Household remote monitoring system for stroke old people

Abstract

The invention relates to the technical field of medical rehabilitation, and discloses a household remote monitoring system for the aged with stroke, and the system comprises a health state collection module, a rehabilitation path optimization module, a feedback adjustment module and a remote monitoring module. The method comprises data acquisition, health state modeling, personalized rehabilitation training optimization, remote monitoring, real-time feedback adjustment, training intensity adaptive adjustment, dynamic training scheme optimization, rehabilitation progress evaluation and abnormity early warning. Physiological, motion and cognitive data are collected in real time through wearable equipment, a rehabilitation scheme is optimized in combination with an intelligent algorithm, the training mode, frequency and intensity are dynamically adjusted based on the state of a patient, and remote doctor monitoring, intelligent early warning and personalized intervention are achieved. Compared with a traditional rehabilitation mode, the precision, the real-time performance and the safety of rehabilitation are improved, sudden risks are reduced, the rehabilitation process is optimized, and the scientificity, the effectiveness and the long-term feasibility of home rehabilitation are improved.

Description

Technical Field

[0001] The present invention relates to the field of medical rehabilitation technology, and in particular to a home remote monitoring system for elderly people suffering from stroke. Background Art

[0002] Elderly stroke patients are those who have suffered a stroke due to cerebral vascular obstruction or rupture. These individuals often face sequelae such as motor dysfunction, cognitive decline, and speech difficulties. Because stroke not only affects patients' ability to live independently but is also associated with high recurrence rates and long-term health risks, rehabilitation therapy is crucial for improving their quality of life. However, the rehabilitation process is more than simple exercise training; it also requires physiological monitoring, cognitive recovery, and personalized rehabilitation plans to ensure safety and effectiveness.

[0003] Currently, rehabilitation for stroke patients typically requires long-term monitoring and professional intervention. However, existing rehabilitation models still have many shortcomings in practical application, resulting in rehabilitation outcomes often failing to meet expectations. Traditional rehabilitation monitoring methods rely on regular offline review sessions, with long intervals between data acquisitions, often failing to capture subtle changes in patients during the recovery process. This delayed data feedback not only affects doctors' judgment but also leads to delayed adjustments to rehabilitation plans, causing patients to miss the optimal time for intervention.

[0004] Furthermore, existing rehabilitation training models often rely on fixed plans developed by therapists with limited real-time adjustments. Each patient's recovery progress and physical tolerance vary, but traditional programs often struggle to dynamically optimize for individual differences. Training intensity can be too high, preventing patients from persisting, or too low, slowing recovery progress, prolonging the recovery period, and even impacting the ultimate outcome. This lack of precise, adaptive training strategies limits the scientific nature and efficiency of the rehabilitation process.

[0005] At the same time, the lack of remote monitoring capabilities is also a significant shortcoming. During home recovery, doctors struggle to obtain timely information about patients' health status, and family members lack effective monitoring methods. If patients experience unexpected issues during recovery, such as abnormal blood pressure, an unbalanced heart rate, or an unstable gait, they often lack a prompt response. This passive monitoring approach can lead to intervention only after an unexpected event occurs, significantly compromising the safety of rehabilitation.

[0006] Therefore, the present invention proposes a home remote monitoring system for elderly people with stroke to solve the shortcomings of the existing technology. Summary of the Invention

[0007] In response to the shortcomings of the existing technology, the present invention provides a home remote monitoring system for elderly people with stroke, which solves the problems of stroke patients' lack of real-time health monitoring during home rehabilitation, insufficient personalization of rehabilitation training, limited remote intervention by doctors, and difficulty in timely warning of sudden health events.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: a home remote monitoring system for elderly people with stroke, comprising:

[0009] The health status acquisition module is used to collect physiological, motor and cognitive data of stroke patients in real time through wearable devices, obtain real-time health data, and transmit the acquired real-time health data to the rehabilitation path optimization module;

[0010] The rehabilitation pathway optimization module models real-time health status data based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, generates a dynamically optimized rehabilitation pathway, and formulates the optimal training plan based on it, which is then passed to the feedback regulation module for adjustment;

[0011] The feedback adjustment module is used to automatically adjust the intensity and method of rehabilitation training for stroke patients based on the optimal training plan output by the rehabilitation path optimization module, and generate training results and adjustment results for feedback to the remote monitoring module;

[0012] The remote monitoring module is used to transmit the real-time health data, optimal training plan and training results of elderly people with stroke to professional doctors or rehabilitation therapists through the cloud platform for remote monitoring and real-time guidance, and provide necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module.

[0013] Preferably, the physiological data includes at least blood pressure, heart rate, and blood oxygen saturation; the motion data includes at least gait recovery and muscle activity; and the cognitive data includes at least cognitive recovery.

[0014] Preferably, the rehabilitation pathway includes a physiological recovery pathway, a motor function recovery pathway, and a cognitive ability recovery pathway;

[0015] The optimal training program includes personalized exercise training intensity, cognitive training intensity and physical therapy intensity, and is used to formulate a rehabilitation training plan for elderly people with stroke.

[0016] Preferably, the training results include changes in physiological indicators, improvement in motor ability, recovery of cognitive function and training compliance data;

[0017] The adjustment results include changes in training intensity, optimization of training methods, adjustment of training frequency and updates of personalized intervention measures.

[0018] Preferably, the cloud platform transmission includes data encryption processing, real-time uploading to the cloud server, storage and classification management, remote access authorization and data synchronization update, which are used to ensure the security and availability of rehabilitation data.

[0019] Preferably, the step of modeling the real-time health status data using the high-dimensional stochastic partial differential equation and the high-dimensional Hamilton-Jacobi-Bellman equation includes:

[0020] Obtain patients’ real-time health status data;

[0021] Define the health state variable X(t) as a multidimensional vector, representing the patient's health indicators at time t;

[0022] The evolution of health state is modeled using high-dimensional stochastic partial differential equations, which can be expressed as:

[0023] dX(t)=f(X(t),u(t),t)dt+σ(X(t),u(t),t)dW(t);

[0024] in: is the patient's health status variable; is the training control variable; f(X,u,t) is It is a state evolution function that describes the deterministic change trend of the patient's health status; σ(X,u,t) is It is a random perturbation coefficient matrix that describes the influence of uncertainty factors in the rehabilitation process; is a d-dimensional standard Wiener process, describing the random disturbances in the rehabilitation process; dt is the time increment, representing the change of the state variable in a small time interval; the initial condition X(0) = X0 gives the initial health state of the patient, and the terminal condition is set according to the rehabilitation goal;

[0025] Solve the optimal value function using the high-dimensional Hamilton-Jacobi-Bellman equation;

[0026] Based on real-time health status data and equation solution results, training plans and intervention strategies are dynamically adjusted to optimize the patient's recovery path.

[0027] Preferably, the high-dimensional Hamilton-Jacobi-Bellman equation is expressed as:

[0028]

[0029] Among them, V(X,t) is the optimal value function, which represents the optimal control value at a given time t and state X; is the partial derivative of the value function V with respect to time t, indicating the change of the optimal value over time; is the gradient of the value function V relative to the state variable X, which indicates the direction and rate of change of the value function in the current state; is the second-order derivative of the value function V with respect to the state X, which represents the change in the curvature of the value function; f(X,u,t) is the state evolution function, which represents the deterministic evolution of the system under the state X and control u; σ is the random perturbation coefficient matrix, which is used to describe the influence of random perturbations in the system on the state evolution; Tr(σσ T ) is the matrix σσ T The trace operation represents the second-order effect of random disturbance on the state; L(X,u,t) is the training cost function, which represents the cost or consumption of training under the current state X and control u; min u To minimize the control strategy u, it means to select the optimal control strategy.

[0030] Preferably, the health status acquisition module further includes a data storage unit for storing historical health status data of elderly people with stroke and providing a reference when adjusting the optimal training program.

[0031] Preferably, the remote monitoring module provides an intelligent alarm function, which automatically notifies the doctor or rehabilitation therapist when an abnormal health status of the patient is detected, and can send early warning information to the patient and his family.

[0032] The present invention also provides a home remote monitoring method for elderly people with stroke, comprising the following steps:

[0033] Health status data collection: Through the health status collection module, wearable devices are used to collect real-time physiological, motor and cognitive data of elderly people with stroke, and the acquired real-time health data is transmitted to the rehabilitation path optimization module;

[0034] Health status modeling and optimization: In the rehabilitation pathway optimization module, real-time health status data is modeled based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, and a dynamically optimized rehabilitation pathway is generated. Based on this pathway, a personalized optimal training plan is generated.

[0035] Training program adjustment: Through the feedback adjustment module, the intensity and method of rehabilitation training for stroke patients are automatically adjusted according to the optimal training program, and training results and adjustment results are generated;

[0036] Remote monitoring and guidance: Through the remote monitoring module, the real-time health data, optimal training plan and training results of the elderly with stroke are transmitted to the cloud platform for professional doctors or rehabilitation therapists to conduct remote monitoring and real-time guidance, and provide necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module;

[0037] Abnormal health status alarm: when the patient's health status is detected to be abnormal, the remote monitoring module will automatically notify the doctor or rehabilitation therapist and send early warning information to the patient and family members.

[0038] The present invention provides a home remote monitoring system for elderly people with stroke. It has the following beneficial effects:

[0039] 1. The present invention adopts the technical solution of wearable devices + remote data collection, which realizes the real-time collection and transmission of multi-dimensional data such as patients' physiology, movement, and cognition. In this way, doctors and rehabilitation therapists can grasp the patient's recovery status at any time. Compared with the traditional method that relies on regular examinations, it not only improves the timeliness and continuity of data, but also solves the problem of delayed monitoring and difficulty in timely detection of status changes during the patient's recovery process.

[0040] 2. The present invention uses high-dimensional stochastic partial differential equations + optimal control theory modeling to dynamically predict the patient's health status and intelligently optimize the rehabilitation training plan. In this way, each patient's training path is personalized and precisely matched. Compared with the existing technology that mainly relies on the experience of rehabilitation therapists to formulate fixed training plans, the present invention solves the shortcomings of rehabilitation plans being difficult to tailor to individuals and having poor adaptability, and improves the accuracy and effectiveness of rehabilitation training.

[0041] 3. The present invention adopts feedback regulation + adaptive adjustment strategy, which can automatically optimize the training intensity and method according to the patient's real-time physiological feedback (such as heart rate, gait, blood oxygen). If the patient recovers quickly, the system will appropriately increase the training difficulty; if the physical condition is weak, the training load will be automatically reduced to prevent the risks of overtraining. Traditional rehabilitation programs often have fixed training intensity and are difficult to adapt to the patient's daily state changes. The dynamic adjustment capability of the present invention truly achieves personalization, strong adaptability, and higher rehabilitation efficiency.

[0042] 4. The present invention relies on cloud data analysis + intelligent early warning mechanism, which enables doctors and family members to remotely view the patient's recovery progress and receive alerts in the first time when abnormalities occur (such as blood pressure surge, abnormal heart rate). Compared with the traditional method of relying on patients to detect abnormalities on their own and seek medical feedback, this intelligent monitoring greatly reduces the response time to sudden risks, and effectively solves the problems of difficulty in timely detection of sudden situations and lack of safety guarantees in the recovery process. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a system architecture diagram of the present invention;

[0044] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] See also Figure 1 The embodiment of the present invention provides a home remote monitoring system for elderly people with stroke, comprising:

[0047] The health status acquisition module is used to collect physiological, motor and cognitive data of stroke patients in real time through wearable devices, obtain real-time health data, and transmit the acquired real-time health data to the rehabilitation path optimization module;

[0048] The health status acquisition module primarily uses wearable devices to collect real-time physiological, motor, and cognitive data from stroke patients. This data is then transmitted to the subsequent rehabilitation pathway optimization module. This module plays a key role in the entire system. As the gateway to the entire remote monitoring system, it provides the necessary foundational data for subsequent data processing and rehabilitation pathway optimization. Through this module, the system can accurately and in real time capture the patient's health status, providing a basis for optimizing the rehabilitation pathway.

[0049] In this embodiment, the health status acquisition module is implemented in the following manner:

[0050] First, the health status collection module includes wearable devices that can monitor and collect various physiological, motor, and cognitive indicators for stroke patients. These devices can take various forms, such as wristbands, bracelets, smartwatches, and electrocardiogram devices, ensuring that patients can wear them easily in their daily lives and use them long-term. The design of wearable devices takes into account patient comfort, reliability, and data collection accuracy. For example, blood pressure, heart rate, and blood oxygen saturation can be measured using wristbands. Gait recovery can be monitored using sensors that capture the patient's steps, and muscle activity and other data can be monitored using accelerometers, gyroscopes, and other devices.

[0051] In a specific embodiment, the health status acquisition module obtains the following data in real time through these devices:

[0052] Physiological data: This includes blood pressure, heart rate, and blood oxygen saturation. This data helps assess the patient's baseline physiological state and monitor physiological changes during recovery. Analysis of this data can determine whether the patient is at risk of excessive fatigue or training overload.

[0053] Motion data: This primarily includes gait recovery and muscle activity. This data can be used to monitor the patient's motor recovery progress in real time, assess gait recovery, and determine whether rehabilitation goals have been achieved. Motion sensors, accelerometers, and other devices can capture subtle movement changes and perform precise calculations.

[0054] Cognitive data: This includes cognitive recovery. Cognitive recovery can be assessed by having patients complete a series of cognitive tasks, which can be performed using specific cognitive training devices or mobile apps. This cognitive data plays an important role in adjusting rehabilitation training programs.

[0055] In some embodiments, the design of the health status collection module ensures real-time and accurate data. Real-time health data is transmitted via wireless networks (such as Bluetooth and Wi-Fi) to subsequent modules of the system, particularly the rehabilitation pathway optimization module. Encryption technology is used during data transmission to ensure data security and privacy.

[0056] In one possible implementation, a health status acquisition module could continuously monitor the health status of stroke patients and automatically upload health data at regular intervals (e.g., every minute or every hour). This data would serve as input for subsequent rehabilitation pathway optimization, where it would be further analyzed and processed by the optimization module. In this way, the system could accurately capture changes in the patient's health and provide real-time information for dynamic adjustments to the rehabilitation pathway.

[0057] In this embodiment, the workflow of the health status acquisition module can be described as follows:

[0058] Wearable devices collect physiological, motor, and cognitive data.

[0059] The data is transmitted to the health status acquisition module via wireless transmission.

[0060] The module transmits data to the rehabilitation pathway optimization module for modeling and analysis.

[0061] In this way, the patient's health data can be monitored in real time throughout the rehabilitation process and provide support for the development of optimal rehabilitation training plans.

[0062] The health status acquisition module in this embodiment, leveraging advanced wearable devices, can efficiently and accurately acquire physiological, motor, and cognitive data from stroke survivors in real time. By transmitting this data to the rehabilitation pathway optimization module, the system can dynamically adjust rehabilitation training plans based on changes in the patient's health, thereby providing patients with personalized and precise rehabilitation plans. By modeling health status using high-dimensional stochastic partial differential equations, the system can reflect the patient's recovery progress in real time and provide theoretical support for further optimization.

[0063] The rehabilitation pathway optimization module models real-time health status data based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, generates a dynamically optimized rehabilitation pathway, and formulates the optimal training plan based on it, which is then passed to the feedback regulation module for adjustment;

[0064] The Rehabilitation Pathway Optimization module, at the core of the entire home remote monitoring system, receives real-time health data from the Health Status Acquisition Module and is responsible for generating a dynamically optimized rehabilitation path based on this data. By using high-dimensional stochastic partial differential equations (SDEs) and high-dimensional Hamilton-Jacobi-Bellman equations (HJBs), this module accurately models the health status of stroke patients, thereby developing a personalized optimal training plan. This plan relies on real-time collection of physiological, motor, and cognitive data, and the resulting optimal training plan is passed to the Feedback Adjustment Module for further adjustment.

[0065] In this embodiment, the rehabilitation path optimization module is implemented through the following steps:

[0066] First, the module models the patient's health status based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations. The patient's health status, X(t), is a multidimensional vector that includes various health indicators of the elderly stroke patient, such as physiological data, motor function, and cognitive ability. This model can describe changes in health status over time and account for external factors, such as adjustments to training intensity and the impact of uncertainty.

[0067] In the specific implementation, we first use a high-dimensional stochastic partial differential equation to model the evolution of health status based on real-time health status data and the patient's personalized situation. This equation can be expressed as:

[0068] dX(t)=f(X(t),u(t),t)dt+σ(X(t),u(t),t)dW(t);

[0069] in: is the patient's health status variable; is the training control variable; f(X,u,t) is It is a state evolution function that describes the deterministic change trend of the patient's health status; σ(X,u,t) is It is a random perturbation coefficient matrix that describes the influence of uncertainty factors in the rehabilitation process; is a d-dimensional standard Wiener process, which describes the random interference in the rehabilitation process; dt is the time increment, which represents the change of the state variable in a small time interval.

[0070] The implication of this equation is that by setting an appropriate control strategy u(t), the optimal evolution of the health state can be achieved under the background of random disturbances. Furthermore, by solving these equations, the system can obtain the optimal health state change at each moment.

[0071] The process of generating a rehabilitation pathway also involves the high-dimensional Hamilton-Jacobi-Bellman equation (HJB). In some embodiments, the HJB equation is used to solve the optimal value function, thereby providing guidance for subsequent training plans. The HJB equation can be expressed as:

[0072]

[0073] Among them, V(X,t) is the optimal value function, which represents the optimal control value at a given time t and state X; is the partial derivative of the value function V with respect to time t, indicating the change of the optimal value over time; is the gradient of the value function V relative to the state variable X, which indicates the direction and rate of change of the value function in the current state; is the second-order derivative of the value function V with respect to the state X, which represents the change in the curvature of the value function; f(X,u,t) is the state evolution function, which represents the deterministic evolution of the system under the state X and control u; σ is the random perturbation coefficient matrix, which is used to describe the influence of random perturbations in the system on the state evolution; Tr(σσ T ) is the matrix σσ T The trace operation represents the second-order effect of random disturbance on the state; L(X,u,t) is the training cost function, which represents the cost or consumption of training under the current state X and control u; min u To minimize the control strategy u, it means to select the optimal control strategy.

[0074] By solving the HJB equation, the system can generate the optimal rehabilitation path for each moment, including personalized training intensity and training methods. This path can further guide the patient's training plan, thereby improving rehabilitation efficiency.

[0075] Specifically, in this implementation, the process of generating a rehabilitation pathway involves the following steps:

[0076] Capture real-time patient health data, such as physiological, motor, and cognitive data.

[0077] High-dimensional stochastic partial differential equations are used to model the patient's health status, taking into account factors such as training control variables and random disturbances.

[0078] The high-dimensional Hamilton-Jacobi-Bellman equation is used to solve the optimal value function and generate the optimal rehabilitation path.

[0079] Based on this pathway, a personalized rehabilitation training plan is developed and passed to the subsequent feedback regulation module for adjustment.

[0080] In one possible implementation, the rehabilitation pathway optimization module iteratively updates the rehabilitation pathway at each point in time. In practice, this module will dynamically adjust the training plan based on the patient's real-time health data and continuously optimize the program throughout the patient's recovery. This dynamic optimization allows the system to effectively respond to changes in the patient's health during recovery and adjust the training plan in a timely manner to ensure maximum recovery.

[0081] The rehabilitation pathway optimization module in this embodiment models and optimizes the health status of stroke patients by introducing high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations. These equations accurately describe the evolution of a patient's health status and provide a theoretical basis for developing optimal training programs. This module can provide patients with personalized, precise rehabilitation pathways, ensuring an efficient and safe rehabilitation process.

[0082] The feedback adjustment module is used to automatically adjust the intensity and method of rehabilitation training for stroke patients based on the optimal training plan output by the rehabilitation path optimization module, and generate training results and adjustment results for feedback to the remote monitoring module;

[0083] The feedback adjustment module is primarily responsible for automatically adjusting the intensity and method of rehabilitation training for stroke patients based on the optimal training plan output by the rehabilitation pathway optimization module. Through this module, patients' rehabilitation training can be dynamically adjusted based on changes in real-time health data, thereby avoiding overtraining or undertraining. This module not only ensures the personalization and accuracy of rehabilitation training, but also generates training and adjustment results, which are fed back to the remote monitoring module for further guidance and modification suggestions from doctors or rehabilitation therapists.

[0084] In this embodiment, the feedback adjustment module is implemented in the following manner:

[0085] First, the Feedback Regulation Module receives the optimal training plan from the Rehabilitation Pathway Optimization Module. This plan is generated using high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations and is optimized based on the patient's real-time health status and rehabilitation goals. These plans typically include personalized training intensity, training methods, and training progress.

[0086] In general, the optimal training plan is dynamically adjusted based on the following factors:

[0087] Training Intensity: The system adjusts training intensity based on the patient's current physiological, motor, and cognitive state. If a patient's physiological indicators (such as blood pressure and heart rate) show signs of high load at a certain stage, the training intensity will be reduced. Conversely, if the patient's health is good, the training intensity may be gradually increased.

[0088] Training methods: Adjust training methods based on changes in the patient's health. For example, if a patient's gait recovery during exercise training is slow, the system may adjust the movements or form of training, adding more rehabilitative movements to help the patient recover.

[0089] Training frequency: The frequency of training may vary at different stages of the rehabilitation process. In some stages, patients may need more frequent rehabilitation training, while in other stages, the system may reduce the frequency of training to avoid excessive fatigue.

[0090] Optionally, the feedback adjustment module can further adjust the training program based on patient feedback data (such as subjective feelings and fatigue). Using real-time data and sensor data collected by the device worn by the patient, the system can infer the patient's level of fatigue and pain, and dynamically adjust the intensity and form of training accordingly.

[0091] In one possible implementation, the feedback regulation module employs a closed-loop control system. By monitoring the patient's health data in real time, the system can adjust the training plan based on actual conditions. For example, if a patient's physiological data (such as blood pressure) exceeds a set safety range during a certain period of time, the feedback regulation module will automatically reduce the training intensity and send an abnormality alert to the remote monitoring module as needed.

[0092] The core algorithm of the feedback regulation module may include regulation methods based on control theory. Through feedback regulation, the system can adjust the training plan in real time as the patient's health status changes, ensuring the optimal recovery path for the patient. The feedback regulation module will also make corresponding adjustments to different patients based on individual differences to ensure maximum recovery results.

[0093] In some embodiments, the feedback adjustment module can generate multiple training program alternatives so that when a sudden health issue arises, the user can quickly switch to the most appropriate program. For example, if a patient shows signs of cognitive decline, the feedback adjustment module can increase the intensity of cognitive training and reduce the load of other types of training.

[0094] Feedback on training results and adjustment results

[0095] The feedback adjustment module not only adjusts the training plan but also generates training results and adjustment results. Training results include changes in physiological indicators (such as blood pressure and heart rate), improvements in motor skills (such as gait recovery and muscle activity), cognitive recovery, and patient training compliance data. Using these results, the system can assess the patient's recovery progress and adjust the next training plan based on actual conditions.

[0096] Adjustment results include changes in training intensity, optimization of training methods, adjustment of training frequency, and updates to personalized intervention measures. These results are transmitted to doctors or rehabilitation therapists through the remote monitoring module, allowing them to provide further remote guidance based on the patient's actual recovery progress.

[0097] Collaborative work of feedback regulation module and remote monitoring module

[0098] Working in conjunction with the remote monitoring module, the feedback and adjustment module's real-time adjustments can help doctors or rehabilitation therapists make precise interventions. In some cases, doctors or rehabilitation therapists can also review the patient's rehabilitation data through the remote monitoring module and further modify the training plan based on the training adjustment suggestions provided by the feedback and adjustment module. This intelligent feedback and adjustment mechanism ensures a more personalized and optimized patient recovery path.

[0099] The feedback adjustment module in this embodiment automatically adjusts the patient's rehabilitation training plan, ensuring personalized and precise rehabilitation training. By monitoring and adjusting training intensity, method, and frequency in real time, this module helps patients avoid the risks of overtraining and undertraining. Working in conjunction with the remote monitoring module, the feedback adjustment module not only improves the effectiveness of rehabilitation training but also provides real-time data feedback and adjustment suggestions, further optimizing the effectiveness of the entire home remote monitoring system.

[0100] The remote monitoring module is used to transmit the real-time health data, optimal training plan and training results of stroke patients to professional doctors or rehabilitation therapists through the cloud platform for remote monitoring and real-time guidance. It also provides necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module.

[0101] The remote monitoring module works closely with the feedback adjustment module to achieve remote, real-time supervision of the home rehabilitation process of elderly people with stroke. This module is not only responsible for receiving the training adjustment results generated by the feedback adjustment module, but also integrates various physiological, motor, and cognitive data to ensure the personalization and safety of rehabilitation training. In addition, this module provides remote data support for doctors or rehabilitation therapists, allowing them to grasp the patient's rehabilitation status at any time and make necessary intervention adjustments. Through remote data transmission, abnormal warning mechanisms, and remote interaction functions, this module can improve the scientific nature of the rehabilitation process and effectively reduce potential risks during the rehabilitation process.

[0102] In this embodiment, the remote monitoring module is implemented in the following ways:

[0103] Typically, the core functions of a remote monitoring module include data collection, data analysis, abnormality warnings, and remote interaction. This module receives data from the patient's terminal device via wireless communication networks (such as Wi-Fi, 4G / 5G, or IoT protocols) and uploads the data to a remote server or cloud platform for real-time viewing and analysis by doctors or rehabilitation therapists.

[0104] Alternatively, the remote monitoring module can employ a distributed data storage architecture to reduce data transmission delays and ensure system stability and fault tolerance. This architecture allows the system to initially store some data locally and then upload it to the remote server in batches based on network conditions, improving data transmission efficiency.

[0105] Specifically, the core data processing flow of this module includes:

[0106] Data Reception and Storage: Receives real-time data on physiological indicators, motor parameters, and cognitive status from patients and stores it in a local cache or remote database. Data can be formatted in a time-series data structure to ensure accuracy and efficiency in subsequent analysis.

[0107] Data Analysis: Real-time processing of patients' health data, including but not limited to heart rate variability analysis (HRV), gait analysis, and cognitive reaction time calculation. Through regression modeling of historical data, this module can predict patients' possible future health trends.

[0108] Abnormality Warning: In some embodiments, the remote monitoring module can set abnormality alarm mechanisms based on dynamic thresholds. When the system detects abnormal health status of the patient (such as excessive heart rate, decreased blood oxygen, or abnormal exercise status), an alarm will be triggered immediately, and the doctor and the patient's family will be notified via mobile applications, text messages, or voice reminders.

[0109] Remote Interaction: This module allows doctors or rehabilitation therapists to interact with patients via a remote terminal (such as a PC or mobile device) and provide remote diagnosis and treatment recommendations. In one possible implementation, doctors can use this module to adjust rehabilitation training parameters or directly modify the training strategy of the feedback adjustment module, making rehabilitation plans more precise.

[0110] During data processing, the module uses time series prediction algorithms, such as a prediction model based on a long short-term memory network (LSTM), to analyze the patient's recovery trend. The basic mathematical expression of this prediction model is:

[0111] Hide status update:

[0112] h t =σ(W h ·h t-1 +W x ·X t +b h );

[0113] Output calculation:

[0114] y t =W y ·h t +b y ;

[0115] in, is the hidden state at time step t, representing the patient's health status feature vector with a dimension of d; The hidden state at time step t-1; The input data at time step t contains the patient's physiological, motor and cognitive parameters, with a dimension of n; is the weight matrix from hidden state to hidden state; is the weight matrix from input data to hidden state; is the bias vector of the hidden state; is the weight matrix from hidden state to output layer, m is the output dimension; is the bias vector of the output layer; σ is a nonlinear activation function (such as sigmoid or tanh), which is used to introduce nonlinear characteristics.

[0116] The introduction of this prediction model enables the remote monitoring module to predict possible health abnormalities in advance and adjust the training plan when necessary to reduce the health risks of patients caused by mismatched training intensity.

[0117] In one possible implementation, the remote monitoring module supports multi-user remote management and provides personalized data visualization. Doctors or rehabilitation therapists can use interactive data panels to view the recovery progress of different patients and compare data patterns of similar cases to develop more appropriate rehabilitation plans. Such data panels can optimize data visualization using the following mathematical model:

[0118]

[0119] in: is the comprehensive health score, which represents the numerical evaluation of the patient's current overall recovery status; n is the number of health indicators, and different recovery stages may contain different numbers of indicators; is the weight of the i-th health indicator, and the sum of the weight coefficients is equal to 1 to ensure that the importance of different indicators is reasonably distributed; is the scoring function for the i-th health indicator, representing the specific value of the indicator. The scoring function can use different calculation methods (such as normalization processing, standardized scoring, etc.). X is the patient's health status vector, including multiple health-related variables, such as physiological, motor, and cognitive indicators.

[0120] The synergy between the remote monitoring module and the feedback adjustment module enables further optimization of rehabilitation training programs through remote intervention by doctors. In some embodiments, doctors can directly adjust training parameters remotely and feed the adjusted program back to the feedback adjustment module, which then performs real-time optimization. Furthermore, this module can also work in conjunction with the rehabilitation pathway optimization module to analyze remote monitoring data and adjust the rehabilitation pathway optimization strategy to ensure the dynamic adaptability of the rehabilitation program.

[0121] In some cases, the module can also be connected to external health management systems, such as hospital information systems (HIS) or electronic medical records (EMR) systems, to provide a more complete health management solution. In this way, patients' recovery data can not only be used for real-time monitoring, but also form a complete health record for doctors to conduct long-term follow-up analysis.

[0122] The remote monitoring module in this embodiment realizes real-time remote supervision of the rehabilitation process of elderly people with stroke through data reception, analysis, abnormal warning and remote interaction. Through the time series prediction model, abnormal alarm mechanism and remote interaction function, this module enables the rehabilitation plan to be dynamically adjusted according to the patient's real-time health status, thereby improving the accuracy and safety of rehabilitation. In addition, the module also has data visualization and multi-user remote management functions, ensuring that doctors or rehabilitation therapists can efficiently provide remote guidance on the patient's rehabilitation process. Through the synergistic effect with the feedback adjustment module and the rehabilitation path optimization module, this module further enhances the intelligence level of the present invention, enabling the home rehabilitation monitoring system to provide more comprehensive and accurate rehabilitation support.

[0123] See also Figure 2 The present invention also provides a home remote monitoring method for elderly people with stroke, comprising the following steps:

[0124] S1. Health status data collection: Through the health status collection module, wearable devices are used to collect real-time physiological, motor and cognitive data of stroke patients, and the acquired real-time health data is transmitted to the rehabilitation path optimization module;

[0125] S2, health status modeling and optimization. In the rehabilitation pathway optimization module, real-time health status data is modeled based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, and a dynamically optimized rehabilitation pathway is generated. Based on this pathway, a personalized optimal training plan is generated.

[0126] S3, training program adjustment, through the feedback adjustment module, automatically adjusts the intensity and method of rehabilitation training for stroke elderly people according to the optimal training program, and generates training results and adjustment results;

[0127] S4. Remote monitoring and guidance: Through the remote monitoring module, the real-time health data, optimal training plan and training results of the stroke elderly are transmitted to the cloud platform for professional doctors or rehabilitation therapists to conduct remote monitoring and real-time guidance, and provide necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module;

[0128] S5. Abnormal health status alarm. When the patient's health status is detected to be abnormal, the remote monitoring module will automatically notify the doctor or rehabilitation therapist and send an early warning message to the patient and his family.

[0129] In general, for step S1, the method first uses a health status acquisition module to collect physiological, motor and cognitive status data of the elderly with stroke in real time using wearable devices, smart sensors or other monitoring devices.

[0130] As an option, wearable devices may include but are not limited to heart rate monitors, blood oxygen monitors, gait analysis sensors, electromyography sensors, electroencephalogram (EEG) devices, body temperature monitors, etc. The types of data collected by different devices include:

[0131] Physiological data: blood pressure, blood oxygen saturation, heart rate, heart rate variability (HRV), body temperature, etc.

[0132] Motion data: gait stability, step length, gait speed, range of motion (ROM), muscle electrical signals (EMG), etc.

[0133] Cognitive data: reaction time, attention span, memory test results, emotional state (e.g., based on EEG signal analysis), etc.

[0134] In some embodiments, data from the wearable device can be transmitted to a data processing terminal via Bluetooth Low Energy (BLE), Wi-Fi, 5G, or Internet of Things (IoT) communication protocols and stored in a local or remote server for subsequent analysis.

[0135] In one possible implementation, the frequency of health data collection can be dynamically adjusted based on the patient's daily activity status. For example:

[0136] In rest state, the frequency of data collection is reduced to save equipment energy consumption;

[0137] During exercise or rehabilitation training, increase the frequency of data collection to ensure high-precision monitoring.

[0138] In step S2, after collecting health status data, the rehabilitation pathway optimization module analyzes the patient's health data and constructs a high-dimensional mathematical model of health status evolution. This model combines physiological, motor, and cognitive status data, using high-dimensional stochastic partial differential equations to describe the dynamic changes in health status. The model then calculates the optimal rehabilitation pathway using the Hamilton-Jacobi-Bellman equation (HJB), ensuring the scientific and personalized nature of the rehabilitation program.

[0139] Health status modeling

[0140] Generally speaking, a patient's health status is not only affected by their own recovery, but also by the effects of rehabilitation training. Therefore, this module considers the health status as a multidimensional dynamic system, and considers the following when modeling:

[0141] Physiological data (such as heart rate, blood pressure, and blood oxygen saturation) to assess the patient's basic health status;

[0142] Movement data (such as gait, stride length, and range of motion) to assess the patient's motor function recovery;

[0143] Cognitive data (e.g., reaction time, attention level) to measure neurocognitive function.

[0144] To accurately predict rehabilitation progress, the module uses stochastic partial differential equations to describe changes in health status over time and consider the impact of external rehabilitation training. This model can dynamically adjust training plans to adapt to the patient's real-time health status and avoid over- or undertraining.

[0145] Solving the optimal rehabilitation path

[0146] After the model is built, the system uses the HJB equation to solve the optimal rehabilitation path. This equation calculates the optimal training intensity and method while ensuring the rehabilitation effect, and generates a personalized training plan based on the individual characteristics of the patient. The optimization process considers:

[0147] Maximizing rehabilitation benefits, that is, how to improve the patient's health status in the shortest possible time;

[0148] Minimize training risks and ensure that training does not impose additional burden on patients;

[0149] Patient adaptability, dynamically adjusting training intensity and frequency based on health data.

[0150] In some embodiments, the system continuously optimizes the training program based on the patient's real-time data and rehabilitation progress. For example:

[0151] When it is detected that the patient's gait has recovered well, the training intensity can be appropriately increased;

[0152] If your heart rate or blood pressure is outside a safe range, reduce your training load to avoid excessive fatigue.

[0153] Personalized training plan generation

[0154] Based on the optimization results, the rehabilitation pathway optimization module generates a personalized training plan, including:

[0155] Exercise training (such as walking training and balance training) to improve muscle strength and balance;

[0156] Cognitive training (such as memory training and reaction ability training) to improve neurocognitive function;

[0157] Auxiliary training (such as electrical stimulation rehabilitation and VR rehabilitation) can enhance the training effect.

[0158] With the cooperation of the remote monitoring module, the solution can be dynamically adjusted according to the patient's real-time feedback, making the rehabilitation process more accurate and efficient.

[0159] Summarize

[0160] This module provides personalized optimal training plans through health status modeling and optimization, and continuously adjusts the rehabilitation path based on the patient's real-time data to ensure the scientificity, safety and effectiveness of rehabilitation training.

[0161] The optimal training plan generated by the rehabilitation path optimization module in step S3 will be transmitted to the feedback adjustment module, which adjusts the training intensity, method and frequency according to the patient's real-time health status.

[0162] As an option, the feedback regulation module adopts a closed-loop control strategy, combining the patient's physiological feedback data to dynamically adjust the training parameters. For example:

[0163] If abnormalities in the patient's heart rate or blood oxygen are detected, the training intensity will be automatically reduced;

[0164] If the patient shows good gait stability during exercise, the training difficulty should be appropriately increased.

[0165] In some embodiments, the training regimen adjustment may adopt a fuzzy control algorithm, whose rules are as follows:

[0166] IF (heart rate increase after training > threshold) THEN (reduce training intensity);

[0167] IF (gait stability score > set value) THEN (increase training difficulty).

[0168] The results of training and adjustment will be recorded and fed back to the remote monitoring module.

[0169] In step S4, the remote monitoring module uploads the patient's real-time health data, optimal training plan and training adjustment results to the cloud server through the wireless communication network, and provides remote access to doctors or rehabilitation therapists.

[0170] In one possible implementation, the module may support:

[0171] Abnormal detection: When the patient's physiological data exceeds the safety threshold, the system automatically triggers an alarm;

[0172] Remote interaction: Doctors can adjust training plans or provide remote guidance through smart terminals;

[0173] Health trend prediction: Using deep learning algorithms such as LSTM (Long Short-Term Memory Network), we can make time series predictions of patients’ health status and identify possible recovery problems in advance.

[0174] For step S5, when the remote monitoring module detects that the patient's health status is abnormal (such as high heart rate, decreased blood oxygen, unstable gait, etc.), the system will automatically trigger the alarm mechanism and notify the doctor, rehabilitation therapist and patient's family.

[0175] Typically, the alarm mechanism includes:

[0176] Instant alarm: If health data exceeds the set safety range, the system will send text messages, phone calls or APP notifications to doctors and patients' families;

[0177] Trend alarm: If the system predicts that the patient's health condition may deteriorate in the future, it will issue an early warning.

[0178] As an option, health status abnormality alarms can be combined with Bayesian network modeling to improve the accuracy of alarms and reduce false alarms.

[0179] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A home remote monitoring system for elderly people with stroke, characterized by: include: The health status acquisition module is used to collect physiological, motor and cognitive data of stroke patients in real time through wearable devices, obtain real-time health data, and transmit the acquired real-time health data to the rehabilitation path optimization module; The rehabilitation pathway optimization module models real-time health status data based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, generates a dynamically optimized rehabilitation pathway, and formulates the optimal training plan based on it, which is then passed to the feedback regulation module for adjustment; The feedback adjustment module is used to automatically adjust the intensity and method of rehabilitation training for stroke patients based on the optimal training plan output by the rehabilitation path optimization module, and generate training results and adjustment results for feedback to the remote monitoring module; The remote monitoring module is used to transmit the real-time health data, optimal training plan and training results of elderly people with stroke to professional doctors or rehabilitation therapists through the cloud platform for remote monitoring and real-time guidance, and provide necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module.

2. A home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The physiological data includes at least blood pressure, heart rate, and blood oxygen saturation; the motion data includes at least gait recovery and muscle activity; and the cognitive data includes at least cognitive recovery.

3. A home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The rehabilitation pathway includes a physiological recovery pathway, a motor function recovery pathway, and a cognitive ability recovery pathway; The optimal training program includes personalized exercise training intensity, cognitive training intensity and physical therapy intensity, and is used to formulate a rehabilitation training plan for elderly people with stroke.

4. A home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The training results include changes in physiological indicators, improvement in athletic ability, recovery of cognitive function, and training compliance data; The adjustment results include changes in training intensity, optimization of training methods, adjustment of training frequency and updates of personalized intervention measures.

5. The home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The cloud platform transmission includes data encryption processing, real-time uploading to the cloud server, storage and classification management, remote access authorization and data synchronization update, which is used for the security and availability of rehabilitation data.

6. The home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The steps of modeling the real-time health status data using the high-dimensional stochastic partial differential equation and the high-dimensional Hamilton-Jacobi-Bellman equation include: Obtain patients’ real-time health status data; Define the health state variable X(t) as a multidimensional vector, representing the patient's health indicators at time t; The evolution of health state is modeled using high-dimensional stochastic partial differential equations, which can be expressed as: dX(t)=f(X(t),u(t),t)dt+σ(X(t),u(t),t)dW(t); in: is the patient's health status variable; is the training control variable; f(X,u,t) is It is a state evolution function that describes the deterministic change trend of the patient's health status; σ(X,u,t) is It is a random perturbation coefficient matrix that describes the influence of uncertainty factors in the rehabilitation process; is a d-dimensional standard Wiener process, describing the random disturbances in the rehabilitation process; dt is the time increment, representing the change of the state variable in a small time interval; the initial condition X(0) = X0 gives the initial health state of the patient, and the terminal condition is set according to the rehabilitation goal; Solve the optimal value function using the high-dimensional Hamilton-Jacobi-Bellman equation; Based on real-time health status data and equation solution results, training plans and intervention strategies are dynamically adjusted to optimize the patient's recovery path.

7. A home remote monitoring system for elderly people with stroke according to claim 6, characterized in that: The high-dimensional Hamilton-Jacobi-Bellman equation is expressed as: Among them, V(X,t) is the optimal value function, which represents the optimal control value at a given time t and state X; is the partial derivative of the value function V with respect to time t, indicating the change of the optimal value over time; is the gradient of the value function V relative to the state variable X, which indicates the direction and rate of change of the value function in the current state; is the second-order derivative of the value function V with respect to the state X, which represents the change in the curvature of the value function; f(X,u,t) is the state evolution function, which represents the deterministic evolution of the system under the state X and control u; σ is the random perturbation coefficient matrix, which is used to describe the influence of random perturbations in the system on the state evolution; Tr(σσ T ) is the matrix σσ T The trace operation represents the second-order effect of random disturbance on the state; L(X,u,t) is the training cost function, which represents the cost or consumption of training under the current state X and control u; min u To minimize the control strategy u, it means to select the optimal control strategy.

8. The home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The health status acquisition module further includes a data storage unit for storing historical health status data of the elderly with stroke and providing a reference when adjusting the optimal training plan.

9. The home remote monitoring system for elderly people with stroke according to claim 1, characterized in that: The remote monitoring module provides an intelligent alarm function. When an abnormal health status of a patient is detected, it automatically notifies the doctor or rehabilitation therapist and sends early warning information to the patient and his family.

10. A home remote monitoring method for elderly people with stroke, applied to a home remote monitoring system for elderly people with stroke as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: Health status data collection: Through the health status collection module, wearable devices are used to collect real-time physiological, motor and cognitive data of elderly people with stroke, and the acquired real-time health data is transmitted to the rehabilitation path optimization module; Health status modeling and optimization: In the rehabilitation pathway optimization module, real-time health status data is modeled based on high-dimensional stochastic partial differential equations and high-dimensional Hamilton-Jacobi-Bellman equations, and a dynamically optimized rehabilitation pathway is generated. Based on this pathway, a personalized optimal training plan is generated. Training program adjustment: Through the feedback adjustment module, the intensity and method of rehabilitation training for stroke patients are automatically adjusted according to the optimal training program, and training results and adjustment results are generated; Remote monitoring and guidance: Through the remote monitoring module, the real-time health data, optimal training plan and training results of the elderly with stroke are transmitted to the cloud platform for professional doctors or rehabilitation therapists to conduct remote monitoring and real-time guidance, and provide necessary guidance and modification suggestions based on the adjustment results of the feedback adjustment module; Abnormal health status alarm: when the patient's health status is detected to be abnormal, the remote monitoring module will automatically notify the doctor or rehabilitation therapist and send early warning information to the patient and family members.

Images