogram, read the position and magnitude of the maximum displacement, and evaluate the impact of the deformation degree on the use function.

[0203] Optimize the structure of the orthosis for problems such as stress concentration areas and areas with excessive displacement, adjust the shape, position, size, or change the overall structural layout of the orthosis, etc. After the design is completed, obtain the 4D printing model, which is used for subsequent 4D printing to prepare the knee orthosis. Specific embodiments

[0205] Perform preprocessing on the obtained three-dimensional model of the fracture and soft tissue injury site, including model size design, hollowing design, and file format conversion.

[0206] Perform mesh division on the model in the finite element analysis software. In principle, appropriate mesh types and sizes should be used for mesh division of the model to ensure the balance between calculation accuracy and efficiency. If there are no problems after inspection, continue with the subsequent steps. If there are problems, re-perform mesh division.

[0207] After importing the model, create a new material and assign the elastic modulus, Poisson's ratio, and density, which are 2000 MPa, 0.34, and 1.25e -9 T / mm 3 . These values are all the basic properties of the PLA polylactic acid material.

[0208] Create a "shell" section with the type of "mean". Select the area to which the interface is to be assigned. In the view window, use the mouse to box select the entire model and click Finish. Select the shell section created in the previous step for the cross-section of the entire model, and fill in the thickness value according to the actual size of the model.

[0209] According to the actual use of the brace and based on biomechanical analysis, apply corresponding loads and boundary conditions to the outer surface of the brace. (Such as simulating the situation of the brace being squeezed, corrective moment, etc.)

[0210] After applying the loads and setting the boundary conditions to the model, create and set an analysis step for static analysis to analyze the simulation. Finally, create a new job, submit it, and wait for the software to run and output the results.

[0211] After the job completes the calculation, the obtained data can be analyzed on the result interface, mainly analyzing the stress nephogram and displacement nephogram obtained from the calculation. In the stress nephogram, determine the stress distribution of the model, and the position and magnitude of the maximum stress can be obtained. Compare it with the yield strength of the material to judge whether there is a risk of yield or failure, and determine whether the model has mechanical strength and stability. In the displacement nephogram, determine the displacement distribution of the model, and read the position and magnitude of the maximum displacement. Evaluate the deformation degree of the model according to the maximum displacement situation, and judge the impact of the generated deformation on the use function.

[0212] Specifically, the printing material used for the orthosis is a shape memory polymer intelligent material, including one or more of polyurethane-based shape memory polymers, polynorbornene-based shape memory polymers, polylactic acid-based shape memory polymers, epoxy resin-based shape memory polymers, styrene-based shape memory polymers, polycaprolactone (PCL)-based shape memory polymers, polymethyl methacrylate (PMMA)-based shape memory polymers, polyetheretherketone (PEEK)-based shape memory polymers, ethylene-vinyl acetate copolymer (EVA)-based shape memory polymers, liquid crystal elastomer shape memory polymers, polycarbonate (PC)-based shape memory polymers, polyvinylidene fluoride (PVDF)-based shape memory polymers, silicone rubber-based shape memory polymers, polypropylene (PP)-based shape memory polymers, nylon (polyamide, PA)-based shape memory polymers, furan resin-based shape memory polymers, polyethylene terephthalate-1,4-cyclohexanedimethanol ester (PETG)-based shape memory polymers, polyimide (PI)-based shape memory polymers, polyvinyl chloride (PVC)-based shape memory polymers, natural rubber-based shape memory polymers, phenolic resin-based shape memory polymers, thermoplastic polyester elastomer (TPEE)-based shape memory polymers, polyarylether nitrile (PEN)-based shape memory polymers, polybutadiene-based shape memory polymers, cellulose-based shape memory polymers, polyurethane acrylate (PUA)-based shape memory polymers, polystyrene (PS)-based shape memory polymers, and polylactic acid (PLA)-based shape memory polymers. The external stimuli include one or more of thermal stimulation, electrical stimulation, optical stimulation, magnetic stimulation, and microwave stimulation. When driving the deformation and stiffness regulation of the orthosis through external stimulation conditions.

[0213] The specific embodiments are as follows.

[0214] Establish a quantitative relationship database for external stimuli and the deformation, stiffness, and correction angle changes of the orthosis. For thermal stimulation, conduct thermal performance tests on the shape memory polymer intelligent material to obtain the deformation curves and stiffness change data of the material under different heating rates and temperature holding times. Place the orthosis in a temperature control device and control the temperature through a program according to the requirements of different stages of the patient's rehabilitation process. Throughout the rehabilitation process, based on the patient's regular medical examination data, such as the changes in the knee joint varus / valgus angle and bone growth, adjust the external stimulation parameters in real time and dynamically change the correction angle of the brace to meet the changing biomechanical needs during the patient's rehabilitation process.

[0215] Specifically, refer to Figure 7 and Figure 8, in the knee orthosis structure based on 4D printing in the present invention, it includes a first orthopedic plate 100 and a second orthopedic plate 200. One end of the first orthopedic plate 100 is provided with a first connecting plate 101, and a second connecting plate 201 is provided on the second orthopedic plate 200. The first connecting plate 101 and the second connecting plate 201 are connected by a clamping plate 300. A first deformation section 102 is provided on the first connecting plate 101, and a second deformation section 202 is provided on the second connecting plate 201;

[0216] The first deformation section 102 and the second deformation section 202 are printed with 4D memory materials.

[0217] A first gear tooth 101a is provided at the end of the first connecting plate 101, a second gear tooth 201a is provided at the end of the second connecting plate 201. The first gear tooth 101a and the second gear tooth 201a are meshed. A first perforation 101b is provided on the first connecting plate 101, a second perforation 201b is provided on the second connecting plate 201. A first pin shaft 101c is inserted through the first perforation 101b to be connected with the clamping plate 300, and a second pin shaft 201c is inserted through the second perforation 201b to be connected with the clamping plate 300;

[0218] A first strap hole 101d is provided on the first orthopedic plate 100, and a second strap hole 201d is provided on the second orthopedic plate 200.

[0219] It should be noted that two pairs of pin holes are correspondingly provided on the clamping plate 300 for the first pin shaft 101c and the second pin shaft 201c to pass through and connect.

[0220] Specifically, both the first orthopedic plate 100 and the second orthopedic plate 200 are in an arc structure, which are respectively used to fit the thigh part and the calf part of the patient. Straps are provided thereon for fixation. The first orthopedic plate 100 and the second orthopedic plate 200 are connected to each other through the first connecting plate 101 and the second connecting plate 201. It should be noted that the meshing of the first gear tooth 101a and the second gear tooth 201a makes the relative rotation between the first connecting plate 101 and the second connecting plate 201 more convenient, facilitating the relative movement between the first connecting plate 101 and the second connecting plate 201 during walking.

[0221] In step S4,

[0222] Set the nozzle temperature in the printing parameters within the range of the melting point of the selected smart polymer material +5°C to ensure that the material can be fully melted without excessive degradation, so as to retain the best molecular chain state for the subsequent shape memory characteristics.

[0223] Set the baseplate temperature in the printing parameters within the range of -15°C to -20°C of the glass transition temperature of the selected intelligent polymer material. When printing at a temperature slightly lower than the glass transition temperature, the activity of the intelligent polymer material molecules can be reasonably regulated. The intermolecular forces can be maintained at a relatively stable level, meeting the stringent requirements of the knee orthosis for material stability and shape accuracy.

[0224] It should be understood that in the development process of any actual implementation, such as in any engineering or design project, a large number of specific implementation decisions can be made. Such development efforts may be complex and time-consuming, but for those of ordinary skill in the art who benefit from this disclosure, without excessive experimentation, the development efforts will be a routine task of design, manufacturing, and production.

[0225] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A preparation and driving method for a knee joint orthosis based on 4D printing, characterized in that: including, S1, collecting data of the patient, including three-dimensional scanned body surface data of the knee joint and medical image CT or MRI data; S2, performing three-dimensional reconstruction on the medical image data to obtain a biomechanical model of the knee joint deformity characteristics; S3, conducting finite element analysis on the initial orthosis structure model, and designing the structure of the orthosis according to the analysis results to obtain a 4D printing model; S4, using shape memory polymer intelligent materials to 4D print the obtained model through the fused deposition modeling technology and assembling the orthosis; S5, calculating the internal / valgus angle based on the knee joint biomechanical model, constructing a knee joint orthopedic target angle prediction model using a deep learning algorithm, and compensating the orthopedic angle to obtain the orthosis structure model; S6, driving the orthosis to deform and regulate the stiffness through external stimulation conditions, changing the correction angle of the orthosis to meet the continuously changing biomechanical needs during the patient's rehabilitation process.

2. The preparation and driving method of the knee joint orthosis based on 4D printing according to claim 1, characterized in that: The biomechanical model for obtaining the knee joint deformity characteristics includes: importing the medical image data; retaining the target knee joint bones through threshold segmentation and removing most soft tissues and noise data; generating a bone model B through three-dimensional reconstruction; Noise data is screened out from the bone model B and the model integrity is repaired to obtain the optimized three-dimensional bone model B * ; Establish a biomechanical model M with knee joint deformity characteristics by combining the known deformity characteristic parameters of the patient bio , expressed as: M bio = (B * , P, Λ) Where: B * is the optimized bone geometric model described above; P is the physical and mechanical properties of bones and soft tissues; Λ is a set of malformation angles, joint force lines, or other malformation characteristic parameters.

3. The preparation and driving method of the knee joint orthosis based on 4D printing according to claim 2, wherein: The threshold segmentation includes setting a preliminary threshold range [I min , I max according to the difference in gray values (or signal intensities) of bone tissue and soft tissue in CT / MRI scans; For the image voxel position x, if I min ≤ I(x) ≤ I max , then this voxel is considered to belong to bone tissue, and its mask label M(x) = 1 is recorded; otherwise M(x) = 0; as follows: According to whether the masked 3D preview image completely displays the patient's lower limb bone and joint structures, if there is a lot of soft tissue or noise residue, then adjust the range of [[I min , I max ; If there are fractures or missing parts in the bone structure, the upper and lower limits of the threshold are modified accordingly.

4. The preparation and driving method of a knee joint orthosis based on 4D printing according to claim 3, wherein: Performing preprocessing on the initial orthosis structure model for finite element analysis, including checking the size of the model and making necessary adjustments: According to the actual use situation of the orthosis and biomechanical analysis, applying corresponding loads and boundary conditions to the outer surface of the orthosis to simulate the actual stress conditions such as gravity, extrusion, and correction moment on the orthosis; Judging the mechanical strength and stability of the model according to the finite element analysis results, determining the displacement distribution of the model through the displacement nephogram, reading the position and magnitude of the maximum displacement, and evaluating the impact of the deformation degree on the use function; Optimizing the design of the orthosis structure for problems such as stress concentration areas and areas with excessive displacement, adjusting the shape, position, size, or changing the overall structure layout of the orthosis, etc. After the design is completed, a 4D printing model is obtained, which is used for subsequent 4D printing to prepare the knee joint orthosis.

5. The preparation and driving method of the knee joint orthosis based on 4D printing according to claim 4, characterized in that: The printing material used for the orthosis is a shape memory polymer intelligent material.

6. The method for preparing and driving a knee joint orthosis based on 4D printing according to claim 5, characterized in that: In step S4, Set the nozzle temperature in the printing parameters within the range of the melting point of the selected intelligent polymer material +5°C to ensure that the material can be fully melted without excessive degradation, so as to retain the best molecular chain state for subsequent shape memory characteristics. Set the build plate temperature in the printing parameters within the range of the glass transition temperature of the selected intelligent polymer material -15°C to -20°C. When printing at a temperature slightly lower than the glass transition temperature, the activity of the intelligent polymer material molecules can be reasonably regulated.

7. The preparation and driving method of the knee orthosis based on 4D printing according to claim 6, characterized in that: including calculating the internal / valgus angle based on the knee joint biomechanical model, constructing a knee joint orthopedic target angle prediction model using a deep learning algorithm, and compensating the orthopedic angle to obtain the orthosis structure model; when correcting again, change the deformity data, compensate the orthopedic angle again to obtain the orthosis structure model for the next stage, and change the correction angle of the orthosis again.

8. The driving method of the knee joint orthosis based on 4D printing according to claim 7, characterized in that: The method for constructing a knee joint orthopedic target angle prediction model using a deep learning algorithm includes: Obtaining three-dimensional data of the patient's knee joint, preprocessing it, and dividing the data set; Defining the network model and network structure; Training the model and constructing a loss function; Performing cross-validation and optimizing the model; Output of the model.

9. The preparation and driving method of the 4D printing-based knee joint orthosis according to claim 8, characterized in that: The step of obtaining three-dimensional data of the patient's knee joint, preprocessing it, and dividing the data set includes: Based on the obtained three-dimensional data of the patient's knee joint and the known target internal / valgus angle after orthopedic correction clinically, forming a training data set and a validation data set; Take the original knee joint deformity angle corresponding to the patient and the finally clinically confirmed corrected angle as training labels, and denote the corrected angle as Y i ; Performing unified formatting processing on multi-modal data and performing structured encoding on relevant mechanical parameters for input into the network.

10. The method for preparing and driving a knee joint orthopedic brace based on 4D printing according to claim 9, wherein: Defining the input layer as: Denoting the preprocessed data as X, where: X = {x1, x2,..., x N} Each piece of training data X i includes the three-dimensional morphological features and mechanical parameters of the patient's knee joint and is input into the network in the form of a vector or tensor; Defining the output layer as: That is, the predicted value of the target varus / valgus angle of the patient is output For regression prediction.

11. The preparation and driving method of the knee joint orthosis based on 4D printing according to claim 10, characterized in that: The loss function L is defined as: Wherein: is the predicted angle of the model for the i-th sample; Y i is the true post-orthopedic angle (label value) of the i-th sample; Ω(W) is the regularization term, and W represents all the trainable parameters in the network; α is the regularization coefficient, which is used to balance the data fitting error and the model complexity.

12. The driving method of the knee joint orthosis based on 4D printing according to claim 11, characterized in that: Driving the orthopedic brace to deform and regulate its stiffness through external stimulus conditions, and combining with the structural model of the orthopedic brace to change the correction angle of the brace to obtain an orthopedic brace with orthopedic ability; driving the orthopedic brace to deform and regulate its stiffness through external stimulus conditions specifically means using one or more of thermal excitation, electrical excitation, optical excitation, magnetic excitation, and microwave excitation to drive the orthopedic brace to change to the target angle.

13. A brace structure of the preparation and driving method of a knee joint orthosis based on 4D printing according to any one of claims 1 to 12, characterized in that: Including: A first orthopedic plate (100) and a second orthopedic plate (200), one end of the first orthopedic plate (100) is provided with a first connecting plate (101), a second connecting plate (201) is provided on the second orthopedic plate (200), and the first connecting plate (101) and the second connecting plate (201) are connected by a clamping plate (300). A first deformation section (102) is provided on the first connecting plate (101), and a second deformation section (202) is provided on the second connecting plate (201); A first gear tooth (101a) is provided at the end of the first connecting plate (101), a second gear tooth (201a) is provided at the end of the second connecting plate (201), the first gear tooth (101a) and the second gear tooth (201a) are meshed, a first through hole (101b) is provided on the first connecting plate (101), a second through hole (201b) is provided on the second connecting plate (201), a first pin shaft (101c) is inserted through the first through hole (101b) to be connected with the clamping plate (300), and a second pin shaft (201c) is inserted through the second through hole (201b) to be connected with the clamping plate (300); A first strap hole (101d) is provided on the first orthopedic plate (100), and a second strap hole (201d) is provided on the second orthopedic plate (200).

Images