Knee-joint postoperative exercise prescription generation system

Abstract

The invention discloses a knee joint postoperative exercise prescription generation system, belongs to the field of medical information health recovery, solves the problem that a personalized exercise prescription of a patient cannot be generated, and comprises a surgical data acquisition module for acquiring first surgical data of a knee joint surgery of the patient through an associated medical data system; the exercise prescription generation module is used for generating a personalized exercise prescription corresponding to the patient according to the first operation data; the exercise data acquisition module is used for acquiring exercise data when the patient executes the personalized exercise prescription through exercise data acquisition equipment worn by the patient, and performing analysis according to the exercise data to obtain physical sign data corresponding to the patient; and the exercise prescription updating module is used for updating the personalized exercise prescription according to the physical sign data. The individual condition of the patient is reflected through the first operation data, the personalized exercise prescription corresponding to the patient is generated, experience of doctors is not relied on any more, and the doctors can be prevented from neglecting some indexes due to subjective deviation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical information health recovery, and particularly relates to a motion prescription generation system for knee surgery. BACKGROUND

[0002] Motion prescription generally refers to a motion scheme formulated by a doctor, a rehabilitation therapist or the like according to the health status, motion ability and specific target of an individual patient after the patient is injured or has experienced treatment.

[0003] At present, the generation of motion prescription after knee surgery is still dominated by manual generation by a doctor or a rehabilitation therapist, which is mainly dependent on clinical evaluation of a patient and experience adjustment of a doctor, and is not only low in standardization but also too dependent on the experience of a doctor, and is difficult to effectively promote.

[0004] In some systems, a historical prescription of a similar case can be matched through a database to serve as a reference, and a motion prescription is generated through static templating. However, it lacks dynamic adaptation to a real-time patient state, resulting in lag in dynamic response, and it is difficult to consider the condition of each patient and generate a personalized motion prescription. SUMMARY

[0005] To solve the above problems, the present application provides a motion prescription generation system for knee surgery, which comprises a surgery data acquisition module, a motion prescription generation module, a motion data acquisition and analysis module and a motion prescription update module. The surgery data acquisition module acquires first surgery data of knee surgery of a patient through an associated medical data system. The motion prescription generation module generates a personalized motion prescription corresponding to the patient according to the first surgery data. The motion data acquisition module acquires motion data of the patient when performing the personalized motion prescription through a motion data acquisition device worn by the patient, and analyzes the motion data to obtain sign data corresponding to the patient. The motion prescription update module updates the personalized motion prescription according to the sign data.

[0006] In one example, the motion prescription generation module finds a plurality of second surgery data closest to the first surgery data in an established database according to the first surgery data, and determines historical motion prescriptions corresponding to the second surgery data. The data difference between the first surgery data and the second surgery data is determined. If the data difference of the second surgery data is lower than a preset difference degree, the personalized motion prescription corresponding to the patient is generated based on adjustment of the historical motion prescription according to the data difference. If all the data differences of the second surgery data are higher than the preset difference degree, a personalized exercise prescription corresponding to the patient is generated according to the first surgery data and recorded in the database.

[0007] In one example, the exercise prescription generation module, according to the knee surgery type of the first surgery data, searches for historical surgery data with the same knee surgery type in the database; According to the knee surgery type, a plurality of corresponding specified dimensions are selected in the preset plurality of surgery data dimensions; For each specified dimension, the difference degree of the first surgery data and each of the historical surgery data in the specified dimension is obtained; If the difference degree of all specified dimensions is lower than the preset confidence threshold, it is determined that the data difference is lower than the preset difference degree; Otherwise, it is determined that the data difference is higher than the preset difference degree.

[0008] In one example, the exercise prescription generation module, for each specified dimension, determines a dynamic confidence threshold corresponding to each specified dimension according to whether there is abnormal data in the specified dimension of the first surgery data; The dynamic confidence threshold is used as the preset confidence threshold corresponding to the specified dimension.

[0009] In one example, the exercise data acquisition module includes a wearable device; The wearable device includes a bundle, two ends of the bundle are respectively fixed on the thigh and lower leg of the patient, and the two ends are connected by an elastic connecting rope, and the bundle is provided with a knee guard at the knee joint position; A plurality of marker points are provided on the bundle, each marker point is embedded with a micro inertial measurement unit, the connecting rope is covered with an array of flexible strain sensors, and the two sides of the knee guard are respectively provided with a micro laser ranging.

[0010] In one example, the exercise data acquisition module establishes a spatial coordinate system according to the marker points, and establishes a state space equation according to the spatial coordinate system; According to the state space equation and the coordinate changes of the plurality of coordinate points, the corresponding knee angle state is obtained, and according to the feedback of the flexible strain sensor, the corresponding knee strength state is obtained.

[0011] In one example, the system further comprises an online doctor module; The online doctor module obtains feedback information of the patient, and assigns a corresponding online doctor according to the feedback information. According to the feedback information, the first surgery data of the patient is called by the surgery data acquisition module, and the personalized exercise prescription of the patient is called by the exercise prescription generation module and the exercise prescription update module, and the exercise data and the physical sign data of the patient are called by the exercise data acquisition module; The first surgery data, the personalized exercise prescription, the exercise data and the physical sign data are displayed to the online doctor, and a communication channel is established between the patient and the online doctor.

[0012] In one example, the system further comprises a VR module. The VR module comprises a VR device, and after the patient wears the VR device, the action contained in the personalized exercise prescription is displayed by the VR device.

[0013] The system provided in the present application can bring the following beneficial effects: 1. The individual situation of the patient is reflected through the first surgery data, and the personalized exercise prescription corresponding to the patient can be generated, which no longer depends on the experience of the doctor, and can also prevent the doctor from ignoring some indicators due to subjective bias.

[0014] 2. Through the multidimensional thinking of the first surgery data, the rigidity of the database template caused by matching only similar cases can be avoided, and the real-time state of the patient is ignored, so that a more personalized exercise prescription is generated for the patient.

[0015] 3. The generation process of the personalized exercise prescription is unified and standardized, which not only does not depend on the experience of the doctor, but also can be quickly promoted and used, and can be applied to other postoperative rehabilitation projects based on the demand by modifying part of the parameters. BRIEF DESCRIPTION OF DRAWINGS

[0016] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 The figure is the architecture diagram of the exercise prescription generation system for the knee joint postoperative in the embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described clearly and completely in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0018] The technical solutions provided by the embodiments of the present application are described in detail below with reference to the drawings.

[0019] As shown in the Figure 1 embodiments of the present application provide a knee surgery postoperative exercise prescription generation system, comprising: a surgery data acquisition module, an exercise prescription generation module, an exercise data acquisition and analysis module, an exercise prescription update module.

[0020] The surgery data acquisition module acquires first surgery data of the patient's knee surgery through the associated medical data system.

[0021] The first surgery data refers to the surgery data corresponding to the patient after the knee surgery, which can include preoperative data, intraoperative data, postoperative data, etc. Of course, each data in the first surgery data can be further divided to obtain multiple dimensions. For example, the dimensions of the preoperative data can include patient basic information (such as age, gender, weight, medical history, etc.), preoperative sign information (such as preoperative joint range of motion, preoperative muscle strength assessment, etc.), etc. The dimensions of the intraoperative data can include intraoperative imaging data, intraoperative sensor data, etc. The dimensions of the postoperative data can include postoperative sign information (such as postoperative joint range of motion, postoperative muscle strength assessment, postoperative inflammation, postoperative pain, etc.), etc.

[0022] The form of the first surgery data can include text, image, video, waveform, etc., which can be selected based on different data.

[0023] When acquiring the first surgery data, the medical data system can be associated through the interface under the premise of complying with the corresponding regulations and standards to acquire the corresponding data. The interface can adopt FHIR standard to interface with PACS / EMR system through OAuth2.0 authentication.

[0024] During the surgery, a corresponding biosensor network (such as a 9-axis IMU sensor (sampling rate ≥ 100 Hz) + flexible strain sensor + electromyography sensor) can be used to collect intraoperative data.

[0025] For some data that is difficult to automatically collect, it can also be obtained by manual input, such as postoperative pain, which can be obtained by asking the patient through a doctor's inquiry or a questionnaire survey and manually inputting.

[0026] Of course, in order to protect the privacy and informed consent of the patient, the patient needs to be communicated in advance to ensure that the patient agrees to use his surgery data to generate the exercise prescription. When using the data, data processing needs to be carried out in a trusted environment (including equipment, software, network, etc.) to prevent the patient's privacy from being leaked.

[0027] The motion prescription generation module generates a personalized motion prescription corresponding to the patient according to the first surgery data.

[0028] The motion prescription module can automatically generate a personalized motion prescription corresponding to the patient based on the first surgery data of the patient that has been obtained. The personalized motion prescription can specify the actions, times, and times required to be performed, or can give a general motion direction and motion target, and the patient can select a suitable action by himself.

[0029] Specifically, the motion prescription module can set two generation methods of personalized motion prescriptions, and select which method is more suitable through the current scene judgment.

[0030] The motion prescription generation module, according to the first surgery data, finds a plurality of second surgery data closest to the first surgery data in the established database, and determines the historical motion prescription corresponding to the second surgery data.

[0031] After the generation of each motion prescription, it and the corresponding surgery data can be stored in the corresponding database for subsequent reference.

[0032] When searching for the second surgery data, each dimension in the first surgery data can be directly compared with the dimensions of each surgery data stored in the database (for example, the cosine distance is calculated), and the second surgery data is obtained by weighted summation.

[0033] However, this method is too general and cannot consider the different needs of different patients.

[0034] Therefore, when searching for the second surgery data, the motion prescription generation module searches for historical surgery data with the same knee surgery type in the database according to the knee surgery type of the first surgery data.

[0035] Among them, the knee surgery type can include bone type, joint replacement type, minimally invasive type, etc. Different types of surgery often have large differences in postoperative recovery and required motion prescriptions, so the first round of rapid screening is performed by surgery type. In addition, the surgery type can be further classified according to the surgery position (for example, knee bone, meniscus, ligament, etc.).

[0036] According to the knee surgery type, a plurality of corresponding specified dimensions are selected from a plurality of preset surgery data dimensions.

[0037] For different surgery types, the corresponding dimensions can be pre-set as specified dimensions. The specified dimension refers to a dimension that needs to be considered first when generating a motion prescription under this surgery type, which has a greater impact.

[0038] For example, for bone type, in preoperative data, age and weight affect the speed of bone healing, preoperative joint range of motion and muscle strength determine the postoperative rehabilitation baseline, so age, weight, preoperative joint range of motion, preoperative muscle strength can be specified dimensions; in intraoperative data, if it is an osteotomy surgery, the osteotomy angle determines the adjustment range of the force line, the fixation method affects the postoperative weight bearing time, so the osteotomy angle, the fixation method (including steel plate, screw) can be specified dimensions; in postoperative data, the bone healing progress directly affects the rehabilitation stage, so the imaging results of bone healing can be selected as specified dimensions.

[0039] For joint replacement type, in preoperative data, osteoporosis affects the stability of prosthesis fixation, and preoperative range of motion determines the postoperative rehabilitation target, so age, osteoporosis degree, preoperative joint range of motion can be specified dimensions; in intraoperative data, the type of prosthesis determines the activity restriction (for example, PS-TKA needs to avoid deep flexion), and the retention of ligament affects the stability training design, so the type of prosthesis, the retention of ligament, and the intraoperative soft tissue balance data can be specified dimensions; in postoperative data, the stability of prosthesis determines the upper limit of exercise intensity, and the degree of swelling reflects the control of inflammation, so the imaging of prosthesis stability, postoperative joint range of motion, postoperative muscle strength, and postoperative swelling degree can be specified dimensions.

[0040] For minimally invasive type, in preoperative data, the type of injury determines the rehabilitation contraindication (for example, meniscus suture needs to limit early flexion), so the type of injury (including meniscus tear, ligament rupture) and preoperative joint stability can be specified dimensions; in intraoperative data, the suture strength affects the rehabilitation progress, so the repair method (including suture, resection, graft type) can be specified dimensions; in postoperative data, the healing stage determines the timing of advanced training, so the imaging of soft tissue healing, postoperative pain score, joint stability, and postoperative muscle strength can be specified dimensions.

[0041] Among these specified dimensions, there can be repeated dimensions, such as preoperative muscle strength, postoperative muscle strength, joint stability, etc.

[0042] Of course, it needs to be pointed out that the above specified dimensions are only an exemplary example, and in the actual judgment process of the specified dimensions, corresponding selection can be made according to the actual situation.

[0043] In this way, based on the comparison between these specified dimensions, a number of second surgery data closest to the first surgery data can be determined and sorted according to the similarity.

[0044] Next, the data difference between the first surgery data and the second surgery data is determined, so that the relationship between the data difference and the preset difference degree is used to judge which way is suitable for generating the individualized exercise prescription.

[0045] In the judgment of the relationship between the data difference and the preset difference degree, for each specified dimension, the difference degree of the first surgery data and each historical surgery data on the specified dimension is obtained.

[0046] For quantifiable data, such as joint range of motion, muscle strength data, etc., the difference degree can be determined directly through specific numerical values or through the difference between the set corresponding levels, and normalized. For data that is difficult to quantify, such as prosthesis type, injury type, etc., it can be considered that the difference degree is zero when they are consistent, and the difference degree is 1 (which is the difference degree after normalization) when they are inconsistent.

[0047] If the difference degree of all specified dimensions is lower than the preset confidence threshold, it is determined that the data difference is lower than the preset difference degree. Of course, this is to be rigorous, so the confidence threshold is selected to be lower than all specified dimensions, and in actual judgment, the requirement can also be appropriately relaxed based on the actual situation, and at least part (such as 80% or more) of the specified dimensions can be selected to be lower than the preset confidence threshold.

[0048] Otherwise, if there is a specified dimension whose difference degree exceeds the confidence threshold, it is determined that the data difference is higher than the preset difference degree.

[0049] The confidence threshold can be set to a fixed value, such as 0.1, or dynamically set based on the special circumstances of each specified dimension.

[0050] For example, the exercise prescription generation module determines the dynamic confidence threshold corresponding to each specified dimension according to whether there is abnormal data in the first surgery data in the specified dimension. Abnormal data refers to data that is too high, too low, or has special circumstances compared to the patient's normal physical data, such as preoperative muscle strength data that is too low, age that is too old, postoperative pain that is too severe, etc. For some specified dimensions (such as prosthesis type, fixation method), there is no such thing as abnormal data, so there is no need to consider it.

[0051] At this time, it is considered that the specified dimension with abnormal data has a greater impact on the patient's cause and postoperative recovery, and is relatively important for the generation of exercise prescription, so a more stringent and lower dynamic confidence threshold is set as the preset confidence threshold corresponding to the specified dimension. For example, when the default fixed value is 0.1, the dynamic confidence threshold of the specified dimension with abnormal data can be set to 0.05, so as to more strictly filter out similar second surgery data.

[0052] After comparison of the data differences, if the data differences of the second surgery data are all lower than the preset difference degree, it is considered that the two are very similar, and the first personalized exercise prescription generation method can be used, that is, according to the data differences, the historical exercise prescription is adjusted to generate the personalized exercise prescription corresponding to the patient. For most cases, the adjustment process of the historical exercise prescription can be adjusted manually, or according to the overall situation of the data differences (such as whether the physical sign data is more serious or healthier), the exercise amount of part of the exercise in the historical exercise prescription is appropriately increased or decreased. Or, the historical exercise prescription can be directly used as the personalized exercise prescription of the patient.

[0053] If all the data differences of the second surgery data are higher than the preset difference degree, it is considered that there is no very similar second surgery data, and the second personalized exercise prescription generation method is used, that is, the personalized exercise prescription corresponding to the patient is directly generated according to the first surgery data and recorded in the database.

[0054] In this way, through the two generation methods, when there is a matching suitable historical exercise prescription, it can also be directly used, since it is added to the database after verification, it is relatively stable. When there is no, the personalized exercise prescription of the patient can also be directly generated to bottom out.

[0055] The exercise data acquisition module acquires the exercise data of the patient performing the personalized exercise prescription through the exercise data acquisition device worn by the patient, and analyzes the corresponding physical sign data of the patient according to the exercise data.

[0056] When the personalized exercise prescription is generated, the patient can perform the exercise according to the action in the personalized exercise prescription, so as to recover. Or, according to the provisions in the personalized exercise prescription (such as requiring the knee joint activity to be within a certain range, or requiring the lower body to be not intense aerobic exercise), the corresponding action is selected to perform exercise rehabilitation.

[0057] At this time, in the exercise process of the patient, the corresponding exercise data acquisition device needs to be worn to collect exercise data, so as to monitor the exercise process and update the personalized exercise prescription.

[0058] Specifically, the exercise data acquisition module includes a wearable device.

[0059] The wearable device includes a bundle, the two ends of the bundle are respectively fixed on the thighs and shanks of the patient, and the two ends are connected by an elastic connecting rope. The bundle is provided with a knee pad at the knee joint position.

[0060] The wearable device can be customized by 3D printing based on the leg type of the patient, and can be optionally integrated with a phase change material temperature control layer to automatically adjust the temperature of the contact surface.

[0061] The bundle is provided with a plurality of marker points, each of which is embedded with a miniature inertial measurement unit (which can be provided with an integrated accelerometer + gyroscope + magnetometer (9-axis) for inertial measurement), and the connecting rope is covered with an array of flexible strain sensors, and the knee pad is provided with a miniature laser ranging on both sides.

[0062] A configuration scheme of the marker points is provided, which can be adjusted based on requirements in actual schemes. The configuration scheme can include: Four anatomical reference layers are provided, including one each for the medial and lateral condyles of the femur (which can be provided with infrared LED + inertial measurement unit IMU), and one each for the tibial tuberosity and fibular head (which can be provided as piezoelectric contact sensors).

[0063] Six dynamic tracking layers are provided, including three arranged in a ring around the patella (which can include miniature laser emitters), and three distributed along the popliteal fossa (which can be provided with flexible strain sensor arrays).

[0064] Two auxiliary verification layers are provided, including one each for the middle thigh and lower leg (which can carry RFID temperature and humidity sensors).

[0065] Through multi-level settings and data collection, the collection process of motion data can be more accurate, and corresponding indication functions (such as LED light prompting function, temperature and humidity detection function) can be added to prompt the user that motion data is being detected and the user's environment temperature and humidity are being detected to determine whether the user is in a suitable temperature and humidity environment for exercise.

[0066] When the user wears the wearable device for exercise, the motion data collection module establishes a spatial coordinate system based on the marker points and establishes a state space equation based on the spatial coordinate system.

[0067] First, the global coordinate system is constructed, and when the patient stands (or for difficult-to-stand knees, the knee can be in a standard action such as sitting or lying down), the marker point near the hip joint is defined as the origin of the global coordinate system (for example, the X-axis is set to horizontal forward, the Y-axis is set to vertical upward, and the Z-axis is set to horizontal right).

[0068] Then, the local coordinate system is aligned, and the other marker points are calibrated to align their respective coordinate systems to the global coordinate system.

[0069] Finally, the spatial position and attitude of each marker point are updated in real time using gyroscope integration and accelerometer data. Of course, during this process, laser rangefinders on both sides of the knee brace can be used to assist in calibration, measuring the real-time distance between the inner and outer sides of the knee joint. This data is then combined with the measured data to correct coordinate system drift errors caused by limb slippage or deformation of the straps.

[0070] Based on the state-space equation and the coordinate changes of multiple coordinate points, the corresponding knee angle state is obtained; and based on the feedback from the flexible strain sensor, the corresponding knee force state is obtained.

[0071] First, define the state variables. For example, angular states include: knee flexion / extension angle (θ), varus / valgus angle (φ), and rotation angle (ψ). Dynamic states include: angular velocity ( ), angular acceleration ( The observed variables include: the acceleration and angular velocity of the IMU, the laser ranging value, and the deformation of the strain sensor.

[0072] Kinematic equations are constructed based on rigid body kinematics, converting the data from each marker point into joint motion parameters. For example, the knee joint angle is calculated using inverse kinematics based on the positional differences (Δx, Δy, Δz) between marker points on the thigh and lower leg. .

[0073] Then, optionally, Kalman filtering fusion is used to suppress integral drift by predicting the angle at the next moment based on the gyroscope angular velocity integral in the prediction step, and correcting the predicted value with accelerometer (gravity direction) and laser ranging data in the update step.

[0074] For knee angle calculation, accelerometer, gyroscope, and magnetometer data from each marker point can be fused into a quaternion for attitude representation, and then converted into Euler angles. The difference in Euler angles between the thigh and lower leg marker points represents the knee flexion / extension angle. For example, let the thigh attitude quaternion be... The quaternion for lower leg posture is Then relative rotation Thus, the buckling angle θ can be extracted.

[0075] Additionally, selective corrections can be made using laser ranging. When the medial knee joint space narrows during flexion, the laser ranging value decreases, allowing the establishment of a distance-angle lookup table (LUT) or fitting formula. When the difference between the measured angle and the laser ranging-calculated angle exceeds a threshold, dynamic calibration is triggered (e.g., by weighted averaging the two).

[0076] For the knee strength state, strain-force calibration can be performed, and the strain sensor array of the connecting rope measures the deformation (ε), and the strain and tension relationship (F = k*ε+b) is calibrated through pre-experiment. Wherein, b is a constant in the calibration relationship. At this time, when the tensile deformation increases, the sensor output ε increases, and the tension F increases.

[0077] For the array composed of multiple strain sensors, the tension distribution diagram of the connecting rope can be constructed, and the resultant force borne by the knee joint can be calculated combined with the geometric layout of the binding belt (such as the angle between the elastic rope and the limb). For example, according to the limb acceleration and mass distribution measured by the inertial measurement unit (the weight of the binding is known in advance), the inertial force is calculated, and then combined with the tension data of the strain sensor, the active force of the knee joint and the external load component are separated.

[0078] Based on this, the moment estimation equation is established by Newton-Euler equation: τ = F•d + I•α, wherein F is the resultant force measured by the strain sensor (that is, the tension mentioned above), d is the force arm length (which can be calculated from the position of the marker point in the spatial coordinate system), I is the moment of inertia (which can be obtained based on the mass model of the patient's limb), and α is the angular acceleration (which can be obtained by twice differentiating the data measured by the inertial detection unit).

[0079] The moment τ obtained in this way can be used as the overall force of the patient's knee, which is the knee strength state.

[0080] The exercise prescription updating module updates the individualized exercise prescription according to the physical sign data.

[0081] After measuring the physical sign data of the patient (including the knee angle state and the knee strength state), the individualized exercise prescription can be updated. For example, if the physical sign data of the patient is stable and recovered, the individualized exercise prescription can not be updated. If the physical sign data of the patient is abnormal, for example, the knee strength state is continuously low and the recovery is slow, the exercise intensity of the subsequent action related to the knee strength is adjusted to adapt to the slow recovery speed of the patient. On the contrary, if the recovery speed of the patient exceeds the expectation, the subsequent exercise intensity can be increased.

[0082] 1. The first surgery data reflects the individual situation of the patient, which can generate the corresponding individualized exercise prescription of the patient, no longer relying on the experience of the doctor, and can also prevent the doctor from ignoring some indicators due to subjective bias.

[0083] 2. Through multidimensional thinking of the first surgery data, the database template rigidity and the neglect of the real-time state of the patient caused by matching only similar cases can be avoided, and a more personalized exercise prescription that matches the patient can be generated.

[0084] 3. The generation process of the personalized exercise prescription is unified and standardized, which is not dependent on the experience of doctors, can be quickly promoted and used, and can be applied to other postoperative rehabilitation projects based on the modification of part of the parameters.

[0085] In one embodiment, as shown in Figure 1 the system further comprises an online doctor module.

[0086] The online doctor module obtains feedback information of the patient and assigns a corresponding online doctor according to the feedback information.

[0087] The online doctor module can be implemented based on the corresponding application program (such as based on an e-commerce platform). After the patient registers an account in the application program, the patient can initiate online communication with the doctor through the application program, or the application program can initiate an online communication request to the doctor when it is detected that the frequency of abnormal signs data is high, or the degree of abnormal signs data is high.

[0088] If the patient initiates the feedback, the feedback information is input by the patient. If the online communication request is initiated by the system, the corresponding feedback information is automatically generated according to the request generation reason after the patient agrees, which can include abnormal signs data.

[0089] At this time, according to the feedback information, the first operation data of the patient is called through the operation data acquisition module, and the personalized exercise prescription of the patient is called through the exercise prescription generation module and the exercise prescription update module, and the exercise data and the signs data of the patient are called through the exercise data acquisition module.

[0090] The first operation data, the personalized exercise prescription, the exercise data, and the signs data are displayed to the online doctor, and a communication channel is established between the patient and the online doctor. The online doctor can quickly review the relevant information of the patient, thereby preliminarily understanding the situation of the patient, so as to facilitate subsequent rapid communication. During the communication process, the online doctor can answer questions, guide actions, make suggestions, etc. for the patient, so as to ensure the safety and efficiency of the patient during the execution of the exercise prescription.

[0091] The communication channel can be in the form of an online chat room, and the two parties can communicate in the form of text, voice, video shooting, online call, online video, etc.

[0092] After initiating the communication, the corresponding fee can be charged for this communication based on the demand, which can be paid by the patient, or paid by the e-commerce platform, or the situation can also be determined (for example, when the patient initiates it, the patient pays, and when the data is abnormal, the generation system of the exercise prescription can replace the patient to pay the fee).

[0093] By linking the online doctor module with an e-commerce platform, if the doctor suggests the purchase of relevant medications, instruments, equipment, etc. during the communication process, the e-commerce platform can be used to make recommendations, so as to facilitate the patient's quick purchase and prevent the patient from buying items that do not meet their requirements.

[0094] In one embodiment, such as Figure 1 As shown, the system also includes a VR module.

[0095] VR modules include VR devices, such as VR headsets and controllers.

[0096] After wearing the VR device, patients can experience a VR demonstration of the movements included in their personalized exercise prescription. This VR demonstration allows patients to gain a deeper and more detailed understanding of the movements, thus improving their ability to perform the exercises correctly.

[0097] In one embodiment, as mentioned above, when generating a personalized exercise prescription for a patient, the first approach can generate a personalized exercise prescription based on adjustments and modifications to the amount of exercise in historical exercise prescriptions. However, this description is relatively general and does not present a standardized process.

[0098] Based on this, a standardized process is provided and scientifically and rigorously adjusted and modified to better suit the current situation of patients.

[0099] Specifically, for each exercise stage in the historical exercise prescription, the corresponding action features are extracted based on the action templates contained therein.

[0100] Typically, exercise prescriptions include multiple exercise phases, each with varying intensities that increase progressively. An exercise template, on the other hand, is a pre-defined type of exercise, not necessarily involving specific intensity. For example, an exercise template might include passive flexion and extension, straight leg raises, single-leg standing, and cycling. Each exercise template involves a different direction of movement and aims to achieve a different training objective. Movement features can be extracted beforehand for each template (e.g., through a large language model) and stored along with the template itself. For instance, passive flexion and extension features include joint range of motion, straight leg raises feature muscle strengthening, single-leg standing features balance, and cycling features low-impact aerobic exercise.

[0101] For each action feature, determine its corresponding specified dimension.

[0102] As mentioned above, the specified dimension refers to the corresponding dimension in the surgical data. The correspondence between the two means that when considering the motion intensity of the motion template in each motion feature, the specific values ​​of which specified dimensions need to be referenced to set the motion intensity.

[0103] For example, regarding the movement characteristic of joint range of motion, the preservation or removal of ligaments directly affects joint stability and range of motion limitations; therefore, the intraoperative ligament management method can be considered a corresponding designated dimension. Furthermore, poor preoperative range of motion (e.g., long-term flexion contracture) requires a prolonged training cycle and gradual increase in intensity; therefore, preoperative joint range of motion can also be considered a corresponding designated dimension. Regarding the movement characteristic of muscle strength enhancement, the level of muscle strength recovery directly determines the resistance intensity (e.g., when muscle strength reaches 50% of the healthy side, light resistance can be introduced); therefore, postoperative muscle strength assessment can be considered a corresponding designated dimension. Furthermore, when healing is incomplete, weight-bearing intensity needs to be limited; therefore, bone healing and soft tissue healing imaging can be considered corresponding designated dimensions. Regarding the movement characteristic of balance, the degree of ligament laxity during surgery determines the difficulty of balance training; therefore, the results of intraoperative ligament stability testing can be considered a corresponding designated dimension. Furthermore, when the stability of the prosthesis or repaired tissue is insufficient, dynamic balance training needs to be delayed; therefore, postoperative joint stability can be considered a corresponding designated dimension. Given the characteristics of aerobic exercise, obese patients need to choose low-impact aerobic exercise to reduce joint stress, so body weight can be used as the corresponding dimension; and pain level determines the intensity of aerobic exercise, so postoperative pain can be used as the corresponding dimension.

[0104] Of course, this is just an example of some specified dimensions. In actual situations, specified dimensions can be deleted or modified based on requirements.

[0105] Once the correspondence between the specified dimensions and motion features is obtained, it can be stored.

[0106] Based on the degree of difference in each specified dimension and the preset weight under the action feature, the corresponding adjustment factor is obtained by weighted summation.

[0107] The correspondence between each specified dimension and action feature has been explained above. Here, you can further set the preset weights between the two. The higher the weight, the greater the influence of the specified dimension on the action feature. If no setting is made, the preset weights of each specified dimension are considered to be the same.

[0108] The exercise intensity of the movement template corresponding to the movement feature is adjusted based on the adjustment factor to obtain a personalized exercise prescription.

[0109] For each motion feature, the adjustment factor is obtained by weighting the degree of difference with a preset weight. The higher the adjustment factor, the greater the adjustment range of the motion intensity for that motion feature. Here, the direction of the motion intensity adjustment can be determined by specifying the positive or negative value of the adjustment factor, and the motion intensity can be adjusted proportionally according to the specific value of the adjustment factor.

[0110] Exercise intensity can include exercise duration, number of repetitions, and range of motion.

[0111] In one embodiment, when generating a personalized exercise prescription for a patient directly based on the first surgical data in a second manner, all pre-set movement features can be determined, and multiple movement stages can be generated based on the type of knee surgery (the number of movement stages is usually the default, for example, set to four stages). In each movement stage, constraints on the movement features are generated based on the first surgical data.

[0112] The restrictions can include type restrictions and intensity restrictions. Type restrictions can include which motion features cannot be executed, or which action templates within a motion feature cannot be executed. Intensity restrictions refer to the total intensity of each motion exercise not exceeding a certain value.

[0113] In each motion phase, motion features that meet the requirements are selected based on the type restrictions in the constraints. When multiple types of motion features can be executed in a certain motion phase, motion features that have not appeared in previous phases are executed first.

[0114] For the required motion characteristics, a matching motion template is selected from the preset motion library, and the motion template is assigned the corresponding motion intensity according to the intensity limit in the constraints.

[0115] As mentioned above, different movement characteristics correspond to different movement templates. For the same movement characteristic, a combination of multiple movement templates can be selected first to increase the fun of the patient's exercise process.

[0116] Each motion template has a standard exercise intensity (for example, it has a corresponding standard exercise intensity when performing a standard motion, standard number of repetitions, or standard time). By modifying the number of repetitions, time, etc., the corresponding exercise intensity can be assigned.

[0117] Based on the action template with assigned exercise intensity, the exercise prescription for each exercise stage is obtained, and the exercise prescriptions for all exercise stages are combined to obtain the corresponding personalized exercise prescription.

[0118] When multiple movement features can exist in a movement phase, multiple movement features are usually selected to provide comprehensive exercise for the patient. In this case, a corresponding movement template is selected for each movement feature and combined to obtain a combination of movement template and movement intensity in each movement phase, thereby generating a corresponding personalized exercise prescription.

[0119] Furthermore, when selecting a matching action template from a preset action library for actions that meet the requirements, the selection can be random or made according to the user's situation to better adapt to the user's environment.

[0120] For each eligible motion feature, a first motion template is obtained by filtering through the motion templates in the preset motion library. All the selected first motion templates correspond to the given motion feature; motion templates corresponding to motion features that do not meet the requirements are then removed.

[0121] In the first exercise template, the second exercise template is obtained by filtering based on the exercise environment uploaded by the patient and the environmental requirements of each first exercise template. The exercise environment can include the range of activity, available equipment, etc., thereby filtering out movement templates that do not meet the exercise environment (for example, some movement templates require the use of special equipment).

[0122] In the second set of motion templates, at least one motion template is selected as the motion template for matching this motion feature. If multiple motion templates remain in the second set, two can be selected to increase the enjoyment for the patient. If only one motion template remains, that template is selected. If no motion template is available, one can be selected from the previously filtered motion templates, and the user should be informed of any additional requirements for that motion template (e.g., necessary equipment).

[0123] In one embodiment, during the patient's exercise according to the personalized exercise prescription, the exercise prescription update module obtains the corresponding patient feedback information and corresponding vital sign data for each exercise stage in the personalized exercise prescription.

[0124] If the patient’s feedback and / or vital signs data are abnormal, the intensity of the exercise at the current stage of exercise will be adjusted (usually by reducing the intensity of exercise).

[0125] Patient feedback can be sent to the exercise prescription generation system through the corresponding application. As long as the feedback type is marked as abnormal (for example, the pain level during exercise is high), it can be considered abnormal.

[0126] When abnormalities in vital signs data are detected frequently, or when abnormalities in vital signs data are detected to be of a high degree, the vital signs data are considered to be abnormal.

[0127] If it is determined that adjustments are needed for multiple consecutive exercise phases (e.g., two consecutive exercise phases), then the current personalized exercise prescription is considered to have significant problems.

[0128] At this point, for the personalized exercise prescription generated by modifying the historical exercise prescription, a new personalized exercise prescription for the patient is generated based on the first surgical data; for the personalized exercise prescription generated based on the first surgical data, it is reported through the online doctor module.

[0129] For personalized exercise prescriptions generated by modifying historical exercise prescriptions, which can be considered cases where the historical exercise prescription does not match the patient, a personalized exercise prescription is directly generated using the first surgical data. If a personalized exercise prescription is generated directly, it is reported through the online doctor module for manual processing by the doctor.

[0130] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of this application.

Claims

1. A system for generating exercise prescriptions after knee surgery, characterized in that, include: Surgical data acquisition module, exercise prescription generation module, exercise data collection and analysis module, exercise prescription update module; The surgical data acquisition module acquires the first surgical data of the patient's knee joint surgery through the associated medical data system; The exercise prescription generation module generates a personalized exercise prescription for the patient based on the first surgical data. The exercise data acquisition module collects exercise data from the patient while the patient is executing the personalized exercise prescription through an exercise data acquisition device worn by the patient, and analyzes the exercise data to obtain the patient's corresponding vital sign data. The exercise prescription update module updates the personalized exercise prescription based on the vital sign data.

2. The exercise prescription generation system for post-knee joint surgery according to claim 1, characterized in that, The exercise prescription generation module, based on the first surgical data, searches for several second surgical data that are closest to the first surgical data in the established database, and determines the historical exercise prescription corresponding to the second surgical data. Determine the data differences between the first surgical data and the second surgical data; If the data difference of the second surgical data is lower than the preset difference level, then based on the data difference, a personalized exercise prescription corresponding to the patient is generated by adjusting the historical exercise prescription. If the differences in all second surgical data exceed a preset difference level, a personalized exercise prescription for the patient is generated based on the first surgical data and recorded in the database.

3. The exercise prescription generation system for post-knee joint surgery according to claim 2, characterized in that, The exercise prescription generation module searches for historical surgical data with the same knee surgery type in the database based on the knee surgery type of the first surgical data. Based on the type of knee surgery, select several corresponding specified dimensions from a preset range of surgical data dimensions; For each specified dimension, the degree of difference between the first surgical data and each of the historical surgical data on that specified dimension is obtained; If the degree of difference in all specified dimensions is lower than the preset confidence threshold, then the data difference is determined to be lower than the preset degree of difference. Otherwise, the data difference is determined to be higher than a preset difference level.

4. The exercise prescription generation system for post-knee joint surgery according to claim 3, characterized in that, The exercise prescription generation module determines the dynamic confidence threshold for each specified dimension based on whether there is abnormal data in the first surgical data for that specified dimension. The dynamic confidence threshold is used as the preset confidence threshold corresponding to the specified dimension.

5. The exercise prescription generation system for post-knee joint surgery according to claim 1, characterized in that, The motion data acquisition module includes a wearable device; The wearable device includes a strap, with both ends of the strap being worn and fixed on the patient's thigh and calf respectively, and the two ends being connected by an elastic connecting rope. The strap is equipped with a knee brace at the knee joint. The binding is equipped with multiple marker points, each of which contains a miniature inertial measurement unit. The connecting rope is covered by an array of flexible strain sensors, and miniature laser ranging devices are installed on both sides of the knee pad.

6. The exercise prescription generation system for post-knee joint surgery according to claim 5, characterized in that, The motion data acquisition module establishes a spatial coordinate system based on the marked points and establishes a state space equation based on the spatial coordinate system. Based on the state space equation and the coordinate changes of the multiple coordinate points, the corresponding knee angle state is obtained; The corresponding knee force state is obtained based on the feedback from the flexible strain sensor.

7. The exercise prescription generation system for post-knee joint surgery according to claim 1, characterized in that, The system also includes: an online doctor module; The online doctor module obtains patient feedback information and assigns a corresponding online doctor based on the feedback information; Based on the feedback information, the patient's first surgical data is retrieved through the surgical data acquisition module, the patient's personalized exercise prescription is retrieved through the exercise prescription generation module and the exercise prescription update module, and the patient's exercise data and vital sign data are retrieved through the exercise data acquisition module. The first surgical data, the personalized exercise prescription, the exercise data, and the vital sign data are displayed to the online doctor, and a communication channel is established between the patient and the online doctor.

8. The exercise prescription generation system for post-knee joint surgery according to claim 1, characterized in that, The system also includes: a VR module; The VR module includes a VR device. After the patient wears the VR device, the movements included in the personalized exercise prescription are displayed in VR through the VR device.

Images