Medical management system and method based on artificial intelligence
By generating personalized rehabilitation goals through an artificial intelligence system, and monitoring and dynamically adjusting them in real time, the system addresses the shortcomings of traditional orthopedic postoperative rehabilitation programs in terms of personalization and interdisciplinary collaboration. It achieves a balance between pain and functional recovery, thereby improving rehabilitation outcomes and patient engagement.
Patent Information
- Application Number
- CN202511443632.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-10
- Publication Date
- 2026-01-06
AI Technical Summary
Traditional orthopedic postoperative rehabilitation programs lack personalization, cannot be dynamically adjusted, and lack interdisciplinary collaboration, resulting in unsatisfactory recovery outcomes, an imbalance between pain and functional recovery, and delays in the program due to changes in the patient's condition. There are also problems of over- or under-training during the rehabilitation process.
The AI-based medical management system generates personalized rehabilitation goals through data collection, sensor monitoring, and dynamic adjustments. It balances pain and functional recovery in real time, facilitates interdisciplinary collaboration, dynamic grouping, and task adjustment, and enables multi-dimensional monitoring of rehabilitation progress.
It improves the personalization and efficiency of rehabilitation programs, ensures a balance between pain and functional recovery, reduces overtraining or undertraining, enhances interdisciplinary support, improves rehabilitation outcomes and patient engagement, and avoids program delays due to changes in patient condition.
Smart Images

Figure CN121281830A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical management technology, specifically to a medical management system and method based on artificial intelligence. Background Technology
[0002] Postoperative rehabilitation after orthopedic surgery is a complex and ongoing process, involving multiple aspects from wound healing to functional recovery, muscle strength enhancement, and improved joint flexibility. The success of postoperative rehabilitation directly impacts a patient's quality of life and future mobility. Therefore, how to ensure patients recover their health as quickly as possible through scientific rehabilitation methods has become an important research area in the modern medical field.
[0003] In current technologies, post-orthopedic rehabilitation often relies on manually formulated rehabilitation plans, with medical staff manually adjusting the program based on individual patient differences and recovery progress. However, this manual intervention approach often presents several problems. First, manually formulated rehabilitation plans lack personalization and often fail to adapt to changes in the patient's recovery process. Second, as rehabilitation progresses, traditional methods cannot dynamically adjust the plan in a timely manner, leading to unsatisfactory recovery outcomes. Third, the patient's recovery process is influenced by psychological state, pain perception, physical strength, and other social factors, which traditional methods struggle to effectively monitor and regulate. Finally, traditional rehabilitation models neglect the potential for interdisciplinary collaboration, often relying solely on medical staff from a single field to develop rehabilitation plans, thus overlooking the patient's comprehensive needs.
[0004] Therefore, this case aims to propose an AI-based medical management system and method that generates personalized rehabilitation plans based on the patient's specific condition. It introduces a dynamic pain management model and a functional recovery monitoring mechanism to adjust training intensity and pain control strategies in the rehabilitation plan in real time, maintaining the optimal balance between pain and functional recovery and avoiding rehabilitation stagnation caused by over-fatigue or pain. Advanced sensor technology is used to monitor the stress distribution of the patient's bones and soft tissues, ensuring that movement during rehabilitation does not lead to excessive load and helping patients avoid secondary injuries. By comprehensively considering multiple dimensions of factors such as the patient's physiological data, motor ability, and psychological state, a comprehensive patient rehabilitation process model is established to evaluate and adjust rehabilitation strategies in real time. Summary of the Invention
[0005] This invention provides an artificial intelligence-based medical management system and method, which helps to solve the problems mentioned in the background art.
[0006] This invention provides the following technical solution: a medical management method based on artificial intelligence, comprising:
[0007] Data is collected from patients' health records, postoperative assessments, and sensor devices;
[0008] The collected data includes authorized basic patient information, postoperative assessment data, and exercise assessment data;
[0009] The authorized basic patient information includes age, gender, weight, height, and activity level, with the activity level divided into three levels: low, medium, and high.
[0010] The postoperative assessment data includes the type of surgery, postoperative date, and surgical complications;
[0011] The motion assessment data includes joint range of motion A(t), pain score P(t), muscle strength M(t), and stress on bones and soft tissues. ;
[0012] The system uses wearable sensors to record physiological data on patients' postoperative physical activity, gait, and joint movement in real time.
[0013] Based on the collected data, a personalized set of rehabilitation goals is set for the patient, denoted as R(t), which includes:
[0014] Functional recovery goal: restore joint range of motion. and gait recovery ;
[0015] The goal of muscle recovery is to gradually increase muscle strength. ;
[0016] Pain management goals, controlling pain scores ;
[0017] Mental health goals to alleviate postoperative anxiety in patients ;
[0018] Set the target function: ;
[0019] in, The overall rehabilitation goal over time t. For the recovery status of the i-th type of target in set R(t), represents the weighting coefficients for various objectives, and n is the total number of elements in set R(t);
[0020] A preliminary rehabilitation plan is generated based on rehabilitation goals:
[0021] Set a set of rehabilitation training tasks, denoted as Q;
[0022] Each planned task is generated as follows:
[0023] ;
[0024] in, A rehabilitation training plan for time t; Let i be the progress of the i-th training task in set Q; The weighting coefficient for each task; The total number of elements in the set of rehabilitation training tasks;
[0025] A dynamic pain and functional recovery balance strategy is set up to provide real-time feedback on pain and functional recovery, ensuring that the patient's pain and recovery are balanced during the rehabilitation process;
[0026] By monitoring the patient's bones and soft tissues in real time, data can be provided for adjusting the rehabilitation plan;
[0027] The patient's recovery process is modeled from multiple dimensions to further quantify and monitor the recovery effect;
[0028] By dynamically adjusting team members and tasks, interaction and cooperation among patients are enhanced, creating a comprehensive and supportive rehabilitation environment.
[0029] Optionally, the dynamic pain and functional recovery balancing strategy specifically includes:
[0030] The progress of function recovery at time t is denoted as F(t);
[0031] Establish a model of the relationship between pain score P(t) and functional recovery progress F(t);
[0032] The relationship between pain score P(t) and functional recovery progress F(t) is represented by the following function:
[0033] ;
[0034] in, The intensity of pain management measures in the treatment plan is specifically defined as follows:
[0035] ;
[0036] in, Indicates time Total dose of analgesics received by the patient, in units of ; Indicates time Total duration of physical therapy received by patients, in units of ; This indicates the number of cold compress treatments the patient received within a given time period, expressed in units of times; each cold compress treatment is counted as one unit.
[0037] in, The pain score at time t; and The coefficient representing the influence between pain and recovery progress; This is the progress of function recovery at time t, calculated using the following formula:
[0038] ;
[0039] in, The range of motion of the joint at time t; The target range of activities set during the rehabilitation process; Muscle strength at time t; The target muscle strength set during the rehabilitation process; These are the joint motion weighting coefficients; The weighting factor for muscle strength;
[0040] Based on pain score and the progress of function recovery The training plan is dynamically adjusted.
[0041] Optionally, the dynamic adjustment of the training plan specifically includes:
[0042] Set a pain score threshold, denoted as... ;
[0043] Let the training intensity at time t be S(t);
[0044] Set the base training intensity, denoted as... ;
[0045] when When this happens, reduce the training intensity S(t);
[0046] Training intensity is adjusted according to the training intensity adjustment formula:
[0047] ;
[0048] in, This represents the coefficient relating training intensity to pain scores.
[0049] Optionally, the step of providing data for adjusting the rehabilitation plan by real-time monitoring of the patient's bones and soft tissues specifically includes:
[0050] Wearable sensors monitor the stress data of bones and soft tissues during rehabilitation and record the stress data in real time. ;
[0051] in, For position The force situation at time t;
[0052] Obtain the center point of the monitored bone and establish a three-dimensional coordinate system with the center point as the origin;
[0053] Obtain the total length of the monitored skeleton, denoted as L;
[0054] Obtain the total width of the monitored bone, denoted as W;
[0055] Obtain the total height of the monitored skeleton, denoted as H;
[0056] Get location The three-dimensional coordinates of the location are denoted as ;
[0057] in The coordinates of the x-axis in a three-dimensional coordinate system; The coordinates of the y-axis in a three-dimensional coordinate system; The coordinates of the z-axis in a three-dimensional coordinate system;
[0058] Assume the force distribution passes through the location Let time t represent the mathematical model:
[0059] ;
[0060] in, This represents the total force on the entire skeletal model at time t. In the three-dimensional coordinate system at time t The total local stress value at the location;
[0061] The patient's load is dynamically adjusted based on the force distribution in the bones and soft tissues.
[0062] Optionally, the dynamic adjustment of the patient's load specifically includes:
[0063] Set the base load, denoted as... ;
[0064] The dynamic load control model is as follows:
[0065] ;
[0066] in, For the adjusted load; This is the coefficient relating the load and the force applied.
[0067] Optionally, the process of modeling the patient's rehabilitation process through multiple dimensions to further quantify and monitor the rehabilitation effect specifically includes:
[0068] Set up a rehabilitation progress set, including joint range of motion. Pain score and muscle strength Add to the recovery progress collection;
[0069] The patient's psychological state at time t is recorded as follows: ;
[0070] The multidimensional model of rehabilitation progress is as follows:
[0071] ;
[0072] in, The overall recovery progress at time t; , , and These are the weights for each progress dimension;
[0073] By optimizing the objective function, we ensure that the recovery progress gradually approaches the target. The optimized objective function is as follows:
[0074] ;
[0075] in, Let be the value of the i-th progress indicator in the set of recovery progress at time t; Let n be the target value of the i-th progress indicator in the set of recovery progress indicators; n is the number of elements in the set of recovery progress indicators. Let be the weight of the i-th progress indicator in the set of recovery progress indicators.
[0076] Optionally, the process of dynamically adjusting group members and tasks enhances interaction and cooperation among patients, creating a comprehensive and supportive rehabilitation environment, specifically including:
[0077] Obtain all patients who need rehabilitation;
[0078] Based on the patient's rehabilitation progress and goals, patients are assigned to different rehabilitation groups, and the assignments are dynamically adjusted through cluster analysis and progress models, specifically as follows:
[0079] Calculate the distance between each patient's recovery progress and their goals:
[0080] Let the recovery progress of each patient at time t be... and the progress towards the ideal goal The distance between the two can be calculated using the following formula:
[0081] ;
[0082] in, This represents the j-th item of the recovery progress of the i-th patient at time t; Let j be the ideal value of the target rehabilitation progress term; n is the total number of target terms.
[0083] Based on the differences in recovery progress among patients, the K-means clustering algorithm was used to group patients. The goal of the K-means clustering algorithm is to minimize the differences in patient progress within a group. The clustering steps are as follows:
[0084] S1. Initialize K cluster centers, i.e., select K patients as initial representatives;
[0085] S2. For each patient, calculate and compare their distance to the center of each cluster, and assign the patient to the cluster with the closest distance.
[0086] S3. Update the center of each cluster, that is, calculate the average recovery progress of all patients in the cluster and use it as the new cluster center;
[0087] S4. Repeat steps S2 and S3 until the cluster center no longer changes;
[0088] When all patients have been traversed, stop the traversal and record each cluster obtained as a rehabilitation group.
[0089] A dynamic adjustment and optimization strategy is adopted to dynamically adjust the composition of each group and the allocation of tasks.
[0090] Optionally, the dynamic adjustment and optimization strategy for dynamically adjusting the composition and task allocation of each group specifically includes:
[0091] For any given rehabilitation group:
[0092] Set the adjustment interval time;
[0093] Set the initial time;
[0094] Starting from the initial time, after each adjustment interval, the group's rehabilitation progress, interaction score, and rehabilitation goals are obtained;
[0095] Set a threshold for differences in recovery progress;
[0096] Calculate the average rehabilitation progress of all patients in the rehabilitation group;
[0097] Calculate the difference between the rehabilitation progress of each member in the rehabilitation group and the average rehabilitation progress, and take the absolute value as the average difference in patient rehabilitation progress.
[0098] If the average difference in patient recovery progress among members exceeds the recovery progress difference threshold, then that member will be removed from the recovery group and a new member will be reassigned.
[0099] A system for implementing the aforementioned AI-based medical management method includes:
[0100] Data collection module: Collects data from patients' health records, postoperative assessments, and sensor devices;
[0101] Sensor module: Uses wearable sensors to record physiological data of the patient's postoperative body activity, gait and joint movement in real time;
[0102] Calculation module: Used to perform data calculations.
[0103] The present invention has the following beneficial effects:
[0104] 1. By collecting patient health records, postoperative assessments, and sensor data, the system addresses the lack of personalization in rehabilitation plans. Utilizing basic patient information, postoperative assessment data, and movement assessment data, the system accurately understands the patient's postoperative condition and rehabilitation needs, generating a personalized set of rehabilitation goals. This process ensures that the rehabilitation plan is not only based on standardized data but also precisely adjusted according to the patient's specific condition, avoiding the one-size-fits-all problem of traditional plans. Setting a personalized set of rehabilitation goals solves the problem of insufficient precise goal management in the rehabilitation process. Including functional recovery, muscle recovery, pain management, and mental health as multiple rehabilitation goals, and dynamically adjusting them based on the patient's postoperative condition, ensures that each goal is tailored to the patient's rehabilitation needs. This not only improves rehabilitation efficiency but also effectively avoids overly vague or singular rehabilitation goals, ensuring comprehensive support for the patient in all aspects. By setting objective functions, the system resolves the trade-offs and balance issues between goals. By setting objective functions and introducing weighting coefficients, the priority of various goals can be dynamically adjusted, ensuring that the patient's overall rehabilitation goals are balanced at different stages. For example, in the early stages, the focus may be more on pain control and joint mobility recovery, while in the later stages, the focus may shift to strengthening muscle strength and gait recovery. This weighted mechanism allows rehabilitation programs to be flexibly adjusted according to the patient's actual needs, avoiding imbalances caused by over-concentration on a single goal. Real-time monitoring and dynamic adjustment address instability issues during rehabilitation. Real-time monitoring based on wearable sensors collects physiological data such as joint movement and gait, allowing for real-time adjustments to the rehabilitation program. This real-time monitoring capability effectively prevents the program from becoming outdated or unsuitable due to changes in the patient's condition. Balancing strategies for pain and functional recovery ensure patient comfort and effectiveness during rehabilitation. Interdisciplinary integration addresses the lack of comprehensive support in rehabilitation programs. Rehabilitation requires not only medical treatment but also collaboration from fields such as psychology and kinesiology. This interdisciplinary integration provides patients with more comprehensive support. For example, rehabilitation programs not only focus on physical recovery but also consider mental health and emotional management, helping patients reduce postoperative anxiety and increase confidence in rehabilitation. This measure significantly improves rehabilitation outcomes and reduces the negative impact of psychological factors on the rehabilitation process. Dynamic team-based rehabilitation programs address insufficient interaction and motivation during rehabilitation. By dynamically adjusting team members and tasks, patients can mutually motivate and support each other. Interaction among group members helps improve patient participation and motivation for rehabilitation, enhancing the effectiveness of group rehabilitation. Especially in the context of group training, patients can maintain high motivation through comparison and collaboration with others, avoiding the isolation and inefficiency of individual training.
[0105] 2. By establishing a model relating pain scores to functional recovery progress, the lack of dynamic balance between pain and recovery during rehabilitation is addressed. This model precisely quantifies the impact of pain on functional recovery, enabling real-time, personalized rehabilitation adjustments for each patient. It dynamically reflects the interplay between pain and functional recovery during rehabilitation, avoiding the problem of traditional rehabilitation programs focusing solely on recovery progress while neglecting pain management. The introduction of pain management intensity addresses the issue of insufficient pain control during rehabilitation. This intensity measures the strength of pain management measures in the treatment plan, such as medication and cold compresses. Based on different pain scores, the intensity of pain management in the treatment plan can be dynamically adjusted to ensure effective pain control during rehabilitation. Especially in the early stages of rehabilitation, patients may experience significant pain; adjusting the intensity can effectively reduce discomfort and prevent excessive pain from hindering their rehabilitation progress. By calculating the functional recovery progress, the system addresses the problem of inaccurate setting and adjustment of rehabilitation goals. This calculation considers not only the recovery of joint range of motion and muscle strength but also incorporates weighting coefficients for each goal, ensuring a reasonable and comprehensive assessment of the recovery progress for each objective. Through formula adjustments, the system can accurately calculate the patient's recovery progress at a given moment and adjust the rehabilitation plan promptly based on the patient's actual performance, making it more aligned with their needs. This adjustment mechanism avoids the problem of traditional methods where single-goal settings cannot dynamically adapt to the patient's recovery progress. By dynamically adjusting the training plan, the system solves the problem of recovery being too fast or too slow during rehabilitation. By calculating the relationship between pain scores and functional recovery progress in real time, the system can dynamically adjust the rehabilitation training plan based on the patient's pain level and functional recovery status. If the patient's pain score is high, the system will reduce the training intensity, allowing more time for recovery; if the functional recovery progress is slow, the training intensity will be appropriately increased to ensure that the patient does not miss the optimal rehabilitation window due to slow recovery. This dynamic adjustment avoids the problem of recovery being too fast or too slow during rehabilitation, ensuring that the patient's rehabilitation process is always conducted in a reasonable and healthy state.
[0106] 3. By setting pain score thresholds, the problem of excessive pain during rehabilitation is addressed. Setting these thresholds triggers an adjustment mechanism when the patient's pain score reaches a certain level. This mechanism prevents patients from experiencing excessive pain during rehabilitation. Specifically, when the pain score exceeds the set threshold, the system promptly identifies and initiates the adjustment process, thus avoiding severe pain caused by overtraining, which could lead to patient discomfort or delays in rehabilitation. This step effectively prevents adverse reactions from overtraining in the early stages of rehabilitation, ensuring patients can rehabilitate comfortably. Adjusting training intensity addresses the issues of overtraining or undertraining during rehabilitation. The dynamic adjustment of training intensity is based on the patient's pain score. When the patient's pain score is high, the training intensity automatically decreases to alleviate discomfort; conversely, when the patient's pain score is low and recovery progress is good, the training intensity increases to promote functional recovery. This adjustment mechanism solves the problem of overtraining or undertraining in traditional rehabilitation programs, ensuring that the training intensity always matches the patient's rehabilitation status and pain level, avoiding instability in rehabilitation progress. By setting a basic training intensity, personalized and progressive training is ensured. This is achieved by adjusting the basic intensity based on the patient's specific condition and rehabilitation progress, ensuring the training program starts from the basics and gradually increases in difficulty. This progressive training model effectively improves patient tolerance and avoids injury caused by overly intense initial training. Simultaneously, the basic training intensity provides a starting point for the patient, preventing the initial training intensity from being too low and failing to achieve effective recovery. The training intensity adjustment formula addresses the mismatch between training intensity and pain scores during rehabilitation. The formula includes a coefficient... This is a coefficient relating training intensity and pain score. Through formula design, when a patient's pain score is high, the formula automatically lowers the training intensity; conversely, when the pain score is low, the training intensity is appropriately increased. This formula achieves a real-time dynamic balance between training intensity and pain score, ensuring that the patient is always at an appropriate load level during training, avoiding overload or undertraining. Through this adjustment, the patient's rehabilitation process is smoother, reducing additional burden on the body and improving the effectiveness of rehabilitation.
[0107] 4. By monitoring bone and soft tissue stress data in real time, this method addresses the issue of excessive or uneven load on bones and soft tissues during rehabilitation. Wearing sensors monitors the stress data of bones and soft tissues during rehabilitation, allowing for real-time acquisition of the patient's stress at each point in time, with a particular focus on stress distribution. This real-time recorded stress data provides a precise basis for adjusting rehabilitation plans, helping doctors and rehabilitation experts better understand the patient's physical load. Traditional rehabilitation methods often rely on patient self-perception and subjective feedback, easily overlooking potential stress problems such as localized excessive stress or uneven load, leading to incomplete rehabilitation or the risk of injury. This step accurately captures the stress at each location, providing detailed data support for subsequent rehabilitation plans. The precise representation of stress distribution using a mathematical model solves the problem of insufficient precision in traditional stress analysis. Using a mathematical model to represent the stress at time t and location x quantifies complex stress distributions into analyzable data. This model provides rehabilitation experts with intuitive and accurate stress analysis results, enabling them to clearly understand the specific stress distribution of the patient during rehabilitation. This precise modeling of force distribution avoids the ambiguity of traditional methods that cannot quantify force, facilitating the customization of more personalized rehabilitation plans based on the patient's actual force state. By dynamically adjusting the load, the system addresses the problem of inadequate force on the patient's bones and soft tissues during rehabilitation. Based on real-time monitored force data and distribution, the system can dynamically adjust the patient's load. For example, when the force on a certain area is too high, the system can promptly adjust the load, reducing the training intensity or changing the training method for that area; conversely, when the force on a certain area is low, the training load can be appropriately increased. This dynamic adjustment mechanism ensures that the load on each area remains at a suitable level during rehabilitation, avoiding the adverse effects of excessive or insufficient load on the rehabilitation process and ensuring that bones and soft tissues recover under reasonable pressure. By analyzing the spatial distribution of force data, the system optimizes rehabilitation plans, avoiding risks caused by uneven force. By representing the force distribution using location x and time t, it can capture the force differences in different areas and adjust the training plan based on the data. This meticulous management approach minimizes the risk of excessive stress on localized areas, especially during joint and soft tissue recovery, preventing new injuries or pathological changes caused by uneven loading during rehabilitation. Through real-time feedback and adjustments, the patient's rehabilitation process becomes more balanced and safer.
[0108] 5. By setting a baseline load, the problem of unreasonable load settings in traditional rehabilitation programs is solved. A suitable starting load is provided for each patient. This baseline load can be adjusted according to the patient's individual situation, surgical type, and postoperative rehabilitation status, avoiding the one-size-fits-all load setting in traditional methods. The baseline load provides a standard for subsequent dynamic adjustments, allowing patients to start with an appropriate load in the early stages of rehabilitation and gradually increase the training intensity. This setting avoids patients failing to recover effectively due to too low a load, and also avoids secondary injuries that may be caused by too high a load. The dynamic load control model solves the problems of untimely and inaccurate load adjustments during rehabilitation. The dynamic load control model dynamically adjusts the load using time t and force. The coefficients in the model... As a coefficient relating load and force, the load can be precisely adjusted based on changes in force. This dynamic adjustment mechanism responds in real time to changes in the patient's physical condition, ensuring that the load remains at a reasonable level throughout the rehabilitation process. For example, when the patient's force is low, the load can be appropriately increased to promote faster recovery; conversely, when the patient's force is too high, the load will be reduced to prevent new injuries caused by overload. This refined, real-time load adjustment method significantly improves the safety and effectiveness of the rehabilitation process. By optimizing the relationship between load and force, the problem of overload or underload during rehabilitation is solved through the relationship coefficient. The new system allows for more flexible and personalized adjustments between load and force. Traditional rehabilitation methods often adjust load based on the patient's subjective feelings and fixed training plans, which may fail to reflect changes in force in real time, leading to excessively high or low loads. However, through precise control of the load-force relationship, the load can be adjusted based on real-time feedback from force data, ensuring that each training stage does not place excessive stress on the patient and avoiding insufficient load during recovery. This allows for better achievement of rehabilitation outcomes and avoids the problem of traditional rehabilitation programs failing to consider individual patient differences. Dynamic load adjustment enhances rehabilitation effectiveness and avoids imbalances in the patient's recovery process. Dynamic load adjustment not only helps reduce the risk of injury but also promotes smooth rehabilitation progress. Flexible load adjustment ensures that the training load required by each patient at different rehabilitation stages is reasonably controlled, resulting in a smoother rehabilitation process. Timely load adjustment allows patients to gradually improve their physical adaptability without overloading, helping them achieve faster recovery. It avoids overtraining or excessively restrictive training, maximizing rehabilitation effects.
[0109] 6. By setting up a rehabilitation progress set, the lack of multi-dimensional and comprehensive assessment in traditional rehabilitation progress monitoring is addressed. By adding multiple key indicators such as joint range of motion, pain score, and muscle strength to the rehabilitation progress set, the patient's rehabilitation status can be monitored in real time across multiple dimensions. Traditional methods typically focus on only a single dimension, such as pain score or joint range of motion, while neglecting the patient's overall rehabilitation progress. The multi-dimensional set not only assesses physiological recovery progress but also considers non-physiological factors such as psychological state, thus providing a comprehensive rehabilitation progress assessment. This method overcomes the limitation of the singular nature of traditional rehabilitation assessments, ensuring comprehensive monitoring of the rehabilitation process. Optimizing the objective function through a multi-dimensional model solves the problems of unbalanced rehabilitation progress optimization and inconsistent progress objectives. During rehabilitation, the recovery of progress indicators such as joint range of motion and muscle strength is not always synchronous, and pain management and improvement in psychological state may also affect the speed of physiological recovery. By constructing a multi-dimensional rehabilitation progress model and optimizing the objective function, the recovery status of each dimension can be balanced, allowing the rehabilitation progress to be optimized towards multi-objective consistency. This method avoids over- or under-recovery in any single dimension, ensuring coordinated progress in joints, muscles, pain, and psychological aspects, thus achieving a comprehensive and balanced rehabilitation effect. By optimizing the objective function, it ensures that rehabilitation progress gradually approaches the goal, avoiding arbitrariness and inconsistency in rehabilitation plans. In traditional rehabilitation methods, progress assessment may only involve periodic checks, lacking a dynamic optimization mechanism. However, by setting an objective function to optimize rehabilitation progress, goals and strategies can be dynamically adjusted according to the patient's actual rehabilitation situation, ensuring that each progress indicator gradually approaches the set target value. This process makes the rehabilitation plan more scientific and systematic, capable of adaptive adjustments based on individual differences among patients. This continuous dynamic adjustment capability significantly improves the accuracy and efficiency of rehabilitation, reducing the problem of unsatisfactory results caused by progress deviations. By incorporating psychological state, it addresses the impact of patient psychological factors on rehabilitation outcomes. During the rehabilitation process, the patient's psychological state often affects the speed and effect of their physiological recovery, especially in the early stages of postoperative rehabilitation. By acquiring patient psychological state data and incorporating it into a multi-dimensional model, the overall rehabilitation situation of the patient can be more accurately reflected, and optimization adjustments can be made within the objective function. This strategy not only allows for the quantification and tracking of patients' physiological recovery, but also enables the effective assessment and management of the impact of their psychological state on the rehabilitation process. Consequently, effective measures can be taken to improve their psychological state and promote the recovery of physiological functions.
[0110] 7. By dynamically grouping patients based on rehabilitation progress and goals, the problem of excessive differences in patient progress is addressed. In orthopedic postoperative rehabilitation, patients exhibit significant differences in their rehabilitation progress. Traditional one-size-fits-all rehabilitation group settings cannot meet the needs of different patients, potentially leading to slower progress or less than ideal recovery for some, and affecting the overall treatment effect of the group. The dynamic grouping method based on patient rehabilitation progress and goals first calculates the distance between each patient's rehabilitation progress and ideal goal, and then uses the K-means clustering algorithm to divide patients into different groups according to their progress differences, ensuring that the difference in rehabilitation progress within each group is minimized. This method optimizes the patient grouping strategy, making patients in each rehabilitation group similar in their rehabilitation goals, avoiding the imbalance caused by excessive progress differences, thereby improving the overall rehabilitation effect. The K-means clustering algorithm is a current technology. Furthermore, the K-means clustering algorithm addresses the adverse effects of individual differences among patients on rehabilitation group interactions. Traditional rehabilitation groups are usually assigned based on the patient's condition or treatment stage, ignoring the differences in rehabilitation progress among individual patients. By introducing the K-means clustering algorithm, dynamic grouping can be performed based on the actual rehabilitation progress and goal differences of patients, minimizing the progress differences within each rehabilitation group. This method optimizes the composition of the group members, ensuring that patients within the group have similar rehabilitation needs and goals, thereby promoting interaction and cooperation within the group. Patients can exchange experiences at similar rehabilitation stages, increasing their motivation and confidence, and further improving rehabilitation outcomes. By dynamically adjusting group members and tasks, the limitations of a single rehabilitation program and its lack of flexibility are overcome. As the rehabilitation process progresses, patients' rehabilitation progress changes. Traditional rehabilitation programs typically use static settings for group members and task allocation, lacking flexibility to adjust according to the actual progress of patients. However, through a dynamic adjustment strategy based on patients' rehabilitation progress, group members and task allocation can be adjusted in real time according to patients' rehabilitation effects, target progress, and grouping. Specifically, as patients' rehabilitation progress changes, they may be reassigned to different rehabilitation groups or their rehabilitation tasks may be adjusted. This dynamic optimization strategy ensures that each patient is rehabilitated in a group and with tasks suitable for them, improving the personalization and flexibility of the rehabilitation program. Dynamic task allocation improves the collaborative efficiency and effectiveness of the rehabilitation group. In the rehabilitation group, not only is interaction between patients crucial, but the rationality of task allocation also directly affects the rehabilitation outcome. By dynamically adjusting task allocation strategies based on progress differences, the collaboration and adaptability of each team member in completing tasks can be ensured. For example, the task intensity may be reduced for patients with slower progress, while the task intensity may be increased for patients with faster progress.In this way, patients can obtain the most suitable rehabilitation tasks according to their current abilities and needs, avoiding discomfort or slow progress caused by tasks that are too light or too heavy, thereby further improving the overall collaborative efficiency of the rehabilitation team.
[0111] 8. By calculating the differences in patients' recovery progress, the problem of excessive differences among members of the rehabilitation group was addressed. Within the rehabilitation group, there are significant individual differences in patients' recovery progress, especially due to variations in each patient's physical condition and recovery speed, leading to large disparities among group members. This not only affects patient recovery outcomes but may also impact cooperation and interaction within the group. By calculating the difference between each patient's recovery progress and the group's average progress, and removing patients with excessively large progress differences based on a preset threshold, members with significant differences are effectively prevented from remaining in the same group. This ensures that the recovery progress of group members is relatively similar, thereby improving overall collaboration and progress consistency within the group. Attached Figure Description
[0112] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation
[0113] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0114] Example, refer to Figure 1 An artificial intelligence-based healthcare management method includes:
[0115] Data is collected from patients' health records, postoperative assessments, and sensor devices;
[0116] The collected data includes authorized basic patient information, postoperative assessment data, and exercise assessment data;
[0117] The authorized basic patient information includes age, gender, weight, height, and activity level, with the activity level divided into three levels: low, medium, and high.
[0118] The postoperative assessment data includes the type of surgery, postoperative date, and surgical complications;
[0119] The motion assessment data includes joint range of motion A(t), pain score P(t), muscle strength M(t), and stress on bones and soft tissues. ;
[0120] The system uses wearable sensors to record physiological data on patients' postoperative physical activity, gait, and joint movement in real time.
[0121] Based on the collected data, a personalized set of rehabilitation goals is set for the patient, denoted as R(t), which includes:
[0122] Functional recovery goal: restore joint range of motion. and gait recovery ;
[0123] The goal of muscle recovery is to gradually increase muscle strength. ;
[0124] Pain management goals, controlling pain scores Ensure that the pain level is within an acceptable range;
[0125] Mental health goals to alleviate postoperative anxiety in patients ;
[0126] Set the target function: ;
[0127] in, The overall rehabilitation goal over time t. For the recovery status of the i-th type of target in set R(t), represents the weighting coefficients for various objectives, and n is the total number of elements in set R(t);
[0128] A preliminary rehabilitation plan is generated based on rehabilitation goals:
[0129] Set a set of rehabilitation training tasks, denoted as Q;
[0130] Each planned task is generated as follows:
[0131] ;
[0132] in, A rehabilitation training plan for time t; Let i be the progress of the i-th training task in set Q; The weighting coefficient for each task; The total number of elements in the set of rehabilitation training tasks;
[0133] A dynamic pain and functional recovery balance strategy is set up to provide real-time feedback on pain and functional recovery, ensuring that the patient's pain and recovery are balanced during the rehabilitation process;
[0134] By monitoring the patient's bones and soft tissues in real time, data can be provided for adjusting the rehabilitation plan;
[0135] The patient's recovery process is modeled from multiple dimensions to further quantify and monitor the recovery effect;
[0136] By dynamically adjusting team members and tasks, interaction and cooperation among patients are enhanced, creating a comprehensive and supportive rehabilitation environment.
[0137] By collecting patient health records, postoperative assessments, and sensor data, the system addresses the lack of personalization in rehabilitation plans. Utilizing basic patient information, postoperative assessment data, and movement assessment data, the system accurately understands the patient's postoperative condition and rehabilitation needs, generating a personalized set of rehabilitation goals. This process ensures that the rehabilitation plan is not only based on standardized data but also precisely adjusted according to the patient's specific condition, avoiding the one-size-fits-all problem of traditional plans. Setting personalized rehabilitation goal sets solves the problem of insufficient precise goal management in the rehabilitation process. Integrating functional recovery, muscle recovery, pain management, and mental health as multiple rehabilitation goals, and dynamically adjusting them based on the patient's postoperative condition, ensures that each goal is tailored to the patient's rehabilitation needs. This not only improves rehabilitation efficiency but also effectively avoids overly vague or singular rehabilitation goals, ensuring comprehensive support for the patient in all aspects. By setting objective functions, the system resolves the trade-offs and balance issues between goals. By setting objective functions and introducing weighting coefficients, the priority of various goals can be dynamically adjusted, ensuring that the patient's overall rehabilitation goals are balanced at different stages. For example, in the early stages, the focus may be more on pain control and joint mobility recovery, while in the later stages, the focus may shift to strengthening muscle strength and gait recovery. This weighted mechanism allows rehabilitation programs to be flexibly adjusted according to the patient's actual needs, avoiding imbalances caused by over-concentration on a single goal. Real-time monitoring and dynamic adjustment address instability issues during rehabilitation. Real-time monitoring based on wearable sensors collects physiological data such as joint movement and gait, allowing for real-time adjustments to the rehabilitation program. This real-time monitoring capability effectively prevents the program from becoming outdated or unsuitable due to changes in the patient's condition. Balancing strategies for pain and functional recovery ensure patient comfort and effectiveness during rehabilitation. Interdisciplinary integration addresses the lack of comprehensive support in rehabilitation programs. Rehabilitation requires not only medical treatment but also collaboration from fields such as psychology and kinesiology. This interdisciplinary integration provides patients with more comprehensive support. For example, rehabilitation programs not only focus on physical recovery but also consider mental health and emotional management, helping patients reduce postoperative anxiety and increase confidence in rehabilitation. This measure significantly improves rehabilitation outcomes and reduces the negative impact of psychological factors on the rehabilitation process. Dynamic team-based rehabilitation programs address insufficient interaction and motivation during rehabilitation. By dynamically adjusting team members and tasks, patients can mutually motivate and support each other. Interaction among group members helps improve patient participation and motivation for rehabilitation, enhancing the effectiveness of group rehabilitation. Especially in the context of group training, patients can maintain high motivation through comparison and collaboration with others, avoiding the isolation and inefficiency of individual training.
[0138] The dynamic pain and functional recovery balancing strategy specifically includes:
[0139] The progress of function recovery at time t is denoted as F(t);
[0140] Establish a model of the relationship between pain score P(t) and functional recovery progress F(t);
[0141] The relationship between pain score P(t) and functional recovery progress F(t) is represented by the following function:
[0142] ;
[0143] in, The intensity of pain management measures in the treatment plan is specifically defined as follows:
[0144] ;
[0145] in, Indicates time Total dose of analgesics received by the patient, in units of ; Indicates time Total duration of physical therapy received by patients, in units of ; This indicates the number of times the patient received cold compress therapy within a given time period, measured in times; each cold compress therapy session is counted as one unit.
[0146] in, The pain score at time t; and The coefficient of influence between pain and recovery progress is determined as follows:
[0147] Clinical observation and data collection: A large number of patients were followed up to record their responses at different recovery stages and pain levels. For example, the range of motion of the patient's joints was monitored. and muscle strength How does the level of pain change when recovery is good, and vice versa?
[0148] Regression analysis: Using regression analysis methods, such as linear regression or multinomial regression, the relationship between patient recovery progress and pain is analyzed. Specifically, the recovery progress... and pain score The data were paired to calculate the coefficient that best describes the impact of recovery progress on pain. Similarly, data analysis was used to determine the inhibitory effect of changes in pain on recovery progress. The specific mathematical formula is: By fitting the data using the least squares method, we obtain... and The value of .
[0149] Personalized adjustments: Because each patient's physiological condition is different, and The value may vary from person to person, and therefore can be adjusted individually based on the patient's recovery. For example, for patients with a high pain tolerance, The value may be large; while for patients with slower recovery, The value may be large.
[0150] Clinical validation: Estimates obtained through actual clinical trials. and The goal is to determine whether the values can accurately predict the relationship between pain and recovery, thereby optimizing the specific values of these two coefficients.
[0151] This is the progress of function recovery at time t, calculated using the following formula:
[0152] ;
[0153] in, The range of motion of the joint at time t; The target range of activities set during the rehabilitation process; Muscle strength at time t; The target muscle strength set during the rehabilitation process; These are the joint motion weighting coefficients; The weighting factor for muscle strength;
[0154] The joint motion weighting coefficient and the muscle strength weighting coefficient were determined using the following methods:
[0155] In the early stages of rehabilitation, the improvement rate of two indicators was recorded based on preliminary training data over 3-5 consecutive days: Let the improvement in joint mobility be... , ,in At the initial moment, This is the first calibration time; let the improvement in muscle strength be... , .
[0156] Then, calculate the relative improvement rate of the two improvement magnitudes to their respective targets:
[0157] ; ;
[0158] ; .
[0159] Based on pain score and the progress of function recovery The training plan is dynamically adjusted.
[0160] By establishing a model of the relationship between pain scores and functional recovery progress, the lack of dynamic balance between pain and recovery progress during rehabilitation is addressed. This model allows for precise quantification of the impact of pain on functional recovery, enabling real-time, personalized rehabilitation adjustments for each patient. It dynamically reflects the interplay between pain and functional recovery during rehabilitation, avoiding the problem of traditional rehabilitation programs focusing solely on recovery progress while neglecting pain management. The introduction of pain management intensity addresses the issue of insufficient pain control during rehabilitation. This intensity measures the strength of pain management measures in the treatment plan, such as medication and cold compresses. Based on different pain scores, the pain management intensity in the treatment plan can be dynamically adjusted to ensure effective pain control during rehabilitation. Especially in the early stages of rehabilitation, patients may experience significant pain; adjusting the intensity can effectively reduce discomfort and prevent excessive pain from hindering their rehabilitation progress. By calculating the functional recovery progress, the system addresses the problem of inaccurate setting and adjustment of rehabilitation goals. This calculation considers not only the recovery of joint range of motion and muscle strength but also incorporates weighting coefficients for each goal, ensuring a reasonable and comprehensive assessment of the recovery progress for each objective. Through formula adjustments, the system can accurately calculate the patient's recovery progress at a given moment and adjust the rehabilitation plan promptly based on the patient's actual performance, making it more aligned with their needs. This adjustment mechanism avoids the problem of traditional methods where single-goal settings cannot dynamically adapt to the patient's recovery progress. By dynamically adjusting the training plan, the system solves the problem of recovery being too fast or too slow during rehabilitation. By calculating the relationship between pain scores and functional recovery progress in real time, the system can dynamically adjust the rehabilitation training plan based on the patient's pain level and functional recovery status. If the patient's pain score is high, the system will reduce the training intensity, allowing more time for recovery; if the functional recovery progress is slow, the training intensity will be appropriately increased to ensure that the patient does not miss the optimal rehabilitation window due to slow recovery. This dynamic adjustment avoids the problem of recovery being too fast or too slow during rehabilitation, ensuring that the patient's rehabilitation process is always conducted in a reasonable and healthy state.
[0161] The dynamic adjustment of the training plan specifically includes:
[0162] Set a pain score threshold, denoted as... ;
[0163] Let the training intensity at time t be S(t);
[0164] Set the base training intensity, denoted as... ;
[0165] when When this happens, reduce the training intensity S(t);
[0166] Training intensity is adjusted according to the training intensity adjustment formula:
[0167] ;
[0168] in, The coefficient representing the relationship between training intensity and pain score;
[0169] The relationship between training intensity and pain score was determined as follows:
[0170] Data collection: Collect different pain scores through clinical practice. Training intensity data.
[0171] Regression analysis: Using linear or nonlinear regression methods, the relationship between pain scores and training intensity is modeled to obtain coefficients. The mathematical model is ;
[0172] Personalized adjustments: The value can be affected by individual patient differences. For example, some patients have a higher pain tolerance and may require an appropriate increase. This allows for maintaining a high training intensity even in cases of pain; however, for patients who are sensitive to pain, the intensity should be reduced. It relies more on pain control.
[0173] Clinical validation: Validation and adjustment in a clinical setting. The value allows it to adaptively adjust the training intensity according to the patient's actual situation, ensuring that the training plan can effectively stimulate rehabilitation while avoiding excessive burden that causes pain.
[0174] By setting pain score thresholds, the problem of excessive pain during rehabilitation is addressed. These thresholds trigger an adjustment mechanism when a patient's pain score reaches a certain level. This mechanism prevents patients from experiencing excessive pain during rehabilitation, especially when the pain score exceeds the set threshold. The system promptly identifies and initiates the adjustment process, preventing severe pain caused by overtraining, which could lead to patient discomfort or delays in rehabilitation. This step effectively prevents adverse reactions from overtraining in the early stages of rehabilitation, ensuring patients can recover comfortably. Adjusting training intensity addresses the issues of overtraining or undertraining during rehabilitation. The dynamic adjustment of training intensity is based on the patient's pain score. When the patient's pain score is high, the training intensity automatically decreases to alleviate discomfort; conversely, when the patient's pain score is low and recovery progress is good, the training intensity increases to promote functional recovery. This adjustment mechanism solves the problems of overtraining or undertraining in traditional rehabilitation programs, ensuring that the training intensity always matches the patient's rehabilitation status and pain level, avoiding instability in rehabilitation progress. By setting a basic training intensity, personalized and progressive training is ensured. This is achieved by adjusting the basic intensity based on the patient's specific condition and rehabilitation progress, ensuring the training program starts from the basics and gradually increases in difficulty. This progressive training model effectively improves patient tolerance and avoids injury caused by overly intense initial training. Simultaneously, the basic training intensity provides a starting point for the patient, preventing the initial training intensity from being too low and failing to achieve effective recovery. The training intensity adjustment formula addresses the mismatch between training intensity and pain scores during rehabilitation. The formula includes a coefficient... This is a coefficient relating training intensity and pain score. Through formula design, when a patient's pain score is high, the formula automatically lowers the training intensity; conversely, when the pain score is low, the training intensity is appropriately increased. This formula achieves a real-time dynamic balance between training intensity and pain score, ensuring that the patient is always at an appropriate load level during training, avoiding overload or undertraining. Through this adjustment, the patient's rehabilitation process is smoother, reducing additional burden on the body and improving the effectiveness of rehabilitation.
[0175] The method involves real-time monitoring of the patient's bones and soft tissues to provide data for adjusting the rehabilitation plan, specifically including:
[0176] Wearable sensors monitor the stress data of bones and soft tissues during rehabilitation and record the stress data in real time. ;
[0177] in, For position The force situation at time t;
[0178] Obtain the center point of the monitored bone and establish a three-dimensional coordinate system with the center point as the origin;
[0179] Obtain the total length of the monitored skeleton, denoted as L;
[0180] Obtain the total width of the monitored bone, denoted as W;
[0181] Obtain the total height of the monitored skeleton, denoted as H;
[0182] Get location The three-dimensional coordinates of the location are denoted as ;
[0183] in The coordinates of the x-axis in a three-dimensional coordinate system; The coordinates of the y-axis in a three-dimensional coordinate system; The coordinates of the z-axis in a three-dimensional coordinate system;
[0184] Assume the force distribution passes through the location Let time t represent the mathematical model:
[0185] ;
[0186] in, This represents the total force on the entire skeletal model at time t. In the three-dimensional coordinate system at time t The total local stress value at the location;
[0187] The patient's load is dynamically adjusted based on the force distribution in the bones and soft tissues.
[0188] By monitoring bone and soft tissue stress data in real time, this method addresses the issue of excessive or uneven load on bones and soft tissues during rehabilitation. Wearing sensors monitors the stress data of bones and soft tissues during rehabilitation, allowing for real-time acquisition of the patient's stress at every point in time, with a particular focus on stress distribution. This real-time recorded stress data provides a precise basis for adjusting rehabilitation plans, helping doctors and rehabilitation specialists better understand the patient's physical load. Traditional rehabilitation methods often rely on patient self-perception and subjective feedback, easily overlooking potential stress problems such as localized excessive stress or uneven load, leading to incomplete rehabilitation or the risk of injury. This step precisely captures the stress at each location, providing detailed data support for subsequent rehabilitation plans. The precise representation of stress distribution using a mathematical model solves the problem of insufficient precision in traditional rehabilitation stress analysis. Using a mathematical model to represent the stress at time t and location x quantifies complex stress distributions into analyzable data. This model provides rehabilitation specialists with intuitive and accurate stress analysis results, enabling them to clearly understand the specific distribution of stress on the patient during rehabilitation. This precise modeling of force distribution avoids the ambiguity of traditional methods that cannot quantify force, facilitating the customization of more personalized rehabilitation plans based on the patient's actual force state. By dynamically adjusting the load, the system addresses the problem of inadequate force on the patient's bones and soft tissues during rehabilitation. Based on real-time monitored force data and distribution, the system can dynamically adjust the patient's load. For example, when the force on a certain area is too high, the system can promptly adjust the load, reducing the training intensity or changing the training method for that area; conversely, when the force on a certain area is low, the training load can be appropriately increased. This dynamic adjustment mechanism ensures that the load on each area remains at a suitable level during rehabilitation, avoiding the adverse effects of excessive or insufficient load on the rehabilitation process and ensuring that bones and soft tissues recover under reasonable pressure. By analyzing the spatial distribution of force data, the system optimizes rehabilitation plans, avoiding risks caused by uneven force. By representing the force distribution using location x and time t, it can capture the force differences in different areas and adjust the training plan based on the data. This meticulous management approach minimizes the risk of excessive stress on localized areas, especially during joint and soft tissue recovery, preventing new injuries or pathological changes caused by uneven loading during rehabilitation. Through real-time feedback and adjustments, the patient's rehabilitation process becomes more balanced and safer.
[0189] The dynamic adjustment of the patient's workload specifically includes:
[0190] Set the base load, denoted as... ;
[0191] The dynamic load control model is as follows:
[0192] ;
[0193] in, For the adjusted load; This is the coefficient relating the load and the force applied.
[0194] The relationship coefficient between load and force is determined by the following method:
[0195] The experiment was designed to have patients exercise at different training intensities, and to record stress data and patient sensations using sensors. The experiment also involved recording different loads. Changes in lower force They also observed the patient's comfort and pain perception, and the correlation between load and stress, thereby inferring... The initial range, based on experimental data, can be estimated using regression analysis, such as linear regression. The initial value can be obtained using the following simple regression model. In this way, the following can be estimated The initial value of the regression model is used to determine the specific linear relationship between load and stress, based on... The initial value can be adjusted during the actual rehabilitation process. The patient's subjective feelings and objective physiological responses, such as pain scores, serve as the basis for adjustment. If, in practice, it is found that some patients are overly sensitive or insensitive to load adjustments, further fine-tuning can be performed. Value, in order to ensure The rationality of the value requires clinical validation. Each patient's physiological characteristics are different, which may necessitate adjustments. Personalized adjustments can be made. For example, for young and healthy patients, The risk may be lower because their recovery is faster; for older patients or those with other health problems, A higher value may be needed to adjust the load more carefully.
[0196] By setting a baseline load, the problem of unreasonable load settings in traditional rehabilitation programs is solved. A suitable starting load is provided for each patient. This baseline load can be adjusted according to the patient's individual situation, surgical type, and postoperative rehabilitation status, avoiding the one-size-fits-all load setting of traditional methods. The baseline load provides a standard for subsequent dynamic adjustments, allowing patients to start with an appropriate load in the early stages of rehabilitation and gradually increase the training intensity. This setting avoids patients being unable to recover effectively due to too low a load, and also avoids secondary injuries that may be caused by too high a load. The dynamic load control model solves the problems of untimely and inaccurate load adjustments during rehabilitation. The dynamic load control model dynamically adjusts the load using time t and force. The coefficients in the model... As a coefficient relating load and force, the load can be precisely adjusted based on changes in force. This dynamic adjustment mechanism responds in real time to changes in the patient's physical condition, ensuring that the load remains at a reasonable level throughout the rehabilitation process. For example, when the patient's force is low, the load can be appropriately increased to promote faster recovery; conversely, when the patient's force is too high, the load will be reduced to prevent new injuries caused by overload. This refined, real-time load adjustment method significantly improves the safety and effectiveness of the rehabilitation process. By optimizing the relationship between load and force, the problem of overload or underload during rehabilitation is solved through the relationship coefficient. The new system allows for more flexible and personalized adjustments between load and force. Traditional rehabilitation methods often adjust load based on the patient's subjective feelings and fixed training plans, which may fail to reflect changes in force in real time, leading to excessively high or low loads. However, through precise control of the load-force relationship, the load can be adjusted based on real-time feedback from force data, ensuring that each training stage does not place excessive stress on the patient and avoiding insufficient load during recovery. This allows for better achievement of rehabilitation outcomes and avoids the problem of traditional rehabilitation programs failing to consider individual patient differences. Dynamic load adjustment enhances rehabilitation effectiveness and avoids imbalances in the patient's recovery process. Dynamic load adjustment not only helps reduce the risk of injury but also promotes smooth rehabilitation progress. Flexible load adjustment ensures that the training load required by each patient at different rehabilitation stages is reasonably controlled, resulting in a smoother rehabilitation process. Timely load adjustment allows patients to gradually improve their physical adaptability without overloading, helping them achieve faster recovery. It avoids overtraining or excessively restrictive training, maximizing rehabilitation effects.
[0197] The process of patient rehabilitation is modeled from multiple dimensions to further quantify and monitor rehabilitation effectiveness, specifically including:
[0198] Set up a rehabilitation progress set, including joint range of motion. Pain score and muscle strength Add to the recovery progress collection;
[0199] The patient's psychological state at time t is recorded as follows: ;
[0200] The multidimensional model of rehabilitation progress is as follows:
[0201] ;
[0202] in, The overall recovery progress at time t; , , and These are the weights for each progress dimension;
[0203] The weights of each progress dimension are determined using the following method:
[0204] Data feedback method: Calculate the improvement rate of various indicators using data from the initial stage of rehabilitation.
[0205] Calculate the relative improvement rate of each indicator:
[0206] ; ; ; ;
[0207] Then, .
[0208] By optimizing the objective function, we ensure that the recovery progress gradually approaches the target. The optimized objective function is as follows:
[0209] ;
[0210] in, Let be the value of the i-th progress indicator in the set of recovery progress at time t; Let n be the target value of the i-th progress indicator in the set of recovery progress indicators; n is the number of elements in the set of recovery progress indicators. Let be the weight of the i-th progress indicator in the set of recovery progress indicators.
[0211] The weights of the schedule indicators are determined using the following method:
[0212] Sensitivity method: For each index i, calculate the sensitivity in the initial stage:
[0213] ;in, The sensitivity of indicator i to the overall rehabilitation progress;
[0214] .
[0215] By establishing a rehabilitation progress ensemble, the lack of multi-dimensional and comprehensive assessment in traditional rehabilitation progress monitoring is addressed. By adding multiple key indicators such as joint range of motion, pain scores, and muscle strength to the ensemble, the patient's rehabilitation progress can be monitored in real time across multiple dimensions. Traditional methods typically focus on only a single dimension, such as pain score or joint range of motion, neglecting the patient's overall rehabilitation progress. The multi-dimensional ensemble not only assesses physiological recovery progress but also considers non-physiological factors such as psychological state, thus providing a comprehensive rehabilitation progress assessment. This method overcomes the limitation of single-dimensional rehabilitation assessment in traditional methods, ensuring all-round monitoring of the rehabilitation process. Optimizing the objective function through a multi-dimensional model solves the problems of unbalanced rehabilitation progress optimization and inconsistent progress objectives. During rehabilitation, the recovery of progress indicators such as joint range of motion and muscle strength is not always synchronous, and pain management and improvement in psychological state may also affect the speed of physiological recovery. By constructing a multi-dimensional rehabilitation progress model and optimizing the objective function, the recovery status of each dimension can be balanced, enabling the rehabilitation progress to be optimized towards multi-objective consistency. This method avoids over- or under-recovery in any single dimension, ensuring coordinated progress in joints, muscles, pain, and psychological aspects, thus achieving a comprehensive and balanced rehabilitation effect. By optimizing the objective function, it ensures that rehabilitation progress gradually approaches the goal, avoiding arbitrariness and inconsistency in rehabilitation plans. In traditional rehabilitation methods, progress assessment may only involve periodic checks, lacking a dynamic optimization mechanism. However, by setting an objective function to optimize rehabilitation progress, goals and strategies can be dynamically adjusted according to the patient's actual rehabilitation situation, ensuring that each progress indicator gradually approaches the set target value. This process makes the rehabilitation plan more scientific and systematic, capable of adaptive adjustments based on individual differences among patients. This continuous dynamic adjustment capability significantly improves the accuracy and efficiency of rehabilitation, reducing the problem of unsatisfactory results caused by progress deviations. By incorporating psychological state, it addresses the impact of patient psychological factors on rehabilitation outcomes. During the rehabilitation process, the patient's psychological state often affects the speed and effect of their physiological recovery, especially in the early stages of postoperative rehabilitation. By acquiring patient psychological state data and incorporating it into a multi-dimensional model, the overall rehabilitation situation of the patient can be more accurately reflected, and optimization adjustments can be made within the objective function. This strategy not only allows for the quantification and tracking of patients' physiological recovery, but also enables the effective assessment and management of the impact of their psychological state on the rehabilitation process. Consequently, effective measures can be taken to improve their psychological state and promote the recovery of physiological functions.
[0216] The method involves dynamically adjusting group members and tasks to enhance interaction and cooperation among patients, creating a comprehensive and supportive rehabilitation environment, specifically including:
[0217] Obtain all patients who need rehabilitation;
[0218] Based on the patient's rehabilitation progress and goals, patients are assigned to different rehabilitation groups, and the assignments are dynamically adjusted through cluster analysis and progress models, specifically as follows:
[0219] Calculate the distance between each patient's recovery progress and their goals:
[0220] Let the recovery progress of each patient at time t be... and the progress towards the ideal goal The distance between the two can be calculated using the following formula:
[0221] ;
[0222] in, This represents the j-th item of the recovery progress of the i-th patient at time t; Let j be the ideal value of the target rehabilitation progress term; n is the total number of target terms.
[0223] Based on the differences in recovery progress among patients, the K-means clustering algorithm was used to group patients. The goal of the K-means clustering algorithm is to minimize the differences in patient progress within a group. The clustering steps are as follows:
[0224] S1. Initialize K cluster centers, i.e., select K patients as initial representatives;
[0225] S2. For each patient, calculate and compare their distance to the center of each cluster, and assign the patient to the cluster with the closest distance.
[0226] S3. Update the center of each cluster, that is, calculate the average recovery progress of all patients in the cluster and use it as the new cluster center;
[0227] S4. Repeat steps S2 and S3 until the cluster center no longer changes;
[0228] When all patients have been traversed, stop the traversal and record each cluster obtained as a rehabilitation group.
[0229] A dynamic adjustment and optimization strategy is adopted to dynamically adjust the composition of each group and the allocation of tasks.
[0230] By dynamically grouping patients based on their rehabilitation progress and goals, this method addresses the problem of excessively large differences in patient progress. In orthopedic postoperative rehabilitation, patients exhibit significant variations in their recovery progress. Traditional one-size-fits-all rehabilitation group settings cannot meet the needs of diverse patients, potentially leading to slower progress or less than ideal recovery for some, thus affecting the overall treatment outcome of the group. The dynamic grouping method based on patient rehabilitation progress and goals first calculates the distance between each patient's rehabilitation progress and their ideal goal. Then, it uses the K-means clustering algorithm to divide patients into different groups according to their progress differences, ensuring that the difference in rehabilitation progress within each group is minimized. This method optimizes the patient grouping strategy, making patients in each rehabilitation group similar in their rehabilitation goals, avoiding the imbalance caused by excessive progress differences, and thus improving the overall rehabilitation effect. The K-means clustering algorithm is a current technology. Furthermore, the K-means clustering algorithm addresses the adverse effects of individual differences among patients on rehabilitation group interactions. Traditional rehabilitation groups are usually assigned based on the patient's condition or treatment stage, ignoring the differences in each patient's rehabilitation progress. By introducing the K-means clustering algorithm, dynamic grouping can be performed based on the actual rehabilitation progress and goal differences of patients, minimizing the progress differences within each rehabilitation group. This method optimizes the composition of the group members, ensuring that patients within the group have similar rehabilitation needs and goals, thereby promoting interaction and cooperation within the group. Patients can exchange experiences at similar rehabilitation stages, increasing their motivation and confidence, and further improving rehabilitation outcomes. By dynamically adjusting group members and tasks, the limitations of a single rehabilitation program and its lack of flexibility are overcome. As the rehabilitation process progresses, patients' rehabilitation progress changes. Traditional rehabilitation programs typically use static settings for group members and task allocation, lacking flexibility to adjust according to the actual progress of patients. However, through a dynamic adjustment strategy based on patients' rehabilitation progress, group members and task allocation can be adjusted in real time according to patients' rehabilitation effects, target progress, and grouping. Specifically, as patients' rehabilitation progress changes, they may be reassigned to different rehabilitation groups or their rehabilitation tasks may be adjusted. This dynamic optimization strategy ensures that each patient is rehabilitated in a group and with tasks suitable for them, improving the personalization and flexibility of the rehabilitation program. Dynamic task allocation improves the collaborative efficiency and effectiveness of the rehabilitation group. In the rehabilitation group, not only is interaction between patients crucial, but the rationality of task allocation also directly affects the rehabilitation outcome. By dynamically adjusting task allocation strategies based on progress differences, the collaboration and adaptability of each team member in completing tasks can be ensured. For example, the task intensity may be reduced for patients with slower progress, while the task intensity may be increased for patients with faster progress.In this way, patients can obtain the most suitable rehabilitation tasks according to their current abilities and needs, avoiding discomfort or slow progress caused by tasks that are too light or too heavy, thereby further improving the overall collaborative efficiency of the rehabilitation team.
[0231] The aforementioned strategy of dynamically adjusting and optimizing the composition and task allocation of each group specifically includes:
[0232] For any given rehabilitation group:
[0233] Set the adjustment interval time;
[0234] Set the initial time;
[0235] Starting from the initial time, after each adjustment interval, the group's rehabilitation progress, interaction score, and rehabilitation goals are obtained;
[0236] Set a threshold for differences in recovery progress;
[0237] Calculate the average rehabilitation progress of all patients in the rehabilitation group;
[0238] Calculate the difference between the rehabilitation progress of each member in the rehabilitation group and the average rehabilitation progress, and take the absolute value as the average difference in patient rehabilitation progress.
[0239] If the average difference in patient recovery progress among members exceeds the recovery progress difference threshold, then that member will be removed from the recovery group and a new member will be reassigned.
[0240] By calculating the differences in patients' recovery progress, the problem of excessive differences among members of the rehabilitation group was addressed. Within the rehabilitation group, there are significant individual differences in patients' recovery progress, especially due to variations in each patient's physical condition and recovery speed, leading to large disparities among group members. This not only affects patient recovery outcomes but may also impact cooperation and interaction within the group. By calculating the difference between each patient's recovery progress and the group's average progress, and removing patients with excessively large progress differences based on a preset threshold, members with significant differences can be effectively prevented from remaining in the same group. This ensures that the recovery progress of group members is relatively similar, thereby improving overall collaboration and progress consistency within the group.
[0241] This embodiment also provides a system for medical management based on artificial intelligence, including:
[0242] Data collection module: Collects data from patients' health records, postoperative assessments, and sensor devices;
[0243] Sensor module: Uses wearable sensors to record physiological data of the patient's postoperative body activity, gait and joint movement in real time;
[0244] Calculation module: Used to perform data calculations.
[0245] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0246] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A medical management method based on artificial intelligence, characterized by, The application relates to a method for setting a personalized rehabilitation plan for a patient, comprising the following steps: Collecting data from the patient's health record, postoperative evaluation and sensor equipment; The collected data includes authorized basic patient information, postoperative evaluation data and motion evaluation data; The authorized basic patient information includes age, gender, weight, height and activity level, wherein the activity level is divided into three levels: low, medium and high; The postoperative evaluation data includes the type of surgery, the date of surgery and surgical complications; The motion evaluation data includes joint range of motion A(t), pain score P(t), muscle strength M(t), stress on bones and soft tissues ; Using a wearable sensor to record the physiological data of the patient's postoperative physical activity, gait and joint movement in real time; According to the collected data, a set of personalized rehabilitation goals is set for the patient, and the set is denoted as R(t), wherein the set includes: Functional restoration goals, restoration of joint range of motion and gait restoration ; Muscle recovery goals, progressively increasing muscle strength ; Pain management goal, control pain score ; Mental health goal, to alleviate post-operative anxiety in patients ; Set the objective function: ; wherein, is the overall rehabilitation goal at time t, is the recovery of the i-th goal in the set R(t), is the weighting coefficient of each goal, and n is the total number of elements in the set R(t). Generating a preliminary rehabilitation plan based on the rehabilitation goals: Setting a set of rehabilitation training tasks, denoted as Q; The generation of each planned task is: ; wherein, is a rehabilitation training plan at time t; is the progress of the i-th training task in the set Q; is the weight coefficient of each task; is the total number of elements in the rehabilitation training task set; Setting a dynamic pain and function recovery balance strategy to provide real-time pain and function recovery feedback and ensure that the pain and recovery of the patient in the rehabilitation process are balanced; By monitoring the patient's bone and soft tissue in real time, data is provided for the adjustment of the rehabilitation plan; The rehabilitation process of the patient is modeled in multiple dimensions to further quantify and monitor the rehabilitation effect; By dynamically adjusting the team members and tasks, the interaction and cooperation among the patients are enhanced, and a comprehensive support rehabilitation environment is formed.
2. The medical management method based on artificial intelligence according to claim 1, characterized in that, The dynamic pain and function recovery balance strategy specifically includes: The function recovery progress at time t is denoted as F(t); A pain score P(t) and function recovery progress F(t) relationship model is established; The relationship between the pain score P(t) and the function recovery progress F(t) is represented by the following function: ; wherein is the intensity of the pain management measure in the treatment regimen, in particular: ; wherein, represents the total dose of analgesic received by the patient during the time ; ; represents the total length of time the patient received physical therapy during the time ; ; represents the number of times the patient received cold compress therapy during the time, in units of times; each time cold compress therapy is performed, it is directly counted as 1 unit; wherein, is the pain score at time t; and is the impact coefficient between pain and recovery progress; is the functional recovery progress at time t, calculated by the formula: ; wherein, ROMt is the range of motion of the joint at time t; ROMT is the target range of motion set during the rehabilitation process; MUSCt is the muscle strength at time t; MUSCT is the target muscle strength set during the rehabilitation process; WROM is the weighting factor for joint movement; WMUS is the weighting factor for muscle strength; According to the pain score and functional recovery progress , dynamically adjust the training plan. 3.The medical management method based on artificial intelligence according to claim 2, characterized in that, The dynamic adjustment of the training plan specifically includes: A pain score threshold is set, denoted as ; Setting the training intensity at time t, denoted as S(t); Setting the base training intensity, denoted as ; When the training intensity S(t) is reduced. The training intensity is adjusted according to the training intensity adjustment formula: ; wherein, is the correlation coefficient of the training intensity and the pain score. 4.The medical management method based on artificial intelligence according to claim 1, wherein, The real-time monitoring of the patient's bone and soft tissue provides data for the adjustment of the rehabilitation plan, specifically including: Monitoring force data of bones and soft tissues in a rehabilitation process through a worn sensor and recording force data in real time ; wherein is the position the force at time t; Obtaining the center point of the monitored bone, and establishing a three-dimensional coordinate system with the center point as the origin; Obtaining the total length of the monitored bone, denoted as L; Obtaining the total width of the monitored bone, denoted as W; Obtaining the total height of the monitored bone, denoted as H; Get location The three-dimensional coordinates of the location are denoted as ; wherein is the coordinate of the x-axis in the three-dimensional coordinate system; is the coordinate of the y-axis in the three-dimensional coordinate system; is the coordinate of the z-axis in the three-dimensional coordinate system; The force distribution is assumed to pass through the location and time t, the mathematical model is: ; wherein, is the total force value of the entire bone model at time t; is the local total force value at the point in the three-dimensional coordinate system at time t; According to the stress distribution of the bone and soft tissue, the load of the patient is dynamically adjusted.
5. The medical management method based on artificial intelligence according to claim 4, characterized in that, The dynamic adjustment of the patient's load specifically includes: Set the base load, denoted by ; The dynamic load control model is: ; wherein, is the adjusted load; is the relationship coefficient between the load and the force.
6. The medical management method based on artificial intelligence according to claim 1, wherein, The rehabilitation process of the patient is modeled in multiple dimensions to further quantify and monitor the rehabilitation effect, specifically including: A rehabilitation progress set is set up, adding joint range of motion , pain score and muscle strength to the rehabilitation progress set; acquiring a psychological state of the patient at a time t, denoted as ; The multi-dimensional model of the rehabilitation progress is: ; wherein, is the overall rehabilitation progress at time t; , , and are the weights for each progress dimension, respectively. By optimizing the objective function, the rehabilitation progress gradually approaches the target, and the optimization objective function is: ; wherein, is the value of the i-th progress indicator in the set of rehabilitation progress at time t; is the target value of the i-th progress indicator in the set of rehabilitation progress; n is the number of elements in the set of rehabilitation progress; is the weight of the i-th progress indicator in the set of rehabilitation progress.
7. The medical management method based on artificial intelligence according to claim 1, wherein, By dynamically adjusting the team members and tasks, the interaction and cooperation among the patients are enhanced, and a comprehensive support rehabilitation environment is formed, specifically including: Obtaining all the patients who need rehabilitation; Based on the rehabilitation progress and target of the patient, the patient is allocated to different rehabilitation groups, and the allocation is dynamically adjusted through clustering analysis and progress model, specifically including: Calculating the distance between the rehabilitation progress and the target of each patient: Let the recovery progress of each patient at time t be and the ideal target progress be The distance between the two is calculated by the following equation: ; wherein, is the jth item of rehabilitation progress for the ith patient at time t; is the ideal value for the jth item of target rehabilitation progress; n is the total number of target items; According to the difference in rehabilitation progress among patients, the patients are grouped by using a K-means clustering algorithm, and the objective of the K-means clustering algorithm is to minimize the difference in progress among patients in a group, and the clustering steps are as follows: S1, initializing K cluster centers, that is, selecting K patients as initial representatives; S2, for each patient, calculating the distance from each cluster center and comparing, and assigning the patient to the nearest cluster; S3, updating the center of each cluster, that is, calculating the average rehabilitation progress of all patients in the cluster as the new cluster center; S4, repeating steps S2 and S3 until the cluster center no longer changes; When all patients are traversed, stop traversing, and record each cluster obtained as a rehabilitation group; The dynamic adjustment and optimization strategy is used to dynamically adjust the member composition of each group. 8.The medical management method based on artificial intelligence according to claim 7, characterized in that, The dynamic adjustment and optimization strategy is used to dynamically adjust the member composition of each group and task allocation, specifically including: For any rehabilitation group: Set the adjustment interval time; Set the initial time; Starting from the initial time, every time an adjustment interval time is passed, the rehabilitation progress, interaction score and rehabilitation target in the group are obtained; Set the rehabilitation progress difference threshold; Calculate the average rehabilitation progress of all patients in the rehabilitation group; In turn, calculate the difference between the rehabilitation progress of the members in the rehabilitation group and the average rehabilitation progress, and take the absolute value, recorded as the average difference in patient rehabilitation progress; If there is a member whose average difference in patient rehabilitation progress is greater than the rehabilitation progress difference threshold, the corresponding member is removed from the rehabilitation group, and a new member is assigned.
9. A system employing the artificial intelligence-based medical management method of claim 1, characterized by, It includes: Data collection module: collect data from patient health records, postoperative evaluation and sensor devices; Sensor module: use wearable-based sensors to record physiological data of postoperative physical activity, gait and joint movement of patients in real time; Calculation module: for performing data calculation.
