Self-adaptive cervical vertebra radian traction system based on AI

Abstract

The invention provides a self-adaptive cervical vertebra radian traction system based on AI, and belongs to the technical field of medical rehabilitation. The system comprises a neck support module, a forehead holding module, a cervical vertebra radian traction module, a double-lifting module, a three-point mechanical structure and an AI control module. The neck support module comprises a neck support belt, and a flexible pressure sensor group is arranged in the neck support belt; the forehead embracing module comprises a manipulator forehead embracing driving mechanism and a forehead band, and a manipulator is provided with a second force sensor to collect forehead embracing force; the double-lifting module comprises a first lifting assembly and a second lifting assembly and can drive the neck support belt to move; the AI control module is used for adjusting the angle of the forehead band and the height of the neck collar band and correcting the position of the neck collar band. According to the system, AI driving is combined with a three-point mechanical structure, precise and safe traction dynamically matched with the cervical vertebra of an individual is achieved, and the risks of deviation and curvature misalignment of traditional traction are thoroughly eliminated.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation technology, and more specifically to an AI-based adaptive cervical curvature traction system. Background Technology

[0002] With the rapid penetration of artificial intelligence into the field of medical image analysis, deep learning-based automatic segmentation of cervical spine images, quantification of physiological curvature, and prognostic assessment of treatment efficacy have entered the clinical validation stage. However, these studies remain in an "offline-static" scenario: after the algorithm completes image analysis in seconds to minutes, it outputs static parameters in the form of a report, and then the traction prescription is manually set, failing to achieve closed-loop control of "millisecond-level perception-decision-execution". The resulting "measurement-calculation-control" time lag directly disrupts the key technology chain of "precise morphological measurement-dynamic adjustment of traction force-instantaneous force system balance", becoming the core bottleneck limiting the efficacy of personalized cervical curvature traction.

[0003] In the existing technology, although CN105662684B proposes a traction strategy that couples three parameters: body position angle, neck force point, and forehead force point, its control logic is still open-loop: the traction force is executed according to a preset arc, which has no dynamic correlation with the real-time cervical curvature and soft tissue stress feedback. This results in the morphological data output by the AI ​​algorithm being unable to drive the mechanical actuator to make adaptive adjustments, forming a structural disconnect between "intelligent assessment and mechanical execution." In addition, in the existing technology, the traction force is executed according to a preset arc, which has no dynamic correlation with the real-time cervical curvature and soft tissue stress feedback. This results in the morphological data output by the AI ​​algorithm being unable to drive the mechanical actuator to make adaptive adjustments, forming a structural disconnect between "intelligent assessment and mechanical execution."

[0004] Therefore, there is an urgent need in this field to construct a closed-loop three-point mechanical architecture that coordinates "neck support (support + bidirectional angle adjustment), robotic arm for forehead support (extension + angle adjustment) / back pressure point (anchor point = reverse fixation)". The dual-lift neck support for the neck serves as the core support, while the robotic arm for forehead support and the back support form independently adjustable force application points. Through precise positioning, coordinated control, and dynamic balance mechanisms, it adapts to the characteristics of the cervical chord and real-time curvature changes to achieve adaptive adjustment. At the same time, it combines AI technology to achieve real-time optimization of traction parameters, thereby improving the targeting and effectiveness of cervical curvature traction. Summary of the Invention

[0005] This invention provides an AI-based adaptive cervical curvature traction system to solve the problem in existing technologies that cannot adaptively adjust based on the individual spine.

[0006] To achieve the above objectives, this invention provides an AI-based adaptive cervical curvature traction system. The system includes: a cervical support module comprising a standing neck-hugging drive mechanism, a cervical support strap, and a flexible pressure sensor group, the latter used to collect head weight data; a forehead-hugging module comprising a robotic arm, a robotic arm forehead-hugging drive mechanism, and a forehead strap, the robotic arm having a second force sensor at its bottom joint for collecting the forehead-hugging force of the forehead strap; a cervical curvature traction module comprising a chassis frame, a traction frame, and a traction component for driving the traction frame, the traction component having a first force sensor for collecting the force of supine cervical curvature traction; and a dual-lifting module comprising a first lifting group within the support frames on both sides of the cervical support strap. The system includes a first lifting component and a second lifting component that drives the neck brace to move vertically; a three-point mechanical structure for coordinated adjustment of the forward flexion angle and the backward flexion angle to target and restore the physiological curvature of the cervical spine; an AI control module for calculating the ideal forehead-hugging angle based on head and neck position data and ambient temperature data using a preset mapping model and driving the robotic arm to adjust the forehead brace; it is also used to calculate the traction force based on head weight data and cervical chord value using a preset model and drive the dual lifting module to adjust the lifting force of the neck brace; it is also used to calculate the midline offset based on head and neck position data, calculate the target cervical spine position using a preset positioning model, drive the first lifting component to adjust the neck brace, and perform multi-dimensional traction based on the three-point mechanical structure.

[0007] Optionally, the adaptive cervical curvature traction system further includes a data acquisition module, which includes: a neck shape recognition unit for acquiring the neck contour and calculating the neck length and cervical curvature chord value; a ranging unit, including an infrared sensor, for acquiring head and neck position data; and a temperature acquisition unit for acquiring ambient temperature data. Optionally, the adjustment of the forehead band includes: performing temperature compensation on head and neck position data based on ambient temperature data; calculating an ideal forehead-holding angle based on the compensated head and neck position data through a preset mapping model, and comparing it with the current forehead-holding angle to obtain an angle difference; driving the robotic arm to adjust the angle of the forehead band according to the angle difference, and adjusting the forehead-holding force of the forehead band in conjunction with the forehead-holding force collected by the second force sensor.

[0008] Optionally, the adjustment of the neck support force includes: acquiring head weight data collected by the pressure flexible sensor group and the neck curvature chord value calculated by the neck shape recognition unit, and performing filtering and steady-state detection on the head weight data; obtaining the curvature compensation amount according to the neck curvature chord value through a preset calculation rule; obtaining the traction force based on the processed head weight data and the calculated curvature compensation amount through a preset traction force calculation model; determining the adjustment parameters of the dual lifting module according to the obtained traction force; and driving the first lifting component and the second lifting component according to the adjustment parameters to adjust the neck support force of the neck brace. Optionally, the adjustment of the neck brace includes: acquiring head and neck position data collected by the ranging unit; calculating the midline offset and shoulder distance based on the head and neck position data; obtaining the target cervical spine position based on the calculated midline offset and shoulder distance using a preset positioning model; comparing the target cervical spine position with the current position of the neck brace and calculating the offset; driving the first lifting assembly to adjust the angle of the neck brace according to the calculated offset, and verifying the adjustment effect based on the data collected by the flexible pressure sensor group.

[0009] Optionally, the adaptive cervical curvature traction system further includes an early warning module, which is used to calculate the displacement between the cervical brace and the forehead band using the collected head and neck position data, and execute the corresponding response strategy according to the preset three-level early warning mechanism.

[0010] Optionally, the three-level early warning mechanism includes: when the displacement reaches a first preset value, automatically correcting the position of the neck brace and the forehead strap; when the displacement reaches a second preset value, activating the second lifting component to adjust the height of the neck brace to enhance the fixation effect; and when the displacement reaches a third preset value, determining it as a high-risk slippage, automatically restoring the system's initial parameters and stopping traction.

[0011] Optionally, the three-point mechanical structure consists of a robotic arm holding the forehead, a neck brace holding the mid-neck, and a back support point. The robotic arm holding the forehead applies force at the eyebrows and provides downward pull based on the cervical vertebral chord value, simultaneously adjusting the extension angle. The cervical vertebral chord value is the vertical distance between the arc height and chord length of the target cervical vertebral physiological curvature, used for extension angle control. The neck brace holding the mid-neck applies force to the mid-cervical spine, using infrared measurement of the shoulder distance to locate the cervical mid-vertebral axis and lock the mid-cervical segment. It also adjusts the support height and angle based on the cervical chord value, used for bidirectional control of extension and flexion angles. The back support applies force to the back, using back friction as a reverse support anchor point, used for flexion angle control.

[0012] Optionally, the forward flexion adjustment method is to adjust the height of the neck brace by adjusting the dual lifting module, thereby changing the height difference between the neck support point and the back support point to adjust the forward flexion angle.

[0013] Optionally, the extension adjustment method is as follows: calculate the deviation value based on the current cervical chord value and the reference cervical chord value; when the deviation value reaches the preset difference, increase the extension angle; when the deviation value is less than the preset difference, adjust the extension angle to the preset value.

[0014] This invention provides an AI-based adaptive cervical curvature traction system that achieves multi-dimensional, closed-loop, and physiologically adapted dynamic cervical traction therapy through AI control. The system utilizes a flexible pressure sensor array to collect head weight data in real time, a second force sensor to monitor the force applied to the forehead, and combines this with head and neck position and ambient temperature data collected by other modules. Through the AI ​​control module and its preset mapping model, calculation rules, and positioning model, the system intelligently and dynamically coordinates the three-point biomechanical structure dimensions. The AI ​​control module calculates and drives the robotic arm forehead-holding mechanism in real time based on temperature-compensated head and neck data. The device precisely controls the angle at which the neck brace embraces the forehead, calculates personalized traction force based on filtered head weight data and the cervical chord value representing the physiological curvature of the cervical spine, and instructs the dual lifting modules to precisely adjust the flexion and extension angles of the neck brace. Simultaneously, the first lifting component dynamically adjusts the horizontal position of the neck brace based on the offset calculated from the head and neck position. This effectively overcomes the problems of traction direction deviation, insufficient physiological curvature adaptation, over-traction or under-traction, and high risk of displacement and slippage caused by fixed parameters and static adjustments in traditional traction devices, significantly improving the accuracy, individual adaptability, comfort, and safety of treatment. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a cross-sectional view of the cervical curvature traction module provided in an embodiment of the present invention; Figure 2 This is a cross-sectional view of the dual lifting module provided in an embodiment of the present invention; Figure 3 This is a cross-sectional view of the chassis frame provided in an embodiment of the present invention; Figure 4 This is a top view of the chassis frame provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the second lifting component structure provided in an embodiment of the present invention; Figure 6 This is a side view of the cervical curvature traction module provided in an embodiment of the present invention.

[0016] Reference numerals: 1. Neck brace module; 11. Neck brace strap; 12. Flexible pressure sensor group; 2. Forehead support module; 21. Robotic arm forehead support drive mechanism; 22. Forehead strap; 23. Second force sensor; 31. Chassis; 32. Traction frame; 33. Traction assembly; 4. Dual lifting module; 41. First lifting assembly; 42. Second lifting assembly. Detailed Implementation

[0017] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0018] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0019] As mentioned above, with the increasing demand for precision in the diagnosis and treatment of cervical spine diseases, traditional cervical traction devices face limitations such as fixed traction angles, rough force matching, and lack of physiological curvature compensation. These limitations make it difficult to dynamically respond to subtle head and neck movements and individual physiological differences, easily leading to problems such as traction position deviation, excessive nerve traction, or insufficient curvature correction, severely restricting treatment effectiveness and safety. Therefore, developing an intelligent, adaptive cervical curvature traction system is crucial.

[0020] To address this issue, this invention provides an AI-based adaptive cervical curvature traction system. Utilizing an AI control module, it dynamically constructs a multi-dimensional physiological traction model, accurately calculates the ideal forehead-holding angle and traction force, and synchronously drives the robotic arm forehead-holding drive mechanism, dual lifting modules, and electric push components to adjust the forehead band angle, cervical support height, and horizontal position in a closed-loop manner. This forms a real-time adaptive mechanism of "perception-decision-execution-verification," effectively achieving precise correction and dynamic stable traction of individualized cervical physiological curvature, significantly improving treatment efficacy and safety.

[0021] The following is combined with Figures 1-6 This invention is described in detail.

[0022] like Figures 1-6 As shown, this embodiment of the invention provides an AI-based adaptive cervical curvature traction system, the adaptive cervical curvature traction system comprising: The neck support module includes a vertical neck-hugging drive mechanism and a neck support strap. The neck support strap has a three-layer structure, with a plastic elastomer placed in the middle layer and a flexible pressure sensor group attached to the plastic elastomer for collecting head weight data. The forehead-holding module includes a robotic arm, a robotic arm forehead-holding drive mechanism, and a forehead band. A second force sensor is provided at the bottom joint of the robotic arm to collect the forehead-holding force of the forehead band. The cervical curvature traction module includes a chassis frame, a traction frame supporting the cervical support module and the forehead support module, and a traction component for driving the traction frame. The traction component is equipped with a first force sensor for collecting the force of cervical curvature traction in a supine position. A dual lifting module includes a first lifting component within the support frames on both sides of the cervical support belt, and a second lifting component for driving the cervical support belt to move vertically. A three-point mechanical structure comprises a force point for the robotic arm supporting the forehead, a force point for the cervical support belt supporting the middle section of the neck, and a force point for the back. This three-point mechanical structure is used for coordinated adjustment of the forward flexion angle and the backward flexion angle to target and restore the physiological curvature of the cervical spine. An AI control module is used to... The system uses head and neck position data and ambient temperature data to calculate the ideal forehead-holding angle using a preset mapping model and drives the robotic arm to adjust the forehead band. It also uses head weight data and cervical curvature chord value to calculate traction force using a preset model and drives the dual lifting module to adjust the lifting force of the neck support strap. Furthermore, it uses head and neck position data to calculate midline offset, calculate the target cervical spine position using a preset positioning model, and drives the first lifting component to adjust the neck support strap. Additionally, based on the three-point mechanical structure, it links the chassis frame and traction frame in the cervical curvature traction module, the robotic arm forehead-holding drive mechanism in the forehead-holding module, and the upright neck-holding drive mechanism in the neck support module to perform multi-dimensional traction.

[0023] Specifically, the neck brace is the component that directly contacts and supports the patient's neck. It features a three-layer composite structure: the upper layer combines sponge and carbon fiber heating cloth for heat therapy; the middle layer is a PTU elastomer with embedded flexible pressure sensor arrays; and the lower layer uses sponge webbing to connect to the support frame, forming a stable support structure. The flexible pressure sensor array consists of multiple highly sensitive flexible pressure sensors. These sensors utilize the principles of piezoresistive or capacitive effects. When the weight of the patient's head is applied to the neck brace, the sensors convert pressure changes into electrical signals, thus accurately collecting head weight data. These flexible pressure sensors have excellent flexibility, conforming to the curve of the neck to ensure accurate data collection and evenly distribute neck pressure, avoiding excessive local pressure that could cause discomfort. The neck brace forms a U-shaped structure through the PTU elastomer, with both ends fixed to the support frame. Using the back of the neck as the twist point, the cushioning properties of the PTU elastomer effectively attenuate the traction impact, solving the problems of force loss and swaying associated with traditional webbing traction. Meanwhile, the neck brace can rely on a flexible pressure sensor array to monitor the weight of the patient's head in real time, providing data support for the AI ​​control module to adjust the height of the neck brace.

[0024] Specifically, the robotic arm forehead-holding drive mechanism is an automated operating device that integrates a standing joint and a telescopic transmission mechanism. It can precisely adjust its height and rotation according to instructions from the AI ​​control module, thereby adjusting the force and angle of the forehead-holding motion. The traction component can drive the entire traction frame to move horizontally. Because the patient's head is fixed by the neck brace and forehead strap, traction can be applied to the patient's cervical spine as the traction frame moves. The first lifting components within the support frames on both sides of the neck brace can be electrically operated push rods. The two first lifting components can operate independently or simultaneously, thereby adjusting the angle or height of the neck brace.

[0025] The adaptive cervical curvature traction system provided in this invention comprises five main components: a neck brace module, a forehead support module, a cervical curvature traction module, a dual-lifting module, and an AI control module. Each module achieves precise traction through the collaboration of hardware sensing and intelligent algorithms. The forehead support module's robotic arm forehead support drive mechanism works in conjunction with the forehead strap, using a second force sensor to acquire the forehead support force, providing a basis for the forehead strap height. The cervical curvature traction module, based on a chassis, drives the traction frame supporting the neck brace and forehead support module to complete the traction action. The first and second lifting components of the dual-lifting module respectively achieve fine-tuning of the neck brace angle and coarse-tuning of the height, forming a multi-dimensional adjustment capability. The AI ​​control module, as the core, calculates the ideal angle based on head and neck position data and ambient temperature data through a mapping model and drives the robotic arm forehead support drive mechanism to adjust it. Combining head weight data and cervical curvature chord value, it derives the traction force through a preset model to adjust the neck brace height. Based on head and neck position data, it calculates the target cervical spine position and drives the electric push component to correct the neck brace position. The system boasts significant technical advantages: First, by integrating hardware such as flexible pressure sensor arrays and secondary force sensors with AI control module algorithms, it achieves closed-loop control from data acquisition to parameter calculation and execution adjustment, solving the problem of poor adaptability caused by traditional equipment relying on manually preset parameters. Second, the multi-module collaborative adjustment mechanism greatly improves the accuracy of traction, ensuring that the traction force matches individual characteristics such as cervical curvature and head weight. Third, by combining real-time data sensed by sensors with preset models, the traction process becomes dynamically adaptable, enabling timely adjustments in response to changes in body position and ensuring patient comfort through pressure feedback, providing reliable equipment support for personalized cervical traction treatment.

[0026] Preferably, the adaptive cervical curvature traction system further includes a data acquisition module, which includes: a neck shape recognition unit for acquiring the neck contour and calculating the neck length and cervical curvature chord value; a ranging unit, including an infrared sensor, for acquiring head and neck position data; and a temperature acquisition unit for acquiring ambient temperature data.

[0027] Specifically, the neck shape recognition unit can accurately calculate the neck length and cervical chord value by acquiring the neck contour. These parameters directly reflect the physiological morphological characteristics of the cervical spine and are important bases for the realization of functions such as traction height adjustment. The ranging unit is equipped with an infrared sensor, which can accurately collect head and neck position data, including information such as the distance from the left and right shoulders to the midline, providing raw data for calculating key parameters such as the distance between the shoulders and the midline offset. The temperature acquisition unit focuses on collecting ambient temperature data to perform temperature compensation on the head and neck position data, eliminating the influence of ambient temperature changes on the sensor measurement accuracy.

[0028] The preferred embodiment of this invention provides a data acquisition module that, through the collaborative acquisition of multiple types of units, achieves comprehensive acquisition of cervical spine-related physiological parameters, positional parameters, and environmental parameters. This provides a complete data foundation for the precise calculation of the AI ​​control module, avoiding adjustment deviations caused by missing or isolated data. The clear division of labor among the units and the interconnection of data enable the system to more comprehensively grasp the individual characteristics of patients and environmental influences, improving the reliability and accuracy of the data. This lays a solid data foundation for subsequent personalized traction adjustment and further enhances the system's adaptive capabilities.

[0029] Preferably, adjusting the forehead band includes: performing temperature compensation on head and neck position data based on ambient temperature data; calculating the ideal forehead-holding angle based on the compensated head and neck position data through a preset mapping model, and comparing it with the current forehead-holding angle to obtain the angle difference; driving the robotic arm forehead-holding drive mechanism to adjust the angle of the forehead band according to the angle difference, and adjusting the height of the forehead band in conjunction with the forehead-holding force collected by the second force sensor.

[0030] Specifically, the mapping model is represented as: (1) in, For the ideal forehead angle; These are the distances from the left and right shoulders to the midline, respectively. This is the shoulder spacing after temperature compensation. Based on shoulder width, The shoulder width influence coefficient, with a clinically calibrated value of 8.2; The correction factor for shoulder asymmetry is set to 15.0; The base angle is 5°. The temperature compensation amount is calculated from the ambient temperature data: =0.02×(T-25), where T is the ambient temperature.

[0031] Specifically, when performing temperature compensation, the AI ​​control module first obtains the real-time ambient temperature from the temperature acquisition unit, compares it with the reference temperature of 25°C to obtain the temperature difference, and then calculates the temperature compensation amount based on the angle compensation standard of 0.02 degrees per degree Celsius. This is used to correct the head and neck position data and avoid the distance measurement deviation caused by temperature changes from affecting subsequent calculations. Then, the ideal angle for the patient's body shape is calculated using formula (1). After comparing the current forehead-holding angle to obtain the angle difference, the robotic arm forehead-holding drive mechanism will act according to the difference and flexibly adjust the forehead band angle. The second force sensor will collect the forehead-holding force in real time. When the force is not within the preset comfort range (such as 1.2-2.5 kPa), the system will make a fine adjustment to the forehead band height to ensure that the angle is accurate while the force on the patient's forehead is appropriate, further improving the accuracy and comfort of the adjustment.

[0032] The forehead band adjustment scheme provided in the preferred embodiment of the present invention uses a temperature compensation mechanism to calculate the compensation amount based on the temperature difference between the real-time ambient temperature and the reference temperature, correcting the head and neck position data. This effectively eliminates the interference of temperature changes on the ranging accuracy, providing accurate basic data for subsequent angle calculation and avoiding angle adjustment errors caused by data deviation. The coefficients and basic angles calibrated by a large amount of clinical data in the preset mapping model, combined with parameters such as shoulder width and shoulder asymmetry after temperature compensation, can accurately calculate the ideal forehead-holding angle suitable for the patient's body shape, realizing the personalization and scientific nature of angle adjustment.

[0033] Preferably, the adjustment of the forehead support force includes: acquiring head weight data collected by the pressure flexible sensor group and the neck curvature chord value calculated by the neck shape recognition unit, and performing filtering and steady-state detection on the head weight data; obtaining the curvature compensation amount according to the neck curvature chord value through a preset calculation rule; obtaining the traction force based on the processed head weight data and the calculated curvature compensation amount through a preset traction force calculation model; determining the adjustment parameters of the dual lifting module according to the obtained traction force; and driving the first lifting component and the second lifting component according to the adjustment parameters to adjust the height of the neck support strap.

[0034] Specifically, the traction force calculation model is expressed as follows: (2) in, Traction force (unit: kg); Head weight data (unit: kg) after filtering and steady-state detection; ΔC is the weight coefficient (clinically recommended range 0.08~0.12, dynamically selected based on the patient's cervical spine weight-bearing capacity); b is the basic traction force (normal range 2.0~3.5kg, ensuring minimum effective traction force); ΔC is the curvature compensation (unit: kg), calculated from the cervical curvature chord value: if the cervical curvature chord value < 12mm, ΔC = 0.15 × (12 - cervical curvature chord value); if the cervical curvature chord value > 17mm, ΔC = -0.10 × (cervical curvature chord value - 17); if 12mm ≤ cervical curvature chord value ≤ 17mm, ΔC = 0. This formula achieves dynamic adaptation of traction force by integrating head weight data and cervical curvature characteristics—ensuring a basic traction force matching the head weight while making targeted adjustments through curvature compensation to make the traction force more consistent with the physiological state of the cervical spine, avoiding excessive or insufficient traction due to abnormal curvature.

[0035] Specifically, when adjusting the pressure on the forehead, the AI ​​control module receives head weight data collected by the flexible pressure sensor group and neck morphology recognition unit calculated by the neck chord value during the data acquisition phase. For head weight data, filtering is performed first. A moving average filtering method can be used, selecting the average of 10 sets of data within the last 5 seconds to eliminate noise interference such as momentary jitter. When calculating the curvature compensation, the preset calculation rules are based on the normal curvature range of the cervical spine (cervical chord value 12mm~17mm): when the cervical chord value is less than 12mm, it indicates that the cervical curvature is straightened, and ΔC needs to be increased to enhance traction and promote curvature recovery; when it is greater than 17mm, it indicates that the curvature is too large, and ΔC needs to be reduced to avoid excessive traction; within the normal range, no compensation is required. When applying the traction force calculation model, the weight coefficient k is dynamically adjusted according to the patient's cervical spine load-bearing capacity. For example, 0.08 is used for elderly patients or those with severe cervical spine injuries, and 0.12 is used for young patients with good cervical spine conditions. The base traction force b is combined with the patient's body type, with 2.0~2.5kg for those who are thin and 3.0~3.5kg for those who are overweight. When determining the adjustment parameters of the dual lifting module, the calculated traction force F is converted into the extension and retraction of the first lifting component and the lifting height of the second lifting component. Finally, based on these adjustment parameters, the two components are driven to work together to precisely adjust the height of the neck brace, so that the traction force is accurately applied to the cervical spine.

[0036] The preferred embodiment of this invention provides a forehead-holding force adjustment scheme that ensures the accuracy of traction force calculation by processing head weight data through sliding average filtering and steady-state detection; the curvature compensation rule based on the cervical chord value makes the traction force adapt to the actual curvature of the cervical spine, ensuring safe and effective traction; the dynamic adjustment of the weight coefficient and the basic traction force realizes personalized customization of the traction force; and the high-precision coordinated adjustment of the dual lifting modules ensures that the traction force is accurately applied to the cervical spine, improving the adaptability and accuracy of the treatment.

[0037] Preferably, the adjustment of the neck brace includes: acquiring head and neck position data collected by the ranging unit; calculating the midline offset and the distance between the shoulders based on the head and neck position data; obtaining the target cervical spine position based on the calculated midline offset and the distance between the shoulders using a preset positioning model; comparing the target cervical spine position with the current position of the neck brace and calculating the offset; driving the first lifting assembly to adjust the angle of the neck brace according to the calculated offset, and verifying the adjustment effect based on the data collected by the flexible pressure sensor group.

[0038] Specifically, the positioning model is represented as follows: (3) in, The horizontal offset of the target cervical spine position relative to the midline of the cervical collar; These are the distances from the left and right shoulders to the midline after temperature compensation; This is the processed headweight data; Neck length; This is the shoulder deviation correction factor (calibrated value 0.35), used to adjust the offset amount according to shoulder asymmetry; The head weight distribution coefficient (calibrated value 1.2) reflects the influence of head weight on cervical spine position. This is the base offset (default 0mm, which can be finely adjusted according to the average physiological position of the cervical spine in a population). This formula achieves precise calculation of the target cervical spine position by integrating parameters such as shoulder position difference and the ratio of head weight to neck length. This provides a direct basis for adjusting the angle of the neck brace in the first lifting component, ensuring the accuracy of the correspondence between the neck brace and the cervical spine.

[0039] Specifically, the AI ​​control module collects the distance from the left and right shoulders to the midline using the infrared sensor of the ranging unit, and calculates the midline offset and the distance between the two shoulders. Combining these data, the horizontal offset of the target cervical spine position is calculated using a preset positioning model (Equation (3)) to obtain the position of the target cervical spine. Then, the offset of the cervical spine is obtained by comparing the position of the target cervical spine with the current position of the cervical support (monitored by the infrared sensor), and the first lifting component is driven to adjust the angle of the cervical support by an amplitude proportional to the offset. After adjustment, the pressure at each point is checked by the flexible pressure sensor group to see if it is uniform (the pressure difference is controlled within 0.5 kPa). If it does not meet the standard, the adjustment is repeated to ensure that the cervical support is accurately aligned with the target cervical spine position.

[0040] The neck brace adjustment scheme provided in the preferred embodiment of the present invention uses an infrared sensor to collect head and neck position data at high frequency, which can capture changes in shoulder position in a timely manner and provide real-time basis for subsequent calculations. Combining the midline offset and the distance between the two shoulders, the target cervical vertebra position is calculated using a clinically validated positioning model, which can accurately lock the correct position of the cervical vertebra and reduce positioning deviations caused by individual differences. The first lifting component is driven to adjust the angle of the neck brace with high precision based on the offset, and the adjustment effect is verified by a flexible pressure sensor group to ensure that the neck brace corresponds precisely to the target cervical vertebra position, avoiding the impact of position offset on the traction effect, and significantly improving the system's dynamic correction capability for cervical vertebra position and the accuracy of traction.

[0041] Preferably, the adaptive cervical curvature traction system further includes an early warning module, which is used to calculate the displacement between the cervical brace and the forehead band using the collected head and neck position data, and execute the corresponding response strategy according to the preset three-level early warning mechanism.

[0042] More preferably, the three-level early warning mechanism includes: when the displacement reaches a first preset value, automatically correcting the position of the neck brace and the forehead strap; when the displacement reaches a second preset value, activating the second lifting component to adjust the height of the neck brace to enhance the fixation effect; when the displacement reaches a third preset value, determining it as a high-risk slippage, automatically restoring the system's initial parameters and stopping traction.

[0043] Specifically, the three-level early warning mechanism executes corresponding strategies based on different preset values ​​of displacement: When the displacement reaches the first preset value (e.g., 4 mm), the system automatically starts a fine-tuning program, using the first lifting component of the dual lifting module and the robotic arm forehead-holding drive mechanism to make small corrections to the position of the neck brace and the angle of the forehead band, quickly eliminating minor deviations and ensuring stable traction position; When the displacement reaches the second preset value (e.g., 8 mm), in addition to position fine-tuning, the second lifting component is activated to appropriately raise the height of the neck brace, increasing the tightness of the fit between the neck brace and the neck, enhancing the overall fixation effect, and preventing further displacement; When the displacement reaches the third preset value (e.g., 12 mm), the system determines it to be a high-risk slippage state, immediately triggering a safety protection program, automatically restoring the initial parameters of all components such as the neck brace and forehead band, and stopping the traction action to maximize patient safety.

[0044] In a preferred embodiment of the invention, the early warning module collects data in real time through the ranging unit and calculates the relative displacement between the cervical brace and the forehead band. This allows for timely detection of subtle changes in their positions, providing a precise basis for the early warning mechanism. Real-time monitoring of positional deviations during traction is achieved, preventing poor traction effects or safety risks due to undetected displacement. The three-level early warning mechanism employs corresponding strategies based on different preset displacement values, forming a multi-layered safety assurance system. The combination of these two mechanisms enables dynamic monitoring and precise adjustment of the traction process while maintaining the continuity and safety of treatment through a graded response mechanism. This significantly improves the system's intelligence and reliability, allowing cervical traction treatment to be performed efficiently and safely.

[0045] Preferably, the robotic arm's forehead-holding force point acts on the eyebrows. This force point provides downward pulling force based on the cervical spine's arc chord value and simultaneously adjusts the extension angle. The cervical spine's arc chord value is the vertical distance between the arc height and chord length of the target cervical spine's physiological curvature, used for the extension angle control correlation. The neck support's neck-holding force point acts on the mid-cervical spine, using infrared measurement of the shoulder distance to locate the cervical spine's mid-axis to lock the mid-cervical spine segment, and adjusting the support height and angle in conjunction with the cervical spine's arc chord value, used for the bidirectional control correlation of extension and flexion angles. The back force point acts on the back, using back friction as a reverse support force anchor point, used for correlation with flexion angle control.

[0046] In a preferred embodiment of the invention, the precision and adaptability of cervical curvature traction are significantly improved through precise positioning and coordinated control of each force point in the three-point mechanical structure. Specifically, the robotic arm, which holds the forehead, applies force at the eyebrows, providing downward pull based on the cervical chord curvature value and simultaneously adjusting the extension angle. This design not only achieves precise control of the extension angle but also allows for targeted adjustments based on individual patient differences in cervical physiological curvature, enhancing the personalization of treatment. The cervical collar, which holds the neck, applies force to the mid-cervical spine. Infrared measurement of the shoulder distance locates the cervical midline, and the support height and angle are adjusted in conjunction with the cervical chord curvature value, enabling bidirectional control of extension and flexion angles. This precisely targets the mid-cervical segment, effectively enhancing the targeted treatment effect on cervical lesions. The back force point applies force to the back, using friction as a reverse support anchor point, and is linked to the flexion angle adjustment, providing a stable and reliable mechanical basis for the entire traction process and ensuring the effective transmission and application of traction force. The linkage adjustment mechanism of the three-point biomechanical structure can more accurately restore the physiological curvature of the cervical spine through multi-dimensional and multi-angle synergistic effects, while taking into account the individual differences of different patients, thus greatly improving the safety, effectiveness and comfort of traction therapy.

[0047] Preferably, the forward flexion adjustment method is as follows: the height of the neck brace is adjusted by using a dual lifting module to change the height difference between the point of force on the neck and the point of force on the back, thereby adjusting the forward flexion angle.

[0048] A further preferred method for adjusting the extension position is as follows: calculate the deviation value based on the current cervical chord curve value and the reference cervical chord curve value; when the deviation value reaches a preset difference, increase the extension angle; when the deviation value is less than the preset difference, adjust the extension angle to the preset value. Preferred embodiments of this invention illustrate the adjustment methods for flexion and extension positions. Through a dual-lifting module and dynamic feedback of the cervical chord curve value, precise control of the cervical traction angle is achieved. During flexion adjustment, the system changes the height difference between the point of contact with the neck and the point of contact with the back by adjusting the neck brace height. For example, when the difference decreases from the initial 8cm to 5cm, the cervical flexion angle increases from 15° to 25°, forming a "forward tilt" traction posture, effectively improving the problem of insufficient cervical lordosis. Extension adjustment dynamically adjusts the angle based on the deviation of the cervical chord curve value. If the patient's current cervical chord curve value is 12mm (the baseline value is 15mm), and the system calculates a deviation of 3mm (reaching the preset difference of 2mm), then the extension angle is automatically increased from 20° to 25°, promoting the recovery of physiological curvature by increasing the degree of cervical extension. When the chord curve value recovers to 14mm after treatment, the deviation is less than the preset difference, and the system adjusts the extension angle back to the preset 22° to maintain the therapeutic effect. This dynamic adjustment mechanism is like a "smart traction plan" customized for each patient. Through real-time monitoring and feedback control, it significantly improves the accuracy of cervical curvature correction while ensuring treatment safety. It is especially suitable for personalized treatment of complex conditions such as straightening or reverse curvature of the cervical spine.

[0049] For example, consider a middle-aged patient with straightened cervical curvature undergoing adaptive cervical curvature traction therapy: After treatment begins, the ranging unit first collects head and neck position data such as shoulder width and distance from the left and right shoulders to the midline. The temperature compensation mechanism calculates an angle compensation of 0.06 degrees based on the temperature difference between the real-time ambient temperature of 28℃ and the reference temperature of 25℃. The corrected data is input into the AI ​​control module. The robotic arm forehead-hugging drive mechanism calculates the ideal forehead-hugging angle based on the mapping model, and, in conjunction with the second force sensor, adjusts the forehead strap force to 2.0 kPa to precisely fit the patient's forehead. The flexible pressure sensor group acquires the patient's head weight data (5.5 kg), which, after filtering and steady-state detection, is combined with the cervical curvature chord value (10 mm) to calculate the target traction force (5.5 × 0.1 + 2.5 + 0.15 × (12 - 10) = 3.3 kg) through the traction force calculation model. The dual lifting module adjusts the height of the neck brace according to this parameter, and the positioning model aligns the neck brace with the target cervical spine position based on the shoulder data. During treatment, the patient slightly turned over, causing a relative displacement of 5mm between the neck brace and the forehead strap. The warning module triggered a level one warning, and the system automatically fine-tuned the position of the two. Subsequently, the patient twisted his head due to discomfort, and the displacement increased to 9mm, triggering a level two warning. The neck brace was raised by 0.5mm to improve the fit.

[0050] In summary, the AI-based adaptive cervical curvature traction system provided by this invention offers significant advantages. Through the coordinated operation of multiple modules, including a cervical support module and a forehead support module, combined with precise adjustment by an AI control module, it achieves personalized treatment: considering the influence of ambient temperature, it calculates the traction force based on head weight and the cervical curvature chord value, dynamically correcting the cervical support position to ensure precise traction; the three-point biomechanical structure works in tandem to regulate the flexion and extension angles, targeting and restoring the physiological curvature of the cervical spine, thus enhancing treatment specificity; by changing the drop or adjusting the angle based on the chord value deviation, it adapts to different treatment needs, enhancing the system's adaptability and effectiveness; simultaneously, real-time monitoring and feedback from various sensors ensure treatment safety, significantly improving overall treatment efficacy and patient comfort.

[0051] The above description is merely a preferred embodiment of the technical solution of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An AI-based adaptive cervical curvature traction system, characterized in that, The adaptive cervical curvature traction system includes: The neck support module includes a vertical neck support drive mechanism, a neck support strap, and a flexible pressure sensor group, which is used to collect head weight data. The forehead-holding module includes a robotic arm, a robotic arm forehead-holding drive mechanism, and a forehead strap. A second force sensor is provided at the bottom joint of the robotic arm to collect the forehead-holding force of the forehead strap. A cervical curvature traction module includes a chassis frame, a traction frame, and a traction component for driving the traction frame. The traction component is equipped with a first force sensor for collecting the force of cervical curvature traction in a supine position. The dual lifting module includes a first lifting component inside the support frames on both sides of the neck brace, and a second lifting component that drives the neck brace to move in the vertical direction; The three-point biomechanical structure is used to coordinate the adjustment of the forward flexion angle and the backward flexion angle, and to target the restoration of the physiological curvature of the cervical spine. The AI ​​control module is used to calculate the ideal forehead-hugging angle based on head and neck position data and ambient temperature data using a preset mapping model, and drive the robotic arm to adjust the forehead strap; it is also used to calculate the traction force based on head weight data and cervical chord value using a preset model, and drive the dual lifting module to adjust the lifting force of the neck brace; it is also used to calculate the midline offset based on head and neck position data, calculate the target cervical spine position using a preset positioning model, and drive the first lifting component to adjust the neck brace; and it performs multi-dimensional traction based on the three-point mechanical structure.

2. The adaptive cervical curvature traction system according to claim 1, characterized in that, The adaptive cervical curvature traction system also includes a data acquisition module, which comprises: The neck shape recognition unit is used to obtain the neck contour and calculate the neck length and neck arc chord value. The ranging unit includes an infrared sensor for acquiring head and neck position data; Temperature acquisition unit, used to collect ambient temperature data.

3. The adaptive cervical curvature traction system according to claim 1, characterized in that, The adjustment of the forehead band includes: Temperature compensation is performed on head and neck position data based on ambient temperature data; Based on the compensated head and neck position data, the ideal forehead-holding angle is calculated through a preset mapping model, and the current forehead-holding angle is compared to obtain the angle difference. The robotic arm is driven to adjust the angle of the forehead band based on the angle difference, and the force of the forehead band is adjusted in conjunction with the force of the forehead band collected by the second force sensor.

4. The adaptive cervical curvature traction system according to claim 2, characterized in that, The adjustment of the pressure applied to the forehead includes: The head weight data collected by the pressure flexible sensor group and the neck arc chord value calculated by the neck shape recognition unit are obtained, and the head weight data is filtered and steady-state detected. Based on the neck arc chord value, the arc compensation amount is obtained through a preset calculation rule; Based on the processed head weight data and the calculated arc compensation, the traction force is obtained through a preset traction force calculation model. The adjustment parameters of the dual lifting module are determined based on the obtained traction force; According to the adjustment parameters, the first lifting component and the second lifting component are driven to adjust the neck support strap's grip strength.

5. The adaptive cervical curvature traction system according to claim 2, characterized in that, The adjustable neck brace includes: Acquire head and neck position data collected by the ranging unit; Calculate the midline offset and the distance between the shoulders based on the head and neck position data; Based on the calculated midline offset and the distance between the shoulders, the target cervical spine position is obtained through a preset positioning model. Compare the target cervical vertebra position with the current position of the cervical brace, and calculate the offset; The first lifting assembly is driven to adjust the angle of the neck brace based on the calculated offset, and the adjustment effect is verified based on the data collected by the flexible pressure sensor group.

6. The adaptive cervical curvature traction system according to claim 1, characterized in that, The adaptive cervical curvature traction system also includes an early warning module, which uses the collected head and neck position data to calculate the displacement between the cervical brace and the forehead band, and executes the corresponding response strategy according to the preset three-level early warning mechanism.

7. The adaptive cervical curvature traction system according to claim 6, characterized in that, The three-level early warning mechanism includes: When the displacement reaches a first preset value, the position of the neck brace and the forehead band is automatically corrected. When the displacement reaches the second preset value, the second lifting component is activated to adjust the height of the neck brace to enhance the fixation effect; When the displacement reaches the third preset value, it is determined to be a high-risk slippage, and the system automatically restores the initial parameters and stops traction.

8. The adaptive cervical curvature traction system according to claim 1, characterized in that, The three-point mechanical structure consists of the force point of the robotic arm holding the forehead, the force point of the neck support holding the middle of the neck, and the force point of the back. The robotic arm's forehead-holding force point acts on the eyebrows. The robotic arm's forehead-holding force point can provide downward pulling force based on the cervical spine arc chord value and simultaneously adjust the extension angle. The cervical spine arc chord value is the vertical distance between the arc height and chord length of the target cervical spine physiological arc, which is used for the extension angle control correlation. The neck brace applies force to the middle segment of the cervical spine, and uses the distance between the two shoulders measured by infrared to locate the cervical spine midline to lock the middle segment of the cervical spine. It also adjusts the support height and angle in combination with the cervical spine chord value, which is used for bidirectional control of the angle in extension and flexion positions. The back pressure point acts on the back, using the back friction as the reverse support anchor point, which is used to correlate with the forward flexion angle adjustment.

9. The adaptive cervical curvature traction system according to claim 8, characterized in that, The forward flexion adjustment method is as follows: adjust the height of the neck brace by adjusting the dual lifting module, thereby changing the difference in height between the point of force on the neck and the point of force on the back, and thus adjusting the forward flexion angle.

10. The adaptive cervical curvature traction system according to claim 8, characterized in that, The method for adjusting the extension position is as follows: Calculate the deviation value based on the current cervical vertebral chord length value and the benchmark cervical vertebral chord length value; When the deviation value reaches the preset difference value, the extension angle is increased; When the deviation value is less than the preset difference value, the extension angle is adjusted to the preset value.

Images