AI-based intelligent guidance system for rehabilitation training of burn patients

Abstract

The present invention relates to the technical field of rehabilitation training for burn patients and discloses an AI-based intelligent guidance system for rehabilitation training for burn patients. The system includes data acquisition, rehabilitation assessment, training planning, intelligent guidance, and effect feedback modules. The data acquisition module collects physiological characteristics and training movement data and converts them into different formats; the rehabilitation assessment module constructs a model to assess burn recovery and functional impairment; the training planning module generates a rehabilitation training plan based on the assessment results; the intelligent guidance module optimizes plan parameters; and the effect feedback module analyzes training data and updates model parameters. This system enables intelligent and precise management of rehabilitation training for burn patients, improving rehabilitation training effectiveness.

Description

Technical Field

[0001] The present invention relates to the technical field of rehabilitation training for burn patients, and in particular to an AI-based intelligent guidance system for rehabilitation training for burn patients. Background Art

[0002] Burns are a common and devastating trauma. Patients not only suffer physical pain but also face a long and complex recovery process. Traditional rehabilitation training for burn patients relies primarily on the experience and manual operation of medical staff. Manual measurement and recording are often used to collect rehabilitation training data. This makes it difficult to achieve real-time, continuous monitoring of patient physiological characteristics, such as skin temperature, heart rate, and blood oxygen levels, and errors are prone to occur during the measurement process. Training movement data cannot accurately capture the details and changes in the patient's movements, resulting in incomplete and inaccurate data.

[0003] During rehabilitation assessments, burn injury assessments and functional impairment analyses are largely based on the physician's subjective judgment and simple examination methods. Burn depth identification relies on visual observation of the wound surface, making it difficult to accurately distinguish the extent of damage to the epidermis, dermis, and subcutaneous tissue. Wound healing progress assessments lack quantitative indicators, making it difficult to accurately grasp the specific progress of healing. The lack of a scientific and systematic assessment model for assessing patient functional impairments, such as range of motion, muscle strength, and coordination, makes it difficult to fully understand the patient's physical function, thus hindering the development of rehabilitation training programs.

[0004] Rehabilitation training programs lack personalization and precision. Traditional methods make it difficult to tailor training programs to individual patient differences, such as burn severity, physical condition, and recovery speed. The selection of training exercises, intensity, and frequency often follow a uniform standard, failing to meet the patient's actual needs. This can lead to poor training results and even delay recovery.

[0005] During rehabilitation training, patients lack effective guidance and feedback. Medical staff are unable to monitor patients' training movements in real time, failing to identify and correct incorrect movements promptly. This makes it difficult for patients to ensure standard and consistent training. Furthermore, feedback on patients' training effectiveness is not timely, preventing doctors from adjusting training plans based on results, making it difficult to ensure the efficiency and quality of rehabilitation training. With the development of artificial intelligence technology, how to apply it to the rehabilitation training of burn patients and address the challenges of traditional rehabilitation training has become a pressing research topic. Summary of the Invention

[0006] The purpose of the present invention is to provide an AI-based intelligent guidance system for rehabilitation training of burn patients to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, the present invention provides an AI-based intelligent guidance system for rehabilitation training of burn patients, which includes a data acquisition module, a rehabilitation assessment module, a training planning module, an intelligent guidance module, and an effect feedback module;

[0008] The data acquisition module is used to set monitoring sites in the burn patient rehabilitation training scene and deploy sensor devices to collect the patient's physiological characteristics data and training movement data in real time, and convert the collected data into a format;

[0009] The rehabilitation assessment module is used to construct a burn injury assessment model and a functional impairment analysis model, and uses machine learning technology to analyze the patient's physiological characteristic data. At the same time, it uses motion analysis technology to identify the standard degree of training movement data. Based on the real-time collected physiological characteristic data and training movement data, it comprehensively evaluates the patient's burn recovery status and functional impairment level.

[0010] The training planning module is used to align the assessed burn recovery status and functional impairment level, perform correlation calculation to obtain a training parameter set, preset a training target value, and perform preliminary matching to generate a rehabilitation training plan for the patient at the current stage;

[0011] After generating a rehabilitation training program, the intelligent guidance module is used to further calculate and obtain an adjustment parameter set based on the individual characteristics of the patient, and to perform a secondary match between the preset adjustment baseline value and the adjustment parameter set to further optimize the specific values ​​of the intensity, frequency and duration of the movement in the rehabilitation training program;

[0012] The effect feedback module is used to perform time-series comparative analysis on the physiological characteristic data and training movement data collected during the execution of the optimized rehabilitation training program to obtain training effect indicators, and dynamically update the parameter configuration of the burn injury assessment model and the functional impairment analysis model based on the training effect indicators.

[0013] Preferably, the data acquisition module includes a physiological signal acquisition unit, a motion trajectory acquisition unit and a data conversion unit;

[0014] The physiological signal acquisition unit includes a skin condition acquisition unit and a vital sign acquisition unit, which is used to monitor and collect the patient's physiological characteristic data in real time by deploying a biosensor group at the patient's burn site and key nodes on the body, and transmit the data to the data conversion unit via wireless transmission. The biosensor group includes a skin temperature sensor group and a heart rate and blood oxygen sensor group. The physiological characteristic data includes skin temperature values ​​and heart rate and blood oxygen values;

[0015] The skin condition acquisition unit is used to monitor the skin temperature data of the patient's burn area in real time based on the skin temperature sensor group. The skin temperature sensor group includes a contact probe, a signal amplifier and a data recorder, which respectively collect the temperature value, fluctuation range and sampling interval of the skin temperature data;

[0016] The vital sign acquisition unit is used to collect the patient's heart rate and blood oxygen data in real time based on the heart rate and blood oxygen sensor group. The heart rate and blood oxygen sensor group includes a photoelectric detector, a filter circuit and a digital-to-analog converter. The heart rate and blood oxygen data include heart rate frequency, blood oxygen saturation, measurement position and signal stability;

[0017] The motion trajectory acquisition unit is used to establish a communication protocol to connect with the control terminal of the motion capture device, read the acquisition parameters of the motion capture device in the control terminal in real time, and extract and summarize the sampling rate, spatial accuracy and angular resolution of the acquisition parameters of the motion capture device in real time to obtain the patient's training motion data.

[0018] Preferably, the data conversion unit is used to filter out interference signals and abnormal values ​​from the collected physiological characteristic data and training movement data, and unify the data formats of different types of sensors. At the same time, the collected physiological characteristic data and training movement data are timestamp-aligned through time synchronization processing to obtain the skin temperature value, heart rate frequency and joint movement angle at the same time point.

[0019] Preferably, the rehabilitation assessment module includes an injury degree assessment unit, a functional impairment analysis unit and a model calibration unit;

[0020] The injury degree assessment unit includes a burn depth identification unit and a wound healing analysis unit;

[0021] The burn depth identification unit extracts the patient's burn area data and wound image information from the electronic medical record system, uses image processing software to establish a three-dimensional texture model of the burn site, identifies epidermal damage, dermal damage, and subcutaneous tissue damage in the wound, and simultaneously annotates typical features of the burn site, including the redness and swelling area, blister distribution, and eschar range. After the initial assessment, a classification algorithm is used to define the thermal conductivity, elastic modulus, and regenerative biological properties of the skin tissue in the constructed three-dimensional texture model. Simultaneously, the physiological signal acquisition cycle, analysis window, and threshold range are set to perform static, dynamic, and continuous assessments to assess the severity of the patient's burn injury.

[0022] The wound healing analysis unit is used to input wound repair parameters, including new epidermal thickness, capillary density, and inflammatory factor levels, and then perform feature extraction and analysis to assess the healing progress of the patient's burn site at different recovery stages;

[0023] The functional impairment analysis unit is used to establish a movement function assessment model, including joint range of motion, muscle strength and coordination indicators, and then apply kinematic equations to analyze the patient's training movement data. The motion analysis technology is used to analyze the patient's functional impairment type and evaluate the degree of joint movement restriction, muscle contraction fatigue and coordination level of movement execution;

[0024] The model calibration unit is used to import the patient's actual recovery data into the burn injury assessment model and the functional disorder analysis model to perform parameter calibration to obtain an updated model, and then collect the patient's physiological characteristics data and training movement data in real time, transmit them to the feature extraction analysis and motion analysis technology for dynamic assessment, and import the dynamic assessment results into the updated model, adjust the weight coefficients of the burn injury assessment model and the functional disorder analysis model in real time, and display the assessment results through the data dashboard to provide doctors with parameter adjustment functions.

[0025] Preferably, the training planning module includes a stage goal setting unit, an action parameter matching unit and a plan generating unit;

[0026] The stage goal setting unit is used to construct a training goal setting algorithm, calculate and obtain the training intensity target of the patient's current rehabilitation stage based on the calibrated assessment results, and set the range of motion and muscle strength improvement indicators of the burned area;

[0027] The motion parameter matching unit is used to construct a motion parameter matching algorithm, and calculate the amplitude, speed and number of repetitions of the training movement based on the calibrated evaluation results to match the rehabilitation training movement type suitable for the patient at the current stage;

[0028] The program generation unit is used to construct a training program generation algorithm, calculate and obtain a rehabilitation training program based on the training intensity target and training action parameters, and generate training guidance content including action names, execution steps and precautions.

[0029] Preferably, the stage goal setting unit is used to calculate and obtain the training intensity target by analyzing the severity of the patient's burn injury and combining it with the calibrated wound healing progress, and set the maximum motion angle of the patient's burned joints and the minimum contraction force of the muscles.

[0030] Preferably, the motion parameter matching unit is used to calculate and obtain training motion parameters by comparing the types of functional impairments in different recovery stages and combining the calibrated motor function assessment results to match the patient's tolerable motion amplitude and execution speed within a safe range at the current stage.

[0031] Preferably, the intelligent guidance module includes a real-time action correction unit and a training prompt unit;

[0032] The real-time motion correction unit is used to compare the motion parameters in the optimized rehabilitation training program with the patient's real-time training motion data in space and align the time series, and then perform correlation calculation to obtain the motion deviation value and perform real-time analysis on the standard degree of the patient's training motion;

[0033] The training prompt unit is used to preset prompt rules based on the cognitive ability and communication habits of the burn patient, and perform secondary matching with the obtained movement deviation value to prompt the patient to adjust the training movement. The specific prompt scheme is as follows: when the movement deviation value is less than the prompt threshold, it means that the patient's training movement meets the requirements of the scheme. At this time, a regular prompt message is generated to encourage the patient to continue the current training; when the movement deviation value is greater than or equal to the prompt threshold, it means that there is a deviation in the patient's training movement. At this time, a correction prompt message is generated to guide the patient to adjust the movement amplitude or speed.

[0034] Preferably, the effect feedback module includes a training data storage unit and a rehabilitation progress analysis unit;

[0035] The training data storage unit is used to further organize the patient's physiological characteristic data and training movement data of each training session in combination with the obtained training effect indicators, and classify and store them in the rehabilitation training database;

[0036] The rehabilitation progress analysis unit is used to preset progress evaluation standards and the obtained training effect indicators, conduct time-series comparative analysis, further analyze the changing trends of the patient's burn recovery status and functional impairment level during the continuous training cycle, and generate corresponding progress reports. The specific analysis plan is as follows; when the training effect indicator improvement rate is greater than the progress benchmark value, it means that the patient's rehabilitation progress is in line with expectations. At this time, a normal progress report is generated to prompt the doctor to maintain the current training plan; when the training effect indicator improvement rate is equal to the progress benchmark value, it means that the patient's rehabilitation progress is basically up to standard. At this time, a focus progress report is generated to prompt the doctor to pay attention to the subsequent training effects; when the training effect indicator improvement rate is less than the progress benchmark value, it means that the patient's rehabilitation progress is lagging behind. At this time, an early warning progress report is generated to prompt the doctor to adjust the training plan.

[0037] Preferably, the system also includes a user interaction module, which includes an operation interface unit; the operation interface unit is used to provide patients and medical staff with a viewing entrance for rehabilitation training programs, demonstration playback of training movements, and data display functions for training effects through a touch screen or voice interaction device.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] The AI-based intelligent guidance system for rehabilitation training of burn patients provided by this invention significantly improves performance through the collaborative operation of multiple modules. The data acquisition module sets monitoring sites and deploys sensing devices in the rehabilitation training scenario for burn patients. The physiological signal acquisition unit uses a biosensor group to collect real-time physiological characteristic data such as skin temperature, heart rate, and blood oxygen levels. The motion trajectory acquisition unit can obtain training movement data. The data conversion unit converts the data format, filters out interference, and synchronizes the data. Compared with traditional manual data collection methods, the data obtained is more real-time, comprehensive, and accurate, providing a rich and reliable data foundation for subsequent rehabilitation training.

[0040] The rehabilitation assessment module constructs burn injury assessment models and functional impairment analysis models, using machine learning to analyze physiological characteristic data and motion analysis to identify the standard degree of training movement data. The burn depth identification unit uses a three-dimensional texture model combined with a classification algorithm to more accurately assess the severity of burn injuries. The wound healing analysis unit assesses healing progress based on repair parameters, and the functional impairment analysis unit uses a motor function assessment model to evaluate functional impairment. This provides a comprehensive and scientific assessment of a patient's burn recovery status and functional impairment level, changing the previous assessment method that relied primarily on subjective judgment.

[0041] The training planning module aligns and correlates the assessment results, extracts training parameter sets, and presets training target values, generating a preliminary matching and rehabilitation training plan. The stage goal setting unit sets training intensity targets based on the severity of the burn injury and the progress of wound healing. The movement parameter matching unit determines training movement parameters based on the type of functional impairment and motor function assessment results. This allows the training plan to be more tailored to the individual patient's specific situation, improving its personalization and precision while overcoming the limitations of traditional, standardized program development.

[0042] After generating a plan, the intelligent guidance module optimizes the intensity, frequency, and duration of the movement within the plan based on the patient's individual characteristics. The real-time movement correction unit compares movement parameters with real-time training movement data to determine deviations and analyze the degree of movement standardization. The training prompt unit presets prompt rules based on the patient's cognitive and communication habits, prompting the patient to adjust their movements based on deviations, ensuring that the patient's training movements are standardized and providing timely and effective guidance during the training process.

[0043] The effectiveness feedback module conducts time-series comparative analysis of training data to obtain training effectiveness indicators and dynamically updates the parameters of the burn injury assessment model and the functional impairment analysis model. The training data storage unit categorizes and stores training data, while the rehabilitation progress analysis unit generates progress reports based on training effectiveness indicators. This helps doctors understand patients' recovery progress and provides a basis for adjusting training plans, thus achieving dynamic optimization and scientific management of rehabilitation training. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a working principle diagram of the AI-based intelligent guidance system for rehabilitation training of burn patients according to the present invention;

[0045] Figure 2 This is the workflow diagram of the data acquisition module;

[0046] Figure 3 This is the workflow diagram of the data conversion unit;

[0047] Figure 4 This is the workflow diagram of the rehabilitation assessment module. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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.

[0049] See also Figures 1-4 The present invention provides an AI-based intelligent guidance system for rehabilitation training of burn patients, and the specific implementation methods are as follows:

[0050] The system comprises a data acquisition module, a rehabilitation assessment module, a training planning module, an intelligent guidance module, and an effectiveness feedback module. During burn patient rehabilitation training, the data acquisition module establishes monitoring points at the burn site and key body nodes, deploying sensing devices such as a biosensor array and a motion capture device. The biosensor array collects real-time physiological data such as the patient's skin temperature, heart rate, and blood oxygen levels, while the motion capture device collects training movement data. This data is then transmitted to the data acquisition module's data conversion unit for format conversion.

[0051] The rehabilitation assessment module constructs a burn injury assessment model and a functional impairment analysis model, uses machine learning technology to analyze physiological characteristic data, uses motion analysis technology to identify the standard degree of training movement data, and comprehensively evaluates the patient's burn recovery status and functional impairment level based on the collected data.

[0052] The training planning module aligns the assessed burn recovery status and functional impairment level, performs correlation calculations to obtain the training parameter set, presets the training target value, and preliminarily matches and generates a rehabilitation training plan for the patient at the current stage.

[0053] After generating a rehabilitation training plan, the intelligent guidance module calculates the adjustment parameter set based on the patient's individual characteristics, performs a secondary match between the preset adjustment baseline value and the adjustment parameter set, and optimizes the specific values ​​of movement intensity, frequency, and duration in the rehabilitation training plan.

[0054] The effect feedback module conducts time-series comparative analysis on the physiological characteristic data and training movement data collected during the execution of the optimized rehabilitation training program to obtain training effect indicators, and dynamically updates the parameter configuration of the burn injury assessment model and functional impairment analysis model based on the training effect indicators.

[0055] Example 1:

[0056] The data acquisition module is specifically composed of a physiological signal acquisition unit, a motion trajectory acquisition unit and a data conversion unit.

[0057] The physiological signal acquisition unit consists of a skin condition acquisition unit and a vital sign acquisition unit. The skin condition acquisition unit is used to monitor the patient's skin temperature at the burn site in real time. The skin temperature sensor assembly used in this unit consists of a contact probe, a signal amplifier, and a data recorder. The contact probe directly contacts the patient's skin at the burn site. Its sensitive temperature sensing element detects temperature changes on the skin surface in real time and accurately captures temperature values. During the temperature acquisition process, the contact probe continuously collects data, recording temperature fluctuations over a specific timeframe and generating temperature fluctuation range data. Furthermore, the probe periodically collects temperature data at a set sampling interval to ensure data continuity and real-time availability. The raw temperature signal collected is relatively weak and may be contaminated by interference. This is where the signal amplifier comes in. It amplifies the raw temperature signal, enhancing its strength. Furthermore, through built-in filtering circuits and other devices, it filters out interference components to ensure a clearer and more accurate output signal. The signal processed by the signal amplifier is transmitted to the data recorder, which has the functions of data storage and preliminary processing. It stores the processed temperature data in a certain format for subsequent data transmission and processing.

[0058] The vital signs acquisition unit is primarily responsible for collecting real-time heart rate and blood oxygen data. The heart rate and blood oxygen sensor assembly it uses consists of a photodetector, a filtering circuit, and a digital-to-analog converter. The photodetector utilizes the principle of photoplethysmography, emitting light of a specific wavelength onto the patient's skin surface. As the light passes through the skin tissue, it is absorbed and reflected by the blood beneath the skin. The beating of the heart causes cyclical changes in blood volume within the blood vessels, which in turn causes cyclical variations in the amount of light absorbed and reflected. The photodetector captures these cyclical variations in light intensity and converts them into corresponding electrical signals, which contain information such as the patient's heart rate and blood oxygen saturation. However, these electrical signals inevitably contain some noise and interference. The filtering circuit filters the electrical signals, removing high-frequency noise and low-frequency interference, resulting in a purer signal. The filtered analog electrical signal is then converted to a digital signal by the digital-to-analog converter, making it easier for computer systems to store, transmit, and process the digital signal. When collecting data, the heart rate and blood oxygen sensor group also records the measurement location information and evaluates the stability of the collected signal to form complete heart rate and blood oxygen data, including heart rate frequency, blood oxygen saturation, measurement location and signal stability.

[0059] The motion trajectory acquisition unit connects to the motion capture device's control terminal via a communication protocol. Motion capture devices are typically deployed around the rehabilitation training environment and can comprehensively capture the patient's movements during training. The control terminal serves as the management and data processing center for the motion capture devices, responsible for receiving and processing the data collected by the motion capture devices. After establishing a connection with the control terminal, the motion trajectory acquisition unit can read the motion capture device's acquisition parameters from the control terminal in real time. These acquisition parameters include the sampling rate, which determines the number of times the motion capture device collects data per unit time. A higher sampling rate results in more precise motion capture; spatial accuracy, which measures the accuracy of the motion capture device's determination of an object's position in space; and angular resolution, which reflects the device's ability to resolve changes in the object's rotation angle. The motion trajectory acquisition unit extracts these acquisition parameters, including sampling rate, spatial accuracy, and angular resolution, in real time and aggregates the extracted data to obtain the patient's training movement data during rehabilitation training. This data accurately describes the spatial motion trajectory and posture changes of various body parts.

[0060] The data conversion unit performs the crucial task of data processing and integration within the entire data acquisition module. After receiving physiological characteristic data from the physiological signal acquisition unit and training motion data from the motion trajectory acquisition unit, the data conversion unit first filters out interference signals and outliers. Since the data acquisition process is inevitably affected by external environmental factors and sensor performance, some interference signals and outliers may be present in the collected data. Using specific algorithms and programs, the data conversion unit analyzes and judges the data, identifying and removing these interference signals and outliers to ensure data accuracy. Next, the data conversion unit unifies the data formats collected by different sensor types. Different sensor types, such as skin temperature sensors, heart rate and blood oxygen sensors, and motion capture devices, often collect data in different formats, which complicates subsequent data processing and analysis. The data conversion unit uses a data format conversion program to convert these various data types into a unified, standardized format, facilitating system processing and analysis. Finally, the data conversion unit synchronizes the timestamps of the collected physiological characteristic data and training motion data through time synchronization. Because the patient's physiological characteristics and training movement data are collected simultaneously by different sensors, they need to be aligned in time to accurately analyze the relationship between physiological characteristics and training movements. The data conversion unit adds a precise timestamp to each data point and sorts and matches the data based on the timestamp. This allows the acquisition of skin temperature, heart rate, joint movement angle, and other data at the same time point, achieving data integration and unification.

[0061] Example 2:

[0062] The rehabilitation assessment module includes an injury degree assessment unit, a functional impairment analysis unit and a model calibration unit.

[0063] The injury severity assessment unit includes a burn depth identification unit and a wound healing analysis unit. The burn depth identification unit extracts patient burn area data and wound image information from the electronic medical record system. The electronic medical record system stores a comprehensive record of the patient's burn condition from the time of injury to the current stage. The burn area data clearly indicates the extent of burns on various parts of the patient's body, while the wound image information provides a high-definition image of the burn site's appearance. Image processing software is used to perform 3D modeling on the wound image. By analyzing and calculating information such as color and texture at each pixel in the image, a 3D texture model of the burn site is constructed. This model clearly distinguishes the epidermal, dermal, and subcutaneous tissue damage within the wound. For example, epidermal damage exhibits specific texture and color characteristics in the model, while dermal damage has a different appearance. Subcutaneous tissue damage also reflects corresponding structural changes in the model. Furthermore, typical burn site features, such as redness and swelling, blister distribution, and eschar extent, are annotated on the model. These annotations provide a visual reference for subsequent assessment. After the initial assessment is completed, the constructed three-dimensional texture model is processed using a classification algorithm. This algorithm assigns biological properties such as thermal conductivity, elastic modulus, and regenerative capacity to skin tissue. These properties can reflect the physical and physiological characteristics of skin tissue under different degrees of damage. The collection cycle, analysis window, and threshold range of physiological signals are also set. Through static assessment, dynamic assessment, and continuous assessment, the severity of the patient's burn injury is comprehensively assessed. Static assessment focuses on observing the state of the burn site at a certain moment, dynamic assessment focuses on changes in the burn site over a period of time, and continuous assessment continuously tracks the development trend of the burn site. Combining these assessment methods, an accurate conclusion on the severity of the burn injury is drawn.

[0064] The wound healing analysis unit receives wound repair parameters, including new epidermal thickness, capillary density, and inflammatory factor levels. New epidermal thickness reflects the regeneration of epidermal tissue at the burn site and is acquired using specialized testing equipment and methods. Capillary density reflects blood circulation within the wound, and its changes are closely related to wound healing. Inflammatory factor levels are a key indicator of the degree of inflammatory response within the wound and reflect whether the wound healing process is normal. The wound healing analysis unit performs feature extraction analysis on these input repair parameters. Using specific algorithms and procedures, it extracts key information related to wound healing and assesses the healing progress of the patient's burn site at different stages of recovery. Through comprehensive analysis and processing of these parameters, a clear understanding of the stage of the wound's progression from initial injury to gradual healing is achieved.

[0065] The Dysfunction Analysis Unit establishes a motor function assessment model that covers joint range of motion, muscle strength, and coordination metrics. Joint range of motion measures the range of motion of a joint and is obtained using specialized measurement tools and methods. Muscle strength reflects the force generated by muscle contraction and can be measured using specialized instruments. Coordination measures reflect the ability of various body parts to coordinate and cooperate in completing movements. Kinematic equations are applied to analyze the patient's training movement data. Based on the principles of physics, kinematic equations quantify information such as the trajectory, velocity, and acceleration of various body parts in the training movement data. Motion analysis techniques are used to analyze the analyzed data in depth to determine the patient's type of dysfunction and assess the degree of joint motion restriction, muscle contraction weakness, and coordination level during movement execution. For example, changes in joint motion angles in the training movement data can be analyzed to determine whether joint motion is restricted and to what extent; muscle contraction weakness can be assessed based on the force exerted by the muscles during the movement; and coordination level can be determined by observing the coordination of various body parts during movement execution.

[0066] The model calibration unit imports the patient's actual recovery data into the burn injury assessment model and functional impairment analysis model. This data, derived from various monitoring data collected during rehabilitation training and physicians' clinical assessment records, includes information on changes in the patient's burn injury status and functional recovery. Model parameter calibration uses a specific algorithm and program to adjust various model parameters based on the discrepancies between actual recovery data and the model's predictions, resulting in an updated model. Real-time patient physiological characteristics and training movement data are collected and transmitted to feature extraction and motion analysis technologies for dynamic assessment. This dynamic assessment provides a timely reflection of the patient's current physical condition and rehabilitation training effectiveness. The dynamic assessment results are then imported into the updated model, adjusting the weighting coefficients of the burn injury assessment model and functional impairment analysis model in real time to ensure that the models more accurately reflect the patient's actual condition. Furthermore, the assessment results are displayed on a data dashboard, allowing medical staff to intuitively view information such as the patient's burn injury extent, wound healing progress, and functional impairment. The dashboard also provides a parameter adjustment function for physicians, allowing them to further adjust and optimize the model based on the patient's specific situation.

[0067] Example 3:

[0068] The training planning module includes a stage goal setting unit, an action parameter matching unit and a plan generation unit.

[0069] The stage goal setting unit constructs a training goal setting algorithm based on the calibrated assessment results from the rehabilitation assessment module. These results include data related to the severity of the patient's burn injury, such as burn depth and burn area percentage, as well as calibrated wound healing progress data, such as changes in neoepidermal thickness over time and fluctuations in inflammatory cytokine levels. Through comprehensive analysis of this data, the training intensity target for the patient's current rehabilitation stage is calculated. The setting of training intensity targets involves determining the range of motion of the joints at the burn site and improving muscle strength. Specifically, the maximum range of motion and minimum muscle contraction force at the burn site are determined based on extensive clinical rehabilitation case data and computational models based on human kinematics. In practice, the range of motion is determined within a reasonable achievable range based on the physiological characteristics of each joint and the patient's current injury status. For example, for a patient with limited knee joint motion due to burns, the desired maximum range of motion for the next stage is determined based on the patient's current range of motion, burn recovery status, and joint motion improvement data from similar patients at similar recovery stages. For muscle strength setting, the patient's current muscle strength data will be obtained through professional muscle strength testing equipment. Combined with the rehabilitation process, a gradually increasing minimum contraction force target will be set. This target must be challenging to promote muscle recovery, but also within the patient's tolerance range to avoid secondary injury.

[0070] The movement parameter matching unit constructs a movement parameter matching algorithm, which performs calculations based on the calibrated assessment results. The calculations are based on the assessment results, including data on functional impairment types at different recovery stages, such as joint stiffness and muscle atrophy, as well as the calibrated motor function assessment results, which include specific quantitative data on joint range of motion, muscle strength, and coordination. By comparing functional impairment types at different recovery stages and combining them with the motor function assessment results, the amplitude, speed, and number of repetitions for training movements are calculated. The range of motion is determined based on the patient's current joint mobility and pain tolerance. For example, for patients with limited joint mobility and pain sensitivity, the initial range of motion is set to a smaller value, which is gradually increased as rehabilitation progresses. Movement speed is determined based on the patient's muscle strength and coordination. For patients with weaker muscle strength and coordination, the speed is set to a slower value to ensure accuracy and safety. The number of repetitions is determined based on a comprehensive consideration of the training goals and the patient's physical condition. For exercises that require muscle strength, an appropriate repetition range is set based on the patient's current muscle endurance. At the same time, the movement parameter matching unit matches the calculated training movement parameters with the appropriate rehabilitation training movement type for the patient's current stage. The rehabilitation training movement type library stores a variety of movements for different burn recovery stages and functional impairment types, such as joint range of motion training, muscle strength training, and coordination training. An algorithm matches the calculated parameters with the movements in the movement type library to select the most suitable rehabilitation training movement for the patient's current stage.

[0071] The program generation unit constructs a training program generation algorithm, which takes the training intensity target and training action parameters as input. Assume that the training intensity target is represented by the vector Indicates that Represents the maximum target range of motion of the joint at the burn site. represents the minimum contraction force target of the muscle, and so on. The number of indicators included in the training intensity target; the training action parameters are represented by vector Indicates that Indicates the range of motion, Indicates the speed of movement, Indicates the number of times the action is repeated. The number of parameters included in the training action parameters. The training plan generation algorithm generates a rehabilitation training plan based on these two vectors through a series of calculations and logical judgments. The rehabilitation training plan includes training instructions such as the action name, execution steps, and precautions. When generating the action name, it is named based on the matching rehabilitation training action type and specific movement characteristics, such as "Knee Flexion and Extension Exercise" and "Upper Limb Muscle Isometric Contraction Exercise." The execution steps are generated to describe in detail each action's starting position, movement process, ending position, and rhythm, ensuring that the patient can accurately understand and execute the movement. Precautions are generated based on factors such as the patient's physical condition and training environment. For example, patients are reminded to maintain correct posture during training to avoid excessive force that may cause pain or injury. For patients with cardiovascular disease, they are reminded to pay attention to the training intensity and rhythm to prevent discomfort. The training plan generation algorithm converts the training intensity target and training action parameters into a detailed, actionable rehabilitation training plan, providing clear guidance for patients' rehabilitation training.

[0072] Example 4:

[0073] The intelligent guidance module consists of a real-time motion correction unit and a training prompt unit. These two units work together to provide precise movement guidance for burn patients' rehabilitation training.

[0074] The real-time motion correction unit processes the motion parameters in the optimized rehabilitation training program with the patient's real-time training motion data. Suppose the patient is undergoing upper limb extension rehabilitation training. The rehabilitation training program stipulates that the arm extension angle of this action should reach 90 degrees, the action duration is 3 seconds, and it is repeated 10 times. During the training process, the motion capture device collects the patient's upper limb motion trajectory data in real time, including the position coordinates of the arm at each moment, joint angle changes and other information. After obtaining this data, the real-time motion correction unit first compares the spatial position of the motion parameters in the rehabilitation training program with the patient's real-time training motion data. It will compare the current position coordinates of the patient's arm with the standard position coordinates specified in the program one by one to determine whether the arm is in the correct spatial position. For example, if the program requires the arm to be extended to the horizontal direction, and the patient's arm tilts during training, the real-time motion correction unit can detect this spatial position deviation.

[0075] Then, time series alignment is performed to match the start and end time of each patient's action with the time requirements in the plan. If the plan stipulates that the action lasts for 3 seconds, but the patient actually completes the action in only 2 seconds, the real-time action correction unit can identify the time difference. Through spatial position comparison and time series alignment, the real-time action correction unit associates and calculates to obtain the action deviation value. This deviation value comprehensively reflects the degree of deviation of the patient's training action from the requirements of the plan in terms of space and time dimensions. For the above-mentioned upper limb extension training, if the arm extension angle only reaches 70 degrees and the duration is insufficient, the action deviation value will increase accordingly, and the real-time action correction unit will analyze the standard degree of the patient's training action in real time based on this.

[0076] The training prompt unit presets prompt rules based on the cognitive ability and communication habits of burn patients. For example, for patients with good cognitive ability and the ability to quickly understand complex instructions, the prompt rules can be set to be relatively detailed and professional; while for patients with weaker cognitive abilities, simpler and more intuitive prompt methods are used. Taking an elderly burn patient with average cognitive ability as an example, when performing lower limb leg lifting training, the training prompt unit performs a secondary match with the obtained movement deviation value. When the movement deviation value is less than the prompt threshold, it means that the patient's training movement meets the requirements of the program. At this time, the training prompt unit generates regular prompt information, such as "Great job, keep maintaining this posture and rhythm", to encourage the patient to continue the current training.

[0077] When the movement deviation value is greater than or equal to the prompt threshold, it indicates that there is a deviation in the patient's training movement. For example, in lower limb leg lift training, the patient does not lift their legs high enough and too quickly, causing the movement deviation value to exceed the threshold. At this time, the training prompt unit generates a correction prompt message, such as "Please lift your legs higher and slow down a bit", to guide the patient to adjust the movement amplitude or speed. The prompt information can be conveyed to the patient through voice broadcast or displayed in text form on the display screen of the training device.

[0078] For example, when a patient is doing finger grasping training, the rehabilitation training program requires the fingers to fully grasp a designated object and hold it for 2 seconds. The real-time motion correction unit monitors and finds that the patient's fingers have not been completely closed, and the holding time is less than 1 second, and calculates a large motion deviation value. Based on the preset prompt rules and the patient's communication habits and preference for concise instructions, the training prompt unit generates a prompt message "Grip hard and hold for a while" to help the patient adjust the training movements in time. Through the close cooperation of the real-time motion correction unit and the training prompt unit, the training movements are continuously monitored and guided during the patient's rehabilitation training process, so that the patient's training movements are as consistent as possible with the requirements of the rehabilitation training program, ensuring the effectiveness and safety of the rehabilitation training.

[0079] Example 5:

[0080] The effect feedback module includes a training data storage unit and a rehabilitation progress analysis unit. The operation interface unit in the user interaction module provides interactive functions for patients and medical staff. The training data storage unit collects various types of data from the data acquisition module during the patient's rehabilitation training. During each training session, the physiological signal acquisition unit obtains physiological characteristic data such as skin temperature, heart rate, blood oxygen saturation, and the motion trajectory acquisition unit records training action data such as joint movement angle, action execution speed and amplitude, etc. These data include the patient's physical reaction and action completion during the training process.

[0081] The training data storage unit organizes these data and classifies them according to dimensions such as training time and training items. For example, the upper limb extension training and lower limb leg lift training data conducted on the same day will be classified and stored separately. The data storage uses a specific data format to facilitate subsequent query and analysis. The stored data not only contains the original collected data, but also includes the patient's basic information, such as age, burn location, burn degree, etc., as well as relevant parameters of the training program, such as movement requirements, intensity settings, etc. Through this classified storage method, a large amount of training data is managed in an orderly manner to form a complete rehabilitation training database.

[0082] The rehabilitation progress analysis unit analyzes training effectiveness indicators based on pre-set progress assessment criteria. These criteria cover multiple aspects of burn recovery status and functional impairment, such as the degree of burn wound healing, improvement in joint range of motion, and changes in muscle strength. Training effectiveness indicators are extracted from training data. For example, wound recovery can be assessed by analyzing the changing temperature trend of the patient's burn site over time; and improvements in joint range of motion can be assessed by comparing the values ​​of joint motion angles at different training stages.

[0083] The rehabilitation progress analysis unit conducts a time-series comparative analysis of the training effect indicators to observe the changing trends of the patients in continuous training cycles. Taking a patient undergoing lower limb rehabilitation training as an example, in the first training cycle, the patient's knee joint motion angle was 30 degrees, which increased to 35 degrees in the second training cycle, and reached 40 degrees in the third training cycle. The rehabilitation progress analysis unit will compare these data and calculate the improvement value and improvement rate for each cycle. When the improvement rate of the training effect indicator is greater than the progress baseline value, it means that the patient's rehabilitation progress is in line with expectations, and a normal progress report is generated at this time. The report lists the training data and changes in each stage in detail, prompting the doctor to maintain the current training plan.

[0084] If the rate of improvement of the training effect indicator is equal to the progress benchmark value, it indicates that the patient's rehabilitation progress is basically up to standard, and the rehabilitation progress analysis unit generates a progress report. The report will specifically mark the indicators that require the doctor's attention, such as the patient's muscle fatigue during training, joint pain, etc., to remind the doctor to pay attention to the subsequent training effects. When the rate of improvement of the training effect indicator is less than the progress benchmark value, it means that the patient's rehabilitation progress is lagging behind, and the system generates an early warning progress report. The report not only presents the training data and progress, but also analyzes the reasons that may lead to progress lags, such as irregular training movements, inappropriate training intensity, etc., prompting the doctor to adjust the training plan.

[0085] The user interaction module's user interface unit implements interactive functions via a touch screen or voice interaction device. Patients can view rehabilitation training plans on the touch screen. The plans are presented in a graphic and text format, with detailed images and text descriptions of the movement names and execution steps, making them easier for patients to understand and remember. Demonstration videos of the training movements can also be played on the interface, allowing patients to intuitively see demonstrations of standard movements. Patients can also view their own training effect data, such as the completion of each training movement and changes in physiological indicators, to understand their rehabilitation progress.

[0086] Medical staff can use the user interface to fully understand the patient's rehabilitation information. They can view the implementation of the patient's rehabilitation training plan, compare training data and assessment results, and evaluate the effectiveness of the training plan. The burn injury assessment and functional impairment analysis results displayed on the data dashboard, as well as the rehabilitation progress report, are all clearly presented on the user interface, making it easier for medical staff to make decisions and adjust treatment plans. For example, after viewing a patient's early warning progress report, medical staff can directly modify the training plan on the user interface, adjusting the intensity, frequency, or training items, and push the new plan to the patient.

[0087] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0088] 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. An AI-based intelligent guidance system for rehabilitation training of burn patients, characterized by: It includes data acquisition module, rehabilitation assessment module, training planning module, intelligent guidance module and effect feedback module; The data acquisition module is used to set monitoring sites in the burn patient rehabilitation training scene and deploy sensor devices to collect the patient's physiological characteristics data and training movement data in real time, and convert the collected data into a format; The rehabilitation assessment module is used to construct a burn injury assessment model and a functional impairment analysis model, and uses machine learning technology to analyze the patient's physiological characteristic data. At the same time, it uses motion analysis technology to identify the standard degree of training movement data. Based on the real-time collected physiological characteristic data and training movement data, it comprehensively evaluates the patient's burn recovery status and functional impairment level. The training planning module is used to align the assessed burn recovery status and functional impairment level, perform correlation calculation to obtain a training parameter set, preset a training target value, and perform preliminary matching to generate a rehabilitation training plan for the patient at the current stage; The intelligent guidance module is used to calculate and obtain an adjustment parameter set based on the individual characteristics of the patient after generating a rehabilitation training plan, and to perform a secondary match between the preset adjustment baseline value and the adjustment parameter set to optimize the specific values ​​of the intensity, frequency and duration of the movement in the rehabilitation training plan; The effect feedback module is used to perform time-series comparative analysis on the physiological characteristic data and training movement data collected during the execution of the optimized rehabilitation training program to obtain training effect indicators, and dynamically update the parameter configuration of the burn injury assessment model and the functional impairment analysis model based on the training effect indicators; The rehabilitation assessment module includes an injury degree assessment unit, a functional impairment analysis unit and a model calibration unit; The injury degree assessment unit includes a burn depth identification unit and a wound healing analysis unit; The burn depth identification unit extracts the patient's burn area data and wound image information from the electronic medical record system, uses image processing software to establish a three-dimensional texture model of the burn site, identifies epidermal damage, dermal damage, and subcutaneous tissue damage in the wound, and simultaneously annotates typical features of the burn site, including the redness and swelling area, blister distribution, and eschar range. After the initial assessment, a classification algorithm is used to define the thermal conductivity, elastic modulus, and regenerative biological properties of the skin tissue in the constructed three-dimensional texture model. Simultaneously, the physiological signal acquisition cycle, analysis window, and threshold range are set to perform static, dynamic, and continuous assessments to assess the severity of the patient's burn injury. The wound healing analysis unit is used to input wound repair parameters, including new epidermal thickness, capillary density, and inflammatory factor levels, and then perform feature extraction and analysis to assess the healing progress of the patient's burn site at different recovery stages; The functional impairment analysis unit is used to establish a movement function assessment model, including joint range of motion, muscle strength and coordination indicators, and then apply kinematic equations to analyze the patient's training movement data. The motion analysis technology is used to analyze the patient's functional impairment type and evaluate the degree of joint movement restriction, muscle contraction fatigue and coordination level of movement execution; The model calibration unit is used to import the patient's actual recovery data into the burn injury assessment model and the functional disorder analysis model to perform parameter calibration to obtain an updated model, and then collect the patient's physiological characteristics data and training movement data in real time, transmit them to the feature extraction analysis and motion analysis technology for dynamic assessment, and import the dynamic assessment results into the updated model, adjust the weight coefficients of the burn injury assessment model and the functional disorder analysis model in real time, and display the assessment results through the data dashboard to provide doctors with parameter adjustment functions.

2. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 1 is characterized by: The data acquisition module includes a physiological signal acquisition unit, a motion trajectory acquisition unit and a data conversion unit; The physiological signal acquisition unit includes a skin condition acquisition unit and a vital sign acquisition unit, which is used to monitor and collect the patient's physiological characteristic data in real time by deploying a biosensor group at the patient's burn site and key nodes on the body, and transmit the data to the data conversion unit via wireless transmission. The biosensor group includes a skin temperature sensor group and a heart rate and blood oxygen sensor group. The physiological characteristic data includes skin temperature values ​​and heart rate and blood oxygen values; The skin condition acquisition unit is used to monitor the skin temperature data of the patient's burn area in real time based on the skin temperature sensor group. The skin temperature sensor group includes a contact probe, a signal amplifier and a data recorder, which respectively collect the temperature value, fluctuation range and sampling interval of the skin temperature data; The vital sign acquisition unit is used to collect the patient's heart rate and blood oxygen data in real time based on the heart rate and blood oxygen sensor group. The heart rate and blood oxygen sensor group includes a photoelectric detector, a filter circuit and a digital-to-analog converter. The heart rate and blood oxygen data include heart rate frequency, blood oxygen saturation, measurement position and signal stability; The motion trajectory acquisition unit is used to establish a communication protocol to connect with the control terminal of the motion capture device, read the acquisition parameters of the motion capture device in the control terminal in real time, and extract and summarize the sampling rate, spatial accuracy and angular resolution of the acquisition parameters of the motion capture device in real time to obtain the patient's training motion data.

3. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 2 is characterized by: The data conversion unit is used to filter out interference signals and abnormal values ​​from the collected physiological characteristic data and training movement data, and unify the data formats of different types of sensors. At the same time, the collected physiological characteristic data and training movement data are timestamp-aligned through time synchronization processing to obtain the skin temperature value, heart rate frequency and joint movement angle at the same time point.

4. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 1 is characterized by: The training planning module includes a stage goal setting unit, an action parameter matching unit and a plan generation unit; The stage goal setting unit is used to construct a training goal setting algorithm, calculate and obtain the training intensity target of the patient's current rehabilitation stage based on the calibrated assessment results, and set the range of motion and muscle strength improvement indicators of the burned area; The motion parameter matching unit is used to construct a motion parameter matching algorithm, and calculate the amplitude, speed and number of repetitions of the training movement based on the calibrated evaluation results to match the rehabilitation training movement type suitable for the patient at the current stage; The program generation unit is used to construct a training program generation algorithm, calculate and obtain a rehabilitation training program based on the training intensity target and training action parameters, and generate training guidance content including action names, execution steps and precautions.

5. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 4 is characterized by: The stage goal setting unit is used to calculate and obtain the training intensity target by analyzing the severity of the patient's burn injury and combining it with the calibrated wound healing progress, and set the maximum motion angle of the patient's burned joints and the minimum contraction force of the muscles.

6. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 4 is characterized by: The motion parameter matching unit is used to calculate and obtain training motion parameters by comparing the types of functional impairments in different recovery stages and combining them with the calibrated motor function assessment results, so as to match the patient's tolerable motion amplitude and execution speed within a safe range at the current stage.

7. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 1 is characterized by: The intelligent guidance module includes a real-time action correction unit and a training prompt unit; The real-time motion correction unit is used to compare the motion parameters in the optimized rehabilitation training program with the patient's real-time training motion data in space and align the time series, and then perform correlation calculation to obtain the motion deviation value and perform real-time analysis on the standard degree of the patient's training motion; The training prompt unit is used to preset prompt rules based on the cognitive ability and communication habits of the burn patient, and perform secondary matching with the obtained movement deviation value to prompt the patient to adjust the training movement. The specific prompt scheme is as follows: when the movement deviation value is less than the prompt threshold, it means that the patient's training movement meets the requirements of the scheme. At this time, a regular prompt message is generated to encourage the patient to continue the current training; when the movement deviation value is greater than or equal to the prompt threshold, it means that there is a deviation in the patient's training movement. At this time, a correction prompt message is generated to guide the patient to adjust the movement amplitude or speed.

8. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 1 is characterized by: The effect feedback module includes a training data storage unit and a rehabilitation progress analysis unit; The training data storage unit is used to combine the acquired training effect indicators, organize the patient's physiological characteristic data and training movement data for each training session, and classify and store them in the rehabilitation training database; The rehabilitation progress analysis unit is used to perform a time-series comparative analysis based on preset progress assessment criteria and the acquired training effect indicators, analyze the changing trends of the patient's burn recovery status and functional impairment level during the continuous training cycle, and generate a corresponding progress report. The specific analysis scheme is as follows: when the improvement rate of the training effect indicator is greater than the progress baseline value, it indicates that the patient's rehabilitation progress is in line with expectations. In this case, a normal progress report is generated to prompt the doctor to maintain the current training plan; When the rate of improvement of the training effect indicator is equal to the progress benchmark value, it means that the patient's rehabilitation progress has basically met the standard. At this time, a progress attention report is generated to remind the doctor to pay attention to the subsequent training effects; when the rate of improvement of the training effect indicator is less than the progress benchmark value, it means that the patient's rehabilitation progress is lagging behind. At this time, an early warning progress report is generated to remind the doctor to adjust the training plan.

9. The AI-based intelligent guidance system for rehabilitation training of burn patients according to claim 1, characterized in that: The system also includes a user interaction module, which includes an operation interface unit; the operation interface unit is used to provide patients and medical staff with a viewing entry for rehabilitation training programs, demonstration playback of training movements, and data display of training effects through a touch screen or voice interaction device.

Images