Orthopedic rod, correction system and control method

Abstract

This application provides an orthotic rod, an orthotic system, and a control method. The orthotic rod includes a rod body and a sensor array. The sensor array includes multiple magnetic passive sensors. During spinal correction, the orthotic rod provides continuous and stable guiding force, adapting to the laws of human growth and development, guiding the deformed spine to gradually return to normal, reducing the number of surgeries and trauma, and preserving the normal growth space of the spine and thoracic cage. Moreover, the sensor array can monitor changes in spinal shape in real time, objectively quantifying the spinal correction status without relying on medical imaging, eliminating dependence on doctors' clinical experience, and achieving unified and standardized spinal correction assessment. In addition, the use of magnetic passive sensors eliminates the need for complex electrical systems such as power supplies, circuits, and processors inside the orthotic rod, resulting in a simple structure that does not affect the strength of the rod body. It also has good biocompatibility and long-term stability, making it suitable for long-term implantation as a growth brace in the human body.

Description

Technical Field

[0001] This specification relates to the field of medical device technology, specifically to an orthotic rod, a correction system, and a control method. Background Technology

[0002] Spinal deformity refers to the curvature of one or more segments of the spine, deviating from the body's midline and resulting in a C-shape or S-shape. Spinal deformities can be classified as kyphosis and scoliosis. For patients with mild or early-onset deformities, braces such as orthotics are often required. Relying on the growth potential of the spine and bones, the orthotics guides the spine to gradually return to its normal physiological curvature during natural growth.

[0003] During orthopedic treatment using orthopedic rods, the patient's growth and development is a dynamic process. The degree of spinal curvature, the force on the orthopedic rod, and the corrective effect will all change with growth. Therefore, it is necessary to monitor the growth and development of the spine and the progress of correction regularly or irregularly in order to assess the treatment effect and determine whether it is necessary to adjust the orthopedic parameters or replace the orthopedic rod.

[0004] Among related technologies, the monitoring of spinal correction mainly relies on medical imaging, which is complex to operate and has high testing costs. Moreover, the detection of correction depends on the doctor's clinical experience and lacks unified and quantitative objective evidence. Summary of the Invention

[0005] To address at least one of the aforementioned technical problems, this specification provides an orthotic rod, an orthotic system, and a control method.

[0006] In a first aspect, embodiments of this specification provide an orthotic rod for providing guiding force to the spine through its own deformation during spinal correction. The orthotic rod includes: Rod; and A sensor array, comprising a plurality of passive sensors spaced apart along the length of the rod, wherein the passive sensors include magnetic elements for acquiring strain signals generated by the deformation of the rod.

[0007] In some possible implementations, the magnetic element includes magnetic blocks, and the sensor array includes a plurality of magnetic blocks spaced apart along the length of the rod.

[0008] In some possible implementations, the passive sensor includes a magnetostrictive sensing layer, the magnetic field strength of which varies with its own stretching and deformation, and the strain signal includes the magnetic field strength signal of the magnetostrictive sensing layer.

[0009] In some possible implementations, a hollow mounting channel is provided inside the rod along its length, and the magnetic blocks of the plurality of passive sensors are spaced apart in the mounting channel.

[0010] In some possible implementations, a hollow mounting channel is formed inside the rod along its length, the plurality of passive sensors are spaced apart in the mounting channel, and the magnetostrictive sensing layer of each passive sensor is attached to the inner wall of the mounting channel.

[0011] In some possible implementations, the surface of the rod is provided with a plurality of mounting slots spaced apart along the length direction, and the magnetostrictive sensing layer of each passive sensor is attached to the mounting slot, and the outer surface of the magnetostrictive sensing layer is provided with a protective layer.

[0012] In some possible implementations, the surface of the rod is provided with a plurality of mounting slots along its length, and the magnetic blocks of the plurality of passive sensors are respectively disposed in the mounting slots.

[0013] In some possible implementations, the rod body includes a plurality of integrally formed segments along its length, the plurality of segments including a apical region located in the middle of the length direction, edge regions located at both ends of the length direction, and a main bending region located between the apical region and the edge regions; The peak temperature of austenite phase transformation in the apical region is lower than that in the main bending region, and the peak temperature of austenite phase transformation in the main bending region is lower than that in the edge region.

[0014] In some possible implementations, the sensor density in the apical region is greater than that in the main curve region, and the sensor density in the main curve region is greater than that in the edge region.

[0015] Secondly, the embodiments of this specification provide a correction system, including: The orthopedic rod described in any of the above embodiments; A data acquisition terminal, comprising a magnetic induction device, wherein the magnetic induction device is used to acquire strain signals from a passive sensor on the orthopedic bar via electromagnetic induction; The data processing terminal includes a controller configured to receive the strain signal and generate a detection result based on the strain signal.

[0016] In some possible implementations, the data acquisition terminal further includes a memory and a first communication module, the memory being used to store strain signals acquired by the magnetic sensing device; The data processing terminal further includes a second communication module, which is communicatively connected to the first communication module and is used to receive strain signals sent by the first communication module.

[0017] Thirdly, this specification provides a control method applied to the correction system described in any of the above embodiments, executed by the controller of the correction system, the method comprising: Acquire the sensor data array for the current period, the sensor data array including sensor data for multiple consecutive unit durations, each unit duration including the strain signal of each passive sensor; The orthopedic rod detection result for the current cycle is determined based on the sensor data array.

[0018] In some possible implementations, determining the orthopedic rod detection result for the current period based on the sensor data array includes: The strain rate of each passive sensor on the orthotic bar is determined based on the sensor data array. The strain trend of the orthotic bar in the current cycle is determined based on the strain rate. Based on the similarity between the strain trend of the orthotic rod in the current period and the strain trend of the orthotic rod in at least one historical period, the orthotic rod recovery result in the current period is determined, and the detection result includes the orthotic rod recovery result; or, based on the strain trend of the orthotic rod in the current period and a pre-constructed change trend template, the detection result is determined according to the matching result.

[0019] In some possible implementations, determining the orthopedic rod detection result for the current period based on the sensor data array further includes: For each passive sensor, based on the sensor data array, the change in the strain signal of the passive sensor within a preset time window is determined; Based on the comparison between the change and a preset threshold, the local strain parameters of the orthotic rod are determined, and the orthotic rod detection results include the local strain parameters.

[0020] In some possible implementations, determining the orthopedic rod detection result for the current period based on the sensor data array includes: The sensor data array is input into a pre-trained correction model, which includes a spatial branch module and a temporal branch module. The correction model extracts the spatial features of the sensor data array through the spatial branch module and the temporal features of the sensor data array through the temporal branch module. Based on the spatial features and the temporal features, it outputs multi-task detection results, which include: the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and the correction effect parameters.

[0021] Fourthly, embodiments of this specification provide a control device applied to the correction system of any of the foregoing embodiments, the device comprising: The data acquisition module is configured to acquire a sensor data array for the current period, the sensor data array including sensor data for multiple consecutive unit durations, each unit duration including strain signals from each passive sensor; The result determination module is configured to determine the orthopedic rod detection result for the current period based on the sensor data array.

[0022] In some possible implementations, the result determination module is configured to: The strain rate of each passive sensor on the orthotic bar is determined based on the sensor data array. The strain trend of the orthotic bar in the current cycle is determined based on the strain rate. Based on the similarity between the strain trend of the orthotic rod in the current period and the strain trend of the orthotic rod in at least one historical period, the orthotic rod recovery result in the current period is determined, and the detection result includes the orthotic rod recovery result; or, based on the strain trend of the orthotic rod in the current period and a pre-constructed change trend template, the detection result is determined according to the matching result.

[0023] In some possible implementations, the result determination module is configured to: For each passive sensor, based on the sensor data array, the change in the strain signal of the passive sensor within a preset time window is determined; Based on the comparison between the change and a preset threshold, the local strain parameters of the orthotic rod are determined, and the orthotic rod detection results include the local strain parameters.

[0024] In some possible implementations, the result determination module is configured to: The sensor data array is input into a pre-trained correction model, which includes a spatial branch module and a temporal branch module. The correction model extracts the spatial features of the sensor data array through the spatial branch module and the temporal features of the sensor data array through the temporal branch module. Based on the spatial features and the temporal features, it outputs multi-task detection results, which include: the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and the correction effect parameters.

[0025] Fifthly, embodiments of this specification provide a storage medium storing computer instructions for implementing the methods of any of the foregoing embodiments.

[0026] Sixthly, embodiments of this specification provide a computer program product for implementing any of the foregoing embodiments of the method.

[0027] The orthotic rod described in this specification includes a rod body and a sensor array. The sensor array includes multiple magnetic passive sensors. During spinal correction, the orthotic rod provides continuous and stable guiding force, adapting to the laws of human growth and development, guiding the deformed spine to gradually return to normal, reducing the number of surgeries and trauma, and preserving the normal growth space of the spine and thoracic cage. Moreover, the sensor array can monitor changes in spinal shape in real time, objectively quantifying the spinal correction status without relying on medical imaging, eliminating dependence on doctors' clinical experience, and achieving unified and standardized spinal correction assessment. In addition, the use of magnetic passive sensors eliminates the need for complex electrical systems such as power supplies, circuits, and processors inside the orthotic rod, resulting in a simple structure that does not affect the rod's strength. It also has good biocompatibility and long-term stability, making it suitable for long-term implantation as a growth brace in the human body. Attached Figure Description

[0028] Figure 1 This is a side view of the orthotic rod and the spine.

[0029] Figure 2 This is a schematic diagram of the orthopedic rod structure in an example embodiment of this specification.

[0030] Figure 3 This is a partial structural diagram of an orthopedic rod according to an example embodiment of this specification.

[0031] Figure 4 This is a schematic cross-sectional view of an orthopedic rod according to an example embodiment of this specification.

[0032] Figure 5 This is a schematic cross-sectional view of the orthopedic rod according to another example embodiment of this specification.

[0033] Figure 6 This is a schematic diagram of different segments of the orthopedic rod in an example embodiment of this specification.

[0034] Figure 7 This is a block diagram of the correction system structure of an example embodiment of this specification.

[0035] Figure 8 This is a block diagram of the correction system structure of an example embodiment of this specification.

[0036] Figure 9 This is a flowchart of a control method according to an example embodiment of this specification.

[0037] Figure 10 This is a flowchart of a control method according to an example embodiment of this specification.

[0038] Figure 11 This is a flowchart of a control method according to an example embodiment of this specification.

[0039] Figure 12 This is a flowchart of a control method according to an example embodiment of this specification.

[0040] Figure 13 This is a schematic diagram of the structure of a correction model according to an exemplary embodiment of this specification.

[0041] Figure 14 This is a block diagram of the control device structure according to an example embodiment of this specification. Detailed Implementation

[0042] Spinal deformity refers to a curvature of one or more segments of the spine that deviates from the midline of the body, resulting in a C-shaped or S-shaped spinal crown. Spinal deformities are divided into scoliosis and kyphosis. Scoliosis refers to lateral curvature of the spine in the coronal plane, while kyphosis refers to excessive backward curvature of the spine in the sagittal plane.

[0043] In clinical practice, the treatment plan for spinal deformities needs to be formulated according to the severity of the deformity and the patient's growth and development. For patients with large deformity angles (e.g., Cobb angle exceeding 40°) or severe deformities, surgical osteotomy to decompress and restore spinal cord and nerve function is often required.

[0044] For patients with milder deformities or those that develop early, invasive spinal osteotomy is unnecessary; correction can be achieved simply by inserting orthotic braces. This approach leverages the growth potential of the spine and bones in children and adolescents, using the biomechanical guidance of the orthotic brace to gradually restore the deformed spine to its normal physiological curvature during natural growth. This effectively controls the progression of the deformity while maximizing the preservation of spinal growth and development space, avoiding the crankshaft effect caused by early spinal fusion surgery and its adverse effects on the development of the patient's thoracic cavity, heart, lungs, and other internal organs.

[0045] During orthopedic treatment using orthopedic rods, the patient's growth and development is a dynamic process. The degree of spinal curvature, the force on the orthopedic rod, and the corrective effect will all change with growth. Therefore, it is necessary to monitor the growth and development of the spine and the progress of correction regularly or irregularly in order to assess the treatment effect and determine whether it is necessary to adjust the orthopedic parameters or replace the orthopedic rod.

[0046] Currently, monitoring of spinal correction primarily relies on medical imaging techniques such as X-rays and CT (Computed Tomography). By taking images of the spine, doctors assess changes in spinal deformity angles and the effectiveness of the correction. These methods are not only complex and costly, but also pose radiation risks, making frequent testing unsuitable. Furthermore, the assessment of correction is highly dependent on the doctor's clinical experience, making it subjective and difficult to achieve precise, real-time monitoring of the spinal correction process. This lack of data support for adjusting correction parameters and determining the timing of orthotic rod replacement hinders the standardization and precision of corrective treatment.

[0047] To address this technical problem, this specification provides an orthotic rod, an orthotic system incorporating the orthotic rod, and a control method. The aim is to optimize the structure and design of the orthotic rod, utilize a built-in sensor array to monitor the growth status of the spine in real time, with little or no reliance on medical imaging, and objectively and accurately reflect the recovery status of the spine through sensor array data, providing a unified and quantitative objective basis for evaluating the effect of spinal correction.

[0048] The orthotic rod, correction system, and control method described in this specification are applicable to scenarios involving spinal correction treatment for patients with spinal deformities. However, it is understood that this specification addresses improvements to the medical device itself and does not involve any intervention in the diagnosis or treatment of human diseases. For example, regarding the orthotic rod and correction system, this specification protects the hardware structure and connectivity of the medical device; regarding the control method, this specification protects the data processing procedures of the computer program, strictly adhering to the relevant provisions of patent law regarding the subject matter of patent protection.

[0049] In some embodiments, this specification provides an orthotic rod that, during spinal correction, uses its own deformation to provide guiding force to the spine, thereby guiding the spine to grow back to its normal physiological curvature. The orthotic rod provided in this specification can be applied to scoliosis correction, kyphosis correction, or other spinal deformity correction scenarios, without limitation.

[0050] For example Figure 1 A schematic diagram of the lateral structure after orthotic rod implantation is shown. See also Figure 1As shown, during spinal correction, firstly, several metal screws 20 are sequentially implanted at the pedicle positions on the deformed spinal segment. The screw tails of the screws 20 are equipped with connecting structures for connecting the orthotic rod 10. Then, the deformed curvature and apical position of the deformed spine are measured and determined. Based on the curvature of the deformed spine, the orthotic rod 10 is bent from its initial straight state to form a curved shape that matches the curvature of the deformed spine, ensuring that the orthotic rod 10 matches and conforms to the corresponding segment of the spine. Afterward, the bent orthotic rod 10 is placed longitudinally along the spine into the connecting structures of each screw 20, and the orthotic rod 10 is locked and fixed to each screw 20 using a locking device, achieving stable installation of the orthotic rod 10 on the spine.

[0051] During the corrective growth of the spine, the elastic restoring force of the orthotic rod 10 applies a continuous and uniform guiding force to the deformed spine, guiding the spine to gradually return to the normal physiological curvature during subsequent growth and development, while preserving the growth space between the spine and the thoracic cage.

[0052] In this specification, in order to continuously monitor the growth of the spine, a sensor array can be set on the body of the orthopedic rod 10. The sensor array includes multiple passive sensors arranged at intervals along the length of the rod.

[0053] Passive sensors are sensors that do not contain their own power source and do not rely on a power supply circuit. It is understandable that after the orthotic rod 10 is implanted in the human body, its service life can reach several months to several years. Therefore, active devices are difficult to provide long-term and stable data acquisition, especially for the orthotic rod 10, which contains complex mechanical structures and electronic circuits, and is basically unusable in clinical applications.

[0054] like Figure 2 As shown in the embodiment of this specification, the orthotic rod 10 includes a rod body 11 and a sensor array. The sensor array includes a plurality of passive sensors 30 spaced apart along the length direction of the rod body 11. In this example, the passive sensors 30 included in the sensor array are distributed throughout the entire rod body 11 along the length direction. The passive sensors 30 are used to collect strain signals generated by the deformation of the rod body.

[0055] It is understandable that during the process of the orthopedic rod 10 guiding the spine to recover, the rod 11 gradually returns from a bent state to a straight state. Thus, the sensor array can collect the strain signal generated by the recovery of the rod 11 during the orthopedic process, and the strain signal can represent the strain of the rod 11.

[0056] In some implementations, the passive sensor may be a magnetic passive sensor, for example, the passive sensor may include a magnetic element, which is used as the sensing end of the sensor.

[0057] For example, in one example, the magnetic component can be a magnetic block, and the shape of the magnetic block can be any shape such as a cylinder, sphere, or cube; this specification does not impose any restrictions on this. Multiple passive sensor magnetic blocks can be pre-embedded inside the rod 11 and spaced apart along the length of the rod 11. During the deformation recovery process of the rod 11, as the shape of the rod 11 changes, the arrangement of the multiple magnetic blocks changes accordingly. Therefore, by observing the strain signal obtained from the shape changes of the multiple magnetic blocks, the recovery status during the spinal correction process can be reflected. This process is described below in this specification.

[0058] In other examples, the magnetic component can be a magnetostrictive sensing layer. The characteristic of a magnetostrictive sensing layer is that the arrangement of magnetic domains within the material changes during stretching and deformation, thereby altering the surface magnetic field strength. Therefore, the magnetostrictive sensing layer can be attached to the orthopedic rod 10. As the degree of bending of the rod 11 changes, the magnetostrictive sensing layer undergoes stretching and deformation, thus changing its own magnetic field strength. Based on this principle, by observing the change in magnetic field strength of each magnetostrictive sensing layer, strain signals from the sensor array can be acquired. These strain signals reflect the recovery status during spinal correction. This process is described below in this specification.

[0059] In some embodiments, the sensor array of the orthotic rod 10 includes S passive sensors. The value of S can be selected according to the specific requirements of the scenario, and this specification does not limit it. For example, in one example, the number of passive sensors S can be selected based on factors such as the length of the rod 11 and the sensor layout density. In some possible embodiments, S can be a positive integer greater than or equal to 3.

[0060] In some embodiments, the S passive sensors can be evenly distributed along the length of the rod 11, meaning the distance between two adjacent passive sensors is a fixed value. In other embodiments, the S passive sensors can be evenly distributed along the length of the rod 11, meaning the distance between two adjacent passive sensors is not fixed, as will be described below.

[0061] In some exemplary embodiments of this specification, the magnetic component is exemplified by a magnetostrictive sensing layer. This layer can be made of magnetic materials, including but not limited to terbium-dysprosium-iron, iron-cobalt alloys, or iron-gallium alloys. The magnetostrictive sensing layer employs a thin-film structure, which can be adhered to the surface of the rod 11 or disposed within the rod 11. Due to the thinness of the thin film structure, it does not affect the structural strength of the rod 11 after being adhered to it. For example, in some embodiments, the thickness of the magnetostrictive sensing layer can be 10 μm to 100 μm.

[0062] When the orthopedic rod 10 undergoes bending deformation, the magnetostrictive sensing layer will stretch and deform along with the rod 11, thereby changing the magnetic domain arrangement inside the magnetic material and the magnetic field strength of the passive magnetic sensor itself.

[0063] For example, in the early stages of spinal correction, the rod 11 is in a state of significant bending, causing the magnetostrictive induction layer to undergo significant tensile deformation due to the bending. This results in a significant shift in the arrangement of its internal magnetic domains, causing the magnetic field strength signal to deviate significantly from the reference value. As the spine gradually grows and repositions, the rod 11 gradually straightens, and the strain gradually decreases. Consequently, the deformation of the magnetostrictive induction layer decreases, the arrangement of its internal magnetic domains recovers, and the magnetic field strength signal deviates less from the reference value.

[0064] The external acquisition terminal collects the magnetic field strength signals of each passive sensor 30 through the principle of electromagnetic induction, thereby obtaining the strain signal of the sensor array. Then, through data processing, the spinal correction status is obtained, which objectively reflects the growth status of each segment of the spine, the degree of correction, whether the orthotic rod is abnormal or needs to be adjusted or replaced, etc. The spinal correction effect can be quantitatively monitored without medical imaging.

[0065] In other exemplary embodiments of this specification, the magnetic component is exemplified by a magnetic block. The magnetic block can be a permanent magnet made of materials such as neodymium iron boron, samarium cobalt, alnico, or ferrite, or it can be made of metals such as iron or iron alloys. Multiple magnetic blocks can be embedded in the rod 11 through pre-filling or drilling. For example, along the length of the rod 11, S mounting channels, the same number as the number of passive sensors, are spaced apart inside the rod 11. The shape of each mounting channel matches the magnetic block, thus ensuring a one-to-one correspondence between each magnetic block and a mounting channel.

[0066] When the orthopedic rod 10 undergoes bending deformation, the arrangement of the multiple magnetic blocks inside the rod body 11 will also change accordingly. For example, in the early stage of spinal correction, the rod body 11 is in a large bending state, so the multiple magnetic blocks inside also exhibit a bent arrangement. As the spine gradually grows and repositions, the rod body 11 gradually tends to straighten, the strain gradually decreases, and the arrangement of the multiple magnetic blocks gradually changes from bending to straight.

[0067] The external acquisition terminal collects the arrangement position of each magnetic block through the principle of electromagnetic induction, thereby obtaining the strain signal of the sensor array. Then, through data processing, the spinal correction status is obtained, which objectively reflects the growth status of each segment of the spine, the degree of correction, whether the orthotic rod is abnormal or needs to be adjusted or replaced, etc. The spinal correction effect can be quantitatively monitored without medical imaging.

[0068] The solution described in this specification uses a magnetic passive sensor, eliminating the need for complex electrical systems such as power supplies, circuits, and processors inside the orthotic rod. It relies on the principle of magnetic induction to detect the deformation of the orthotic rod, resulting in a simple structure and strong stability, making it suitable as a growth support for long-term implantation in the human body.

[0069] As described above, in the embodiments of this specification, during spinal correction, the orthotic rod provides continuous and stable guiding force, adapting to the laws of human growth and development, guiding the deformed spine to gradually return to normal, reducing the number of surgeries and trauma, and preserving the normal growth space of the spine and thoracic cage. Furthermore, the sensor array can monitor changes in spinal shape in real time, objectively quantifying the spinal correction status without relying on medical imaging, eliminating dependence on doctors' clinical experience, and achieving unified and standardized spinal correction assessment. In addition, the use of magnetic passive sensors eliminates the need for complex electrical systems such as power supplies, circuits, and processors inside the orthotic rod, resulting in a simple structure that does not affect the rod's strength, good biocompatibility, and long-term stability, making it suitable for long-term implantation as a growth support in the human body.

[0070] In some embodiments, the body 11 of the orthotic rod 10 is made of a nickel-titanium (NiTi) shape memory alloy, which combines biocompatibility and superelasticity. Compared to titanium alloy, the nickel-titanium alloy is more elastic, capable of recovering more than 8% of its deformation, while titanium alloy only recovers 0.2%. Therefore, in spinal orthopedic scenarios, the rod body can have a larger deformation and can provide a continuous and stable restoring force to the spine after deformation, reducing the frequency of orthotic rod replacement.

[0071] The cross-section of the rod 11 is generally circular, and some rods 11 have an elliptical cross-section. Taking the circular cross-section rod 11 as an example, the cross-sectional radius is generally 2mm to 5mm. The length can be set according to the specific needs of the spinal segment. For example, the length of the rod 11 can generally be adapted to spinal segments of 80mm to 650mm.

[0072] The rod 11 can be processed using forging, 3D printing, and other forming technologies. Through heat treatment or 3D printing, part or all of the rod 11 can be made to have a specific austenitic phase transformation temperature. Therefore, during spinal correction, the rod 11 can be bent to fit the curvature of the spine as needed. After implantation, relying on the superelasticity of the austenitic phase, it continuously outputs a stable elastic recovery force as a guiding force for spinal growth, achieving precise correction and adapting to spinal growth activities.

[0073] In some embodiments of this specification, the magnetic component of the passive sensor is exemplified by a magnetostrictive sensing layer. The assembly method between the magnetostrictive sensing layer and the rod 11 can include various approaches. For example, in one example, the magnetostrictive sensing layer can be pre-embedded inside the rod 11. In another example, a hollow pipe can be formed inside the rod 11, and the magnetostrictive sensing layer can be attached to the inner wall of the pipe. In yet another example, an installation groove can be formed on the outer surface of the rod 11, allowing the magnetostrictive sensing layer to be attached during installation.

[0074] For example Figure 3 A partial structural schematic diagram of the rod 11 according to some embodiments of this specification is shown. In this example, a plurality of mounting slots 12 are spaced apart along the length direction on the surface of the rod 11. Each mounting slot 12 is used to mount one passive sensor 30, so the number of mounting slots 12 corresponds to the number of passive sensors 30. It can be understood that, for the sake of clarity, Figure 3 Only one mounting slot 12 and a passive sensor 30 are shown in the image.

[0075] In this example, the mounting slot 12 can be a slot formed on the outer surface of the rod 11 by a one-piece molding (such as 3D printing) process or a milling process. For example, in one example, the depth of the mounting slot 12 can be set to 0.2mm~0.8mm. The magnetostrictive sensing layer of the passive sensor 30 is attached to the mounting slot 12, thereby adhering to the inner wall of the mounting slot 12. For example, in one example, the magnetostrictive sensing layer can be attached to the inner wall of the mounting slot 12 using a biocompatible adhesive. When the rod 11 is bent and deformed, the magnetostrictive sensing layer attached to the mounting slot 12 can undergo tensile deformation accordingly.

[0076] In some embodiments, after the magnetostrictive sensing layer is attached to the mounting slot 12, a protective layer can be provided on the outer surface of the magnetostrictive sensing layer. The protective layer can be made of biocompatible material, which on the one hand encapsulates the sensor to prevent the sensor from shifting or falling off in the rod body and ensures the stability and reliability of the sensing signal, and on the other hand isolates the magnetostrictive sensing layer from damage caused by human tissue fluid, cells and mechanical friction.

[0077] exist Figure 3In this implementation, by pre-setting mounting slots on the surface of the rod 11, the passive sensor 30 is embedded and fitted to the rod 11. This ensures that the sensor and the rod deform synchronously, improving the detection accuracy of the strain signal, while also preventing the sensor from protruding from the rod surface, reducing stimulation and wear on surrounding soft tissues. Furthermore, the slot structure facilitates positioning, installation, and processing, and the slot depth can be adapted to the thickness of the magnetostrictive sensing layer. This achieves integrated sensor-rod integration with the rod structure with almost no weakening of the rod's structural strength, resulting in a more stable overall structure that is better suited for long-term implantation in the body.

[0078] For example Figure 4 The assembly structure of the magnetostrictive layer and the rod in other embodiments of this specification is shown. Figure 4 In the example, to clearly show the internal structure of the rod, the rod 11 is cut open along its length and half of the rod structure is hidden.

[0079] exist Figure 4 In the example implementation, a hollow mounting channel 13 can be formed along the length direction inside the rod 11. The mounting channel 13 is used to assemble the magnetostrictive sensing layer of the passive sensor 30. The number of mounting channels 13 can correspond to the number of passive sensors 30, so that one magnetostrictive sensing layer of the passive sensor 30 is disposed in each mounting channel 13. It can be understood that, for the sake of clarity, Figure 4 Only one mounting pipe 13 and a passive sensor 30 are shown in the diagram.

[0080] In this example, the mounting pipe 13 can be created using 3D printing technology. In some embodiments, the mounting pipe 13 can adopt a hollow pipe structure with a circular cross-section. The diameter of the circular cross-section can be selected according to specific needs, as long as the side wall area of ​​the mounting pipe 13 is sufficient to fit the magnetostrictive induction layer. For example, in one example, the inner diameter of the mounting pipe 13 can be set to 0.5mm~1.5mm.

[0081] The magnetostrictive sensing layer of the passive sensor 30 is attached to the inner wall of the mounting pipe 13, thereby adhering to the inner wall of the mounting pipe 13. For example, in one example, the magnetostrictive sensing layer can be attached to the inner wall of the mounting pipe 13 using a biocompatible adhesive. When the rod 11 is bent and deformed, the magnetostrictive sensing layer attached to the mounting pipe 13 can undergo tensile deformation accordingly.

[0082] exist Figure 4In this implementation, the sensor is integrated into the rod by creating a pipe inside the rod body. This internal placement of the sensor does not occupy surface space, does not compromise the integrity of the rod body surface, and avoids contact between the sensor material and human soft tissue, thus improving the stability and safety of long-term implantation. Furthermore, the magnetostrictive sensing layer adheres to the inner wall of the pipe, enabling more precise transmission of rod bending deformation and improving the accuracy of strain signal detection.

[0083] In some embodiments of this specification, the magnetic component of the passive sensor is a magnetic block, which can be installed inside the rod 11 by pre-filling or drilling.

[0084] For example Figure 5 The assembly structure of the magnetic block and the rod body in some embodiments of this specification is shown. Figure 5 In the example, to clearly show the internal structure of the rod, the rod 11 is cut open along its length and half of the rod structure is hidden.

[0085] See Figure 5 As shown, hollow mounting channels 13 can be formed along the length of the rod body 11. These channels 13 are used to assemble the magnetic blocks of the passive sensor 30. The number of mounting channels 13 corresponds to the number of passive sensors 30, so that one magnetic block of a passive sensor 30 is placed in each mounting channel 13. It is understood that this arrangement is for clarity of illustration. Figure 4 Only one mounting pipe 13 and a passive sensor 30 are shown in the diagram.

[0086] In this example, the mounting pipe 13 can be created using a 3D printing process. In some embodiments, the mounting pipe 13 can be a hollow pipe structure with a circular cross-section. Correspondingly, the magnetic block can be a cylindrical structure, so that the magnetic block can be placed adaptably in the mounting pipe 13.

[0087] Of course, the installation method of the magnetic block and the rod is not limited to Figure 5 As shown, any other suitable method can also be used, such as forming an assembly groove by drilling holes on the surface of the rod 11 and embedding the magnetic block into the assembly groove of the rod 11. This specification will not elaborate further on this.

[0088] It is worth noting that most spinal orthotic rods in related technologies are designed with a single material and uniform mechanical properties, maintaining consistent overall stiffness and elasticity. However, research has revealed significant regional differences in spinal deformities, for example... Figure 5 Different regions of the deformed spine are shown. For example... Figure 5As shown, the two ends of the curve of the deformed spine are defined as the marginal zone, the middle of the curve as the apical zone, and the area between the marginal zone and the apical zone as the main curve zone. Among them, the apical zone is the most severely deformed area with the greatest curvature and requires the strongest guiding force, while the guiding force required for the main curve zone and the marginal zone decreases in that order.

[0089] In related technologies, orthotic rods with uniform stiffness and elasticity are prone to problems such as insufficient guiding force in the apical region and excessive stress concentration in the peripheral region. This not only affects the overall orthopedic effect but may also lead to problems such as screw loosening and rod fatigue fracture. Especially for children and adolescents in the growth and development stage, the single homogeneous mechanical properties cannot achieve precise and personalized mechanical guidance, resulting in unsatisfactory correction efficiency.

[0090] Based on this, in some embodiments of this specification, the rod body 11 of the orthotic rod 10 is divided into multiple segments based on the region of the deformed spine, for example, according to Figure 6 The spinal region shown divides the orthotic rod 10 into multiple segments, including the marginal zone, the main curve zone, and the apical zone. Combined with... Figure 5 As shown, in some embodiments, the rod 11 includes edge regions located at both ends in the length direction, a apical region located in the middle in the length direction, and a main bending region located on both sides of the apical region connecting the edge regions.

[0091] In some embodiments of this specification, the rod 11 is made of nickel-titanium shape memory alloy, and its segments are integrally formed by forging or 3D printing. For example, taking forging as an example, after the rod 11 is integrally formed, different heat treatment processes can be applied to each segment to achieve different austenitic phase transformation peak temperatures (Ap). Similarly, taking 3D printing as an example, during the additive printing process of the rod 11, different printing parameters can be set for each segment to achieve different austenitic phase transformation peak temperatures at each stage. By setting different Ap values ​​for different segments, the restorative force requirements of different areas of spinal correction can be adapted.

[0092] In materials science, austenite and martensite are phases that exhibit different mechanical properties at different temperatures. Austenite, also known as the A phase, is present in nickel-titanium alloys. In the austenite phase, nickel-titanium alloys exhibit high hardness, high strength, and superelasticity, automatically springing back after the removal of external force without permanent deformation. In the martensite phase, the material is soft, easily bent, and retains its bent state after deformation.

[0093] In this specification, the austenitic phase transformation peak temperature (Ap) varies for different segments of the rod 11. For example, the austenitic phase transformation peak temperature in the apex region is lower than that in the main bend region, and the austenitic phase transformation peak temperature in the main bend region is lower than that in the edge region. In one example, the austenitic phase transformation peak temperature in the apex region is 0℃~5℃, the austenitic phase transformation peak temperature in the main bend region is 8℃~12℃, and the austenitic phase transformation peak temperature in the edge region is 15℃~35℃.

[0094] As discussed above, the lower the peak temperature of the austenitic phase transformation, the earlier the material transforms into austenite at body temperature (36°C), and thus the greater the elastic force it provides. Conversely, the higher the peak temperature of the austenitic phase transformation, the later the material transforms into austenite at body temperature (36°C), and thus the smaller the elastic force it provides.

[0095] Therefore, in this specification, the peak temperature of the austenite phase transformation in the apical region of the rod 11 is set to be the lowest. This results in the apical region exhibiting superelasticity at body temperature, providing greater elastic force that matches the maximum corrective force required by the apical region of the spine. The peak temperature of the austenite phase transformation in the main curvature region and the peripheral region increases sequentially, thus the elastic force provided at body temperature decreases sequentially, matching the magnitude of the corrective force required by the main curvature region and the peripheral region of the spine, satisfying spinal correction while maintaining a certain degree of flexibility.

[0096] Specifically, the Ap (apex) region has the lowest Ap, exhibiting super-elasticity at body temperature, providing the greatest corrective force and addressing the correction challenges in the most severely deformed areas. The main curvature region has a moderate Ap, with super-elastic spacing and strong resilience, achieving stable correction while avoiding stress concentration. The peripheral region has the highest Ap, offering relatively better flexibility, reducing stimulation to the vertebral body and soft tissues. Furthermore, the peripheral region's highest Ap also indicates lower resistance to deformation, maintaining a certain degree of mobility and enhancing the effectiveness of spinal correction and repair.

[0097] In the embodiments described in this manual, the orthotic rod is designed in segments for different regions, combined with differentiated austenitic phase transformation temperature control, to achieve a gradient distribution of the orthotic force, precisely matching the guiding force requirements of different segments of the deformed spine. Compared to traditional orthotic rods with single mechanical properties, the segmented corrective force of the orthotic rod in this manual is more precise, improving the orthotic effect and implantation safety, adapting to the spinal growth characteristics of children and adolescents, and achieving personalized and precise spinal orthodontics.

[0098] In the above-mentioned scheme of zonal design for the orthotic bar, the orthotic force generated by different segments of the bar is different. During the long-term spinal orthopedic process, there are significant differences in the degree of curvature and strain rate of each segment, so the monitoring accuracy of the strain signal also varies.

[0099] Therefore, in some embodiments of this specification, the S passive sensors included in the sensor array are not uniformly distributed along the length of the rod 11, but rather different sensor densities are set for different segments. It can be understood that sensor density refers to the number of passive sensors distributed per unit length of the rod 11, and sensor density can be represented by the spacing between adjacent sensors.

[0100] In some implementations, the apex region provides the greatest corrective force, exhibits the greatest stress variation, and requires the highest monitoring accuracy. Therefore, the sensor density in the apex region is set to the maximum; for example, in one example, the spacing between adjacent sensors in the apex region is 3 mm to 10 mm. The stress variation in the main bending region is relatively gradual, resulting in a lower sensor density compared to the apex region; for example, in one example, the spacing between adjacent sensors in the main bending region is 6 mm to 10 mm. The edge region has even less corrective force, thus allowing for a lower sensor density layout; for example, in one example, the spacing between adjacent sensors in the edge region is 10 mm to 15 mm. Of course, those skilled in the art will understand that the distribution of the sensor array on the rod is not limited to this example, and this specification will not elaborate further.

[0101] As can be seen from the above, in the embodiments of this specification, by setting different sensor density distributions for different rod segments, it is possible to accurately detect the top cone area with large stress changes, while reducing the number of sensors for segments with less stress changes, thereby reducing the processing difficulty and signal interference caused by redundant sensors. While accurately reflecting the strain trend of each segment, it improves the overall stability and reliability of the system, and provides strong data support for zoned precise correction.

[0102] Based on the aforementioned orthotic bar, this specification provides an orthopedic system, such as... Figure 7 As shown, the system includes the orthopedic rod 10 of any of the above embodiments, as well as a data acquisition terminal 40 for acquiring sensor data and a data processing terminal 50 for processing data.

[0103] The data acquisition terminal 40 can be a handheld mobile terminal. Figure 8 The architecture of the data acquisition terminal 40 is shown, such as... Figure 8 As shown, the data acquisition terminal 40 may include a magnetic sensing device, which is a sensing element used to monitor the passive sensor signal on the orthopedic rod 10 through the principle of electromagnetic induction.

[0104] In some embodiments, when the magnetic component of the passive sensor of the orthotic bar 10 is a magnetic block, the magnetic sensing device can detect the spatial position of each magnetic block through the principle of non-contact electromagnetic induction, and determine the strain signal based on the spatial position of each magnetic block. The strain signal represents the deformation state of the orthotic bar 10.

[0105] In other embodiments, when the magnetic component of the passive sensor of the orthotic bar 10 is a magnetostrictive sensing layer, the magnetic sensing device can non-contactly detect the magnetic field strength of each magnetostrictive sensing layer and determine the strain signal based on the magnetic field strength of each magnetostrictive sensing layer. This strain signal represents the deformation state of the orthotic bar 10. For example, the magnetic sensing device generates a high-frequency alternating excitation magnetic field. When the magnetic sensing device approaches the magnetostrictive sensing layer, the magnetostrictive sensing layer of the sensor generates a magnetic response under the action of the excitation magnetic field. The magnetic sensing device senses this magnetic response and generates a corresponding electromagnetic signal. Then, the magnetic sensing device obtains the strain signal of the orthotic bar 10 based on the electromagnetic signal.

[0106] In this specification, when the data acquisition terminal collects data from the orthopedic rod 10 implanted in the human body, the magnetic sensing device can be brought close to or attached to the back of the human body, corresponding to the position of the human spine. The distance between the magnetic sensing device and the back of the human body shall not exceed 5cm to ensure the accuracy of the detection signal.

[0107] During a single data acquisition, the user can hold the data acquisition terminal and sequentially collect data from each passive sensor on the orthotic bar 10 along the human spine from top to bottom (or from bottom to top), obtaining the single acquisition data M=[M1, M2, M3, ..., M S M represents sensor data, and S represents the number of passive sensors.

[0108] In some implementations, the data acquisition terminal also includes a memory, which serves as a data storage area to store the sensor data collected by the magnetic sensing device locally. For example, in one embodiment, the data acquisition terminal collects sensor data from a patient once a day and stores the collected sensor data in the local memory. Then, at preset detection cycle intervals, the data stored in the local memory is uploaded to the data processing terminal.

[0109] The data acquisition terminal also includes a first communication module, which is a functional module used for data communication with the data processing terminal. For example, in some embodiments, the data processing terminal is provided with a second communication module, and the data acquisition terminal connects to the second communication module in the data processing terminal through the first communication module to realize data transmission.

[0110] In some implementations, the first communication module and the second communication module can be either wired or wireless communication modules. For example, in wired communication, the first and second communication modules can be data cables based on the USB (Universal Serial Bus) protocol or the RS232 serial port protocol. In wireless communication, the first and second communication modules can be wireless communication modules based on protocols such as Bluetooth, WiFi, or NFC.

[0111] To enable long-term monitoring of spinal deformities in patients, corresponding testing cycles can be pre-defined, with each cycle consisting of T units of time. For example, in one scenario, a 7-day testing cycle would be used, meaning each cycle would consist of T = 7 units of time, with each unit of time being 1 day.

[0112] In an exemplary scenario, a user can collect sensor data once a day through the data acquisition terminal 40, obtaining a single set of sensor data M = [M1, M2, M3, ..., M...]. S The data is collected and stored in the current memory. The sensor data array, which contains T data collections over a period of T units, is represented as follows: in, Let T represent the sensor data array in the i-th cycle, where T represents the unit duration of that cycle, and S represents the number of passive sensors. This represents the data from the S-th sensor within the T-th unit of time.

[0113] Subsequently, the data acquisition terminal 40 transmits the sensor data array stored in its memory to the data processing terminal 50 via the first communication module. The data processing terminal 50 includes a controller, which can detect the spinal orthopedic condition and obtain the detection results by executing the control method steps described below, based on the received sensor data array.

[0114] Based on the above embodiments, this specification provides a control method that is applied to the correction system of any of the above embodiments and is executed by the controller of the data processing terminal of the correction system.

[0115] like Figure 9 As shown, in some embodiments, the control method exemplified in this specification includes: S810: Obtain the sensor data array for the current period.

[0116] As mentioned above, spinal correction is a long-term process. In order to monitor the entire correction process regularly, the process can be divided into independent testing cycles according to a preset time length. For example, in one example, 7 days, 14 days, 30 days, etc. can be used as a testing cycle, and each testing cycle includes multiple units of time, with each unit of time being 1 day.

[0117] For example, in the previous example, taking a detection cycle consisting of T units of time as an example, the sensor data array collected by the data acquisition terminal in each cycle... That is: In this array, each row of elements represents the data collected from S passive sensors per unit time. The array consists of T rows, which means it contains a total of T×S data values.

[0118] In some implementations, the sensor data array for the current period is acquired. After that, you can also... Preprocessing is performed, including processes such as noise reduction and normalization.

[0119] For example, in one instance, noise can be denoised for the data from each passive sensor, for instance, by using a moving average filter, as shown below: (1) In formula (1), Let W represent the denoised signal, W represent the sliding window length, t represent the current time point, and j represent the i-th sensor.

[0120] For example, in another example, the denoised data can be normalized and represented as: (2) In formula (2), This represents the historical minimum value of the j-th sensor. This represents the historical maximum value of the j-th sensor.

[0121] S820: Determine the orthopedic rod detection result for the current cycle based on the sensor data array.

[0122] In the embodiments described in this specification, after obtaining the sensor data array through the aforementioned data acquisition process, the data processing terminal can perform data analysis based on the sensor data array to obtain the orthopedic detection results for the current period.

[0123] The testing of spinal correction effectiveness in this specification may include at least one of the following aspects: 1) Does the strain trend of the orthotic bar in the current cycle match or closely resemble the trend in historical cycles? It is understood that spinal correction is a slow and continuous process. During the spinal growth and recovery process, the strain trend of the orthotic bar should remain stable and consistent. By comparing it with historical trends, we can reflect whether the correction process is smooth and whether there are any abnormalities in the orthotic bar's recovery.

[0124] 2) Does the strain trend of the orthotic bar in the current cycle conform to the standard template for spinal correction? It is understood that there are standard strain patterns in the clinical correction and recovery of spinal deformities. A standard trend template can be pre-set based on the patient's age and deformity type. This template can cover the standard template for the strain trend of the orthotic bar in each cycle throughout the entire correction process. By matching the strain trend of the orthotic bar in the current cycle with the standard template, it can be determined whether the spinal recovery in the current cycle is within a reasonable range.

[0125] 3) Whether the orthotic rod experiences sudden stress changes in a short period of time. It is understandable that the normal growth and recovery of the spine will not involve drastic stress changes. Therefore, sudden stress changes in a short period of time often indicate abnormal events, such as breakage or detachment of the orthotic rod. By detecting sudden stress changes in the orthotic rod in a short period of time, the effective status of the orthotic rod can be reflected.

[0126] Figures 10 to 12 The specific methods and procedures for detecting the results of the three types of orthotic rods mentioned above are shown below. Figures 10 to 12 Each will be explained separately. Figures 10 to 12 In the example, the magnetic component of the passive sensor of the orthopedic rod 10 is illustrated using a magnetostrictive sensing layer as an example.

[0127] like Figure 10 As shown, in some embodiments, the control method exemplified in this specification, the process of determining the orthotic rod detection result based on a sensor data array, includes: S910. Determine the strain rate of each passive sensor on the orthotic bar based on the sensor data array.

[0128] Based on the foregoing, the sensor data array for the current period can be represented as follows: In this matrix, each column represents sensor data from a specific sensor on the orthotic bar over multiple consecutive unit time periods. In the embodiments described in this specification, the corresponding strain rate can be calculated for each passive sensor, expressed as follows: (3) In formula (3), This represents the strain rate of the j-th sensor. This represents the stress signal of the j-th sensor in the last unit of the current cycle. This represents the stress signal of the j-th sensor in the first unit of time of the current cycle.

[0129] Using the above formula (3), the strain rate of each passive sensor on the orthotic bar in the current cycle can be calculated based on the sensor data array.

[0130] S920. Determine the strain trend of the orthotic bar in the current cycle based on the strain rate.

[0131] In this specification, the strain trend of the orthotic rod is the trend vector V formed by the strain rate of each sensor. curr , represented as: V curr =[ , , ..., ].

[0132] S930. Based on the similarity between the strain trend of the orthotic rod in the current cycle and the strain trend of the orthotic rod in at least one historical cycle, determine the orthotic rod recovery result for the current cycle.

[0133] As mentioned above, the strain trend of the orthotic rod in the current cycle can be compared with the trends of one or more historical cycles to determine whether the recovery of the orthotic rod is stable.

[0134] In some implementations, after calculating the strain trend of the orthotic bar for each cycle using the aforementioned S910-S920 steps, it can be stored in local memory. Thus, taking the current cycle as an example, after calculating the strain trend V of the orthotic bar for the current cycle... curr Subsequently, the historical strain trend V of the orthotic bar from the previous one or several cycles can be read from the memory. prev And calculate the similarity between the two.

[0135] For example, in some implementations, V can be calculated using a cosine similarity algorithm. curr With V prev The similarity is expressed as: (4) In formula (4), sim represents the cosine similarity, and V curr V represents the current cycle trend. prev This indicates the trend of a historical cycle (such as the previous cycle). After calculating the similarity between the strain trend of the orthotic rod in the current cycle and that of the historical cycle, the recovery result of the orthotic rod in the current cycle can be determined based on the similarity value.

[0136] For example, in some implementations, a similarity threshold can be pre-set for the strain trend of the orthotic rod, and the calculated similarity score (sim) can be compared with the similarity threshold. If it is greater than or equal to the similarity threshold, it indicates that the strain trend of the orthotic rod in the current period is consistent with the historical trend, and the orthotic rod is in a stable recovery state. Conversely, if it is less than the similarity threshold, it indicates that the strain trend of the orthotic rod in the current period has deviated from the historical trend, and the orthotic rod has failed or is abnormal, requiring further detection through medical imaging or other means.

[0137] As can be seen from the above, in the embodiments of this specification, by comparing the strain trend of the orthotic bar in the current cycle with that in the historical cycle, it is possible to accurately reflect whether the spinal correction process is proceeding smoothly and whether the recovery state of the orthotic bar is abnormal or ineffective, thus providing objective and accurate test results for clinical use.

[0138] like Figure 11 As shown, in some embodiments, the control method exemplified in this specification, the process of determining the orthotic rod detection result based on a sensor data array, includes: S1010. Determine the strain rate of each passive sensor on the orthotic bar based on the sensor data array.

[0139] S1020. Determine the strain trend of the orthotic bar in the current cycle based on the strain rate.

[0140] For the process of S1010 to S1020, refer to the above S910 to S920, and will not be repeated here.

[0141] S1030. Match the strain trend of the orthopedic rod in the current cycle with the pre-constructed trend template, and determine the detection result based on the matching result.

[0142] In this specification, a standardized trend template for the recovery process of the orthotic bar can be pre-constructed. This template includes the expected strain trend of the orthotic bar in each cycle throughout the entire orthopedic process. That is, the trend template serves as a reference for the spinal correction effect. If the actual strain trend of the orthotic bar matches the trend template, it indicates that the spinal correction effect of the current cycle meets expectations. Conversely, if the actual strain trend of the orthotic bar does not match the trend template, it indicates that the spinal correction effect of the current cycle has deviated from expectations and further testing with medical imaging is required.

[0143] In some implementations, the trend template can be generated based on clinical big data statistics, combined with effective strain trend data from patients of the same age and type, and then fitted after data screening. In some implementations, the trend template can be constructed by simulating the strain pattern of the orthotic rod during the correction process using simulation software. In some implementations, the trend template can be standardized by medical staff setting expected trend ranges for different correction stages based on clinical experience. This specification does not limit these implementations.

[0144] In other implementations, the trend of change in one or more future cycles can be predicted using a neural network model based on the historical strain trends of the same patient, as a trend template, as described below.

[0145] In the current cycle detection, after calculating the strain trend V of the orthotic rod for the current cycle through the aforementioned method, the change trend V of the current cycle can be compared with the change trend template V. std For matching, the matching algorithm can use Euclidean distance, which is expressed as: (5) In formula (5), d represents the Euclidean distance. The trend template is used to represent the change trend. The matching result between the strain trend of the orthotic rod in the current cycle and the standard trend template can be calculated by formula (5). This matching result represents the detection result of spinal correction.

[0146] As can be seen from the above, in the embodiments of this specification, a standardized trend template is pre-constructed. By matching the current period's correction status with the trend template, an objective and real-time assessment of the spinal correction effect can be achieved, accurately reflecting whether the current spinal recovery status meets expectations.

[0147] like Figure 12 As shown, in some embodiments, the control method exemplified in this specification, the process of determining the orthotic rod detection result based on a sensor data array, includes: S1110. For each passive sensor, based on the sensor data array, determine the change in the strain signal of the passive sensor within a preset time window.

[0148] As mentioned above, the sensor data array includes sensor data from each sensor on the orthotic bar over a continuous period of T units of time. In this embodiment of the specification, it is necessary to detect whether there are any sudden changes in the data based on the sensor data of each sensor over a continuous period of T units of time. For example, if there is a sudden change in the data of one or more sensors, it indicates that the stress of the orthotic bar has changed suddenly, and the orthotic bar is likely to be damaged or detached and fail.

[0149] Based on this, in the scheme described in this specification, after obtaining the aforementioned sensor data array, for each sensor, the change in the sensor over a continuous period of T time units can be calculated using a sliding window method, expressed as follows: (6) In formula (6), Let L represent the short-term change of the j-th sensor, and L represent the preset time window, the value of which can be preset. Based on the above formula (6), the change of each sensor in each window can be calculated by means of a sliding window, and the change can reflect the degree of change of the sensor data.

[0150] S1120. Based on the comparison results between the change and the preset threshold, determine the local strain parameters of the orthotic rod.

[0151] In this specification, a preset threshold can be set for the change in strain signal from the sensors. This preset threshold represents the critical value at which the strain is in a safe state. For any j-th sensor, if the change in strain of that sensor is greater than or equal to the preset threshold, it indicates that the strain of the rod at that sensor location exceeds the safe range, and a stress mutation occurs at that location. Conversely, if the changes in strain of all sensors are less than the preset threshold, it indicates that the strain of the entire rod does not exceed the safe range, and the orthotic rod is in a stable recovery state.

[0152] Therefore, the local strain parameters of the orthotic rod can be obtained by comparing the change of each sensor in the sensor array with the preset threshold. These local strain parameters can reflect the strain of the rod at each sensor position, and the orthotic rod detection results can include these local strain parameters.

[0153] As can be seen from the above, in the embodiments of this specification, by using a preset time window to capture short-term stress changes of the sensor, the strain of the orthotic rod can be accurately reflected, and the location of the change can be precisely located. This provides reliable data for timely investigation and resolution of abnormalities such as loosening, displacement, breakage, and detachment of the orthotic rod, thereby improving the safety and reliability of the spinal correction process.

[0154] In some embodiments of this specification, a correction model for spinal deformity correction can be constructed based on a deep neural network. The correction model can output multi-task detection results based on the sensor data array of the current cycle. The multi-task detection results include, but are not limited to: the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and correction effect parameters.

[0155] Figure 13 The architecture of the correction model for some embodiments of this specification is shown; see [link to documentation]. Figure 13As shown, the correction model includes an input layer, an output layer, a spatial branch module, and a temporal branch module. The input layer serves as the input to the correction model, and the input data is the sensor data array of the current period obtained by the aforementioned data acquisition terminal, using the sensor data array as the input tensor.

[0156] In this specification, the correction model employs a network architecture that integrates spatial and temporal branches. The spatial branch module can be a network layer module based on a Convolutional Neural Network (CNN), while the temporal branch module can be a network layer module based on a Long Short-Term Memory (LSTM) network. The spatial branch module is responsible for extracting the spatial distribution characteristics of the sensor data array, such as the strain relationship between different segments. The temporal branch module is responsible for extracting temporal features representing trends, rates of change, and growth patterns over time.

[0157] The spatial branch module extracts spatial features, and the temporal branch module extracts temporal features, which are then fused and fed into the output layer. The output layer includes a fully connected layer for outputting multi-task detection results. These results may include the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and correction effect parameters, as shown in the example above.

[0158] In this specification, after the correction model is constructed, it needs to be trained. The model training process is described below.

[0159] First, training samples need to be collected. Historical sensor data and clinical data from patients of the same age and type of spinal deformity can be selected as sample data. After preprocessing the sample data, a continuous multi-cycle sensor data array is constructed as the training samples. The label information corresponds to the multi-task output of the correction model, including, for example, the strain trend of the next cycle, the actual replacement time of the orthotic rod, and the parameters of the correction effect assessed clinically. Then, during training, the training samples are input into the correction model in batches. The spatial branch module and the temporal branch module extract features and fuse them respectively, and output the multi-task detection results through the output layer.

[0160] In some implementations, a multi-task joint model training strategy is adopted to construct a multi-task loss function, expressed as: (7) In formula (7), L represents the total loss. This represents the trend prediction loss term. This indicates the time loss due to orthotic rod replacement. This indicates the loss of corrective effect. The weight values ​​for each loss item can be set.

[0161] During model training, the total loss value L for each loss term is calculated using formula (7). Then, based on this total loss value L, the model parameters of the correction model are adjusted and optimized using the backpropagation algorithm, thus achieving one round of iterative training. By repeating the above process, the correction model is trained in multiple rounds until the model converges, completing the training of the correction model.

[0162] During the model prediction phase, the sensor data array for the current period is obtained through the aforementioned data acquisition terminal. The sensor data array is then input into the trained correction model. The spatial branch module of the correction model extracts spatial features, and the temporal branch module extracts temporal features. After feature fusion, the output layer outputs multi-task detection results, including: the strain trend of the orthotic rod in the next period, the orthotic rod replacement time, and correction effect parameters.

[0163] It is worth noting that, in some implementations, the strain trend of the orthotic rod output by the correction model for the next cycle can be used as a template for the change trend in the next cycle, thereby in the aforementioned Figure 10 In this implementation, the strain trend of the orthotic rod in the current cycle can be matched with the change trend template output by the correction model, thereby eliminating the need to pre-build the change trend template, further simplifying the process and improving detection efficiency.

[0164] As described above, the embodiments in this specification employ a multi-dimensional feature extraction model architecture that combines spatial and temporal features to accurately extract the spatial distribution characteristics and temporal evolution patterns of sensor data, avoiding the limitations of single-dimensional feature extraction and improving model accuracy. Furthermore, during model training, a multi-task joint training loss function is used, balancing the training accuracy of each task through weight allocation. This achieves simultaneous output of strain trend prediction, change time judgment, and correction effect evaluation, balancing detection efficiency and accuracy. Additionally, the trend predicted by the model for the next cycle can serve as a trend template, simplifying the template construction process, improving detection efficiency, reducing manual intervention costs, and enhancing the safety and controllability of spinal orthopedic treatment.

[0165] In some embodiments, this specification provides a control device, such as Figure 14 As shown, the device includes: The data acquisition module 1 is configured to acquire the sensor data array of the current period. The sensor data array includes sensor data for multiple consecutive unit durations, and each unit duration of sensor data includes the strain signal of each passive sensor. Result determination module 2 is configured to determine the orthopedic rod detection result for the current period based on the sensor data array.

[0166] In some implementations, the result determination module 2 is configured to: The strain rate of each passive sensor on the orthotic bar is determined based on the sensor data array. The strain trend of the orthotic bar in the current cycle is determined based on the strain rate. Based on the similarity between the strain trend of the orthotic rod in the current period and the strain trend of the orthotic rod in at least one historical period, the orthotic rod recovery result in the current period is determined, and the detection result includes the orthotic rod recovery result; or, based on the strain trend of the orthotic rod in the current period and a pre-constructed change trend template, the detection result is determined according to the matching result.

[0167] In some implementations, the result determination module 2 is configured to: For each passive sensor, based on the sensor data array, the change in the strain signal of the passive sensor within a preset time window is determined; Based on the comparison between the change and a preset threshold, the local strain parameters of the orthotic rod are determined, and the orthotic rod detection results include the local strain parameters.

[0168] In some implementations, the result determination module 2 is configured to: The sensor data array is input into a pre-trained correction model, which includes a spatial branch module and a temporal branch module. The correction model extracts the spatial features of the sensor data array through the spatial branch module and the temporal features of the sensor data array through the temporal branch module. Based on the spatial features and the temporal features, it outputs multi-task detection results, which include: the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and the correction effect parameters.

[0169] In some embodiments, this specification provides a storage medium storing computer instructions for implementing the methods of any of the foregoing embodiments.

[0170] In some embodiments, this specification provides a computer program product for implementing the methods of any of the foregoing embodiments.

Claims

1. An orthotic rod, characterized in that, The orthotic bar, used to provide guiding force to the spine through its own deformation during spinal correction, comprises: Rod; and A sensor array, comprising a plurality of passive sensors spaced apart along the length of the rod, wherein the passive sensors include magnetic elements for acquiring strain signals generated by the deformation of the rod; The rod body comprises multiple integrally formed segments along its length direction. The multiple segments include a apical region located in the middle of the length direction, edge regions located at both ends of the length direction, and a main bending region located between the apical region and the edge regions. Wherein, the peak temperature of austenite phase transformation in the apex region is lower than that in the main bend region, and the peak temperature of austenite phase transformation in the main bend region is lower than that in the edge region; The sensor density in the apex region is greater than that in the main curve region, and the sensor density in the main curve region is greater than that in the edge region.

2. The orthotic rod according to claim 1, characterized in that, The magnetic component includes magnetic blocks; a hollow mounting pipe is formed inside the rod along its length, and the magnetic blocks of the plurality of passive sensors are spaced apart in the mounting pipe; or, a plurality of assembly slots are formed on the surface of the rod along its length, and the magnetic blocks of the plurality of passive sensors are correspondingly arranged in the assembly slots.

3. The orthotic rod according to claim 1, characterized in that, The magnetic component includes a magnetostrictive sensing layer, the magnetic field strength of which varies with its own expansion and contraction, and the strain signal includes the magnetic field strength signal of the magnetostrictive sensing layer. The rod body has a hollow installation channel along its length, and the plurality of passive sensors are spaced apart in the installation channel. The magnetostrictive sensing layer of each passive sensor is attached to the inner wall of the installation channel. Alternatively, the surface of the rod body has a plurality of installation slots spaced apart along its length, and the magnetostrictive sensing layer of each passive sensor is attached to the installation slot. The outer surface of the magnetostrictive sensing layer is provided with a protective layer.

4. A correction system, characterized in that, include: The orthopedic rod according to any one of claims 1 to 3; A data acquisition terminal, comprising a magnetic induction device, wherein the magnetic induction device is used to acquire strain signals from a passive sensor on the orthopedic bar via electromagnetic induction; The data processing terminal includes a controller configured to receive the strain signal and generate a detection result based on the strain signal.

5. The correction system according to claim 4, characterized in that, The data acquisition terminal also includes a memory and a first communication module, wherein the memory is used to store strain signals acquired by the magnetic induction device; The data processing terminal further includes a second communication module, which is communicatively connected to the first communication module and is used to receive strain signals sent by the first communication module.

6. The correction system according to claim 4, wherein the controller is configured to: Acquire the sensor data array for the current period, the sensor data array including sensor data for multiple consecutive unit durations, each unit duration including the strain signal of each passive sensor; The orthopedic rod detection result for the current cycle is determined based on the sensor data array.

7. The correction system according to claim 6, characterized in that, The controller is configured to: The strain rate of each passive sensor on the orthotic bar is determined based on the sensor data array. The strain trend of the orthotic bar in the current cycle is determined based on the strain rate. Based on the similarity between the strain trend of the orthotic rod in the current period and the strain trend of the orthotic rod in at least one historical period, the orthotic rod recovery result in the current period is determined, and the detection result includes the orthotic rod recovery result; Alternatively, the detection result can be determined by matching the strain trend of the orthotic rod in the current cycle with a pre-constructed trend template.

8. The correction system according to claim 6, characterized in that, The controller is configured to: For each passive sensor, based on the sensor data array, the change in the strain signal of the passive sensor within a preset time window is determined; Based on the comparison between the change and a preset threshold, the local strain parameters of the orthotic rod are determined, and the orthotic rod detection results include the local strain parameters.

9. The correction system according to claim 6, characterized in that... The controller is configured to: The sensor data array is input into a pre-trained correction model, which includes a spatial branch module and a temporal branch module. The correction model extracts the spatial features of the sensor data array through the spatial branch module and the temporal features of the sensor data array through the temporal branch module. Based on the spatial features and the temporal features, it outputs multi-task detection results, which include: the strain trend of the orthotic rod in the next cycle, the orthotic rod replacement time, and the correction effect parameters.

Images