Postoperative rehabilitation management system for patients in hand and foot surgery department

Abstract

The invention relates to the field of postoperative rehabilitation management, in particular to a postoperative rehabilitation management system for hand and foot surgical patients. The postoperative rehabilitation management system for the patient in the hand and foot surgery department comprises a data monitoring module, a rehabilitation plan prediction module and a rehabilitation plan dynamic adjustment module. By comprehensively analyzing the postoperative data of the patient, including the wound healing condition and the function recovery progress, the recovery trend of the patient in the postoperative training period can be predicted, a scientific initial function recovery training plan is generated, and clear guidance is provided for rehabilitation training of the patient; in the actual training process, the system further performs real-time dynamic adjustment on the rehabilitation training plan by comparing the difference between the actual recovery condition and the prediction trend; the method ensures that the rehabilitation training plan is always consistent with the actual recovery state of the patient, so that the rehabilitation effect is maximized, the rehabilitation time is shortened, and the risk of complications is reduced.

Description

Technical Field

[0001] The present invention relates to the field of postoperative rehabilitation management, and in particular to a postoperative rehabilitation management system for hand and foot surgery patients. Background Art

[0002] Hand and foot surgery usually involves delicate anatomical structures and complex functional recovery processes. Postoperative rehabilitation training plays a vital role in this field. Effective postoperative rehabilitation training can not only promote wound healing and reduce the risk of complications, but also speed up the recovery of the function of the affected limb. If you wait until the wound is completely healed before performing functional recovery, it may cause joint stiffness, soft tissue adhesion or muscle atrophy, which will seriously affect the subsequent recovery effect. Therefore, in postoperative rehabilitation, wound monitoring and functional training are often carried out at the same time. Postoperative rehabilitation training is an indispensable part of the hand and foot surgery treatment process.

[0003] Traditional postoperative rehabilitation training management methods mainly rely on the doctor's experience and regular evaluation; these methods often adopt standardized rehabilitation processes, lack personalized considerations, and cannot be adjusted in a timely manner according to the specific circumstances of each patient; in addition, traditional methods have limitations in data collection and analysis, mainly relying on the patient's subjective description and the doctor's observation, and lack objective quantitative data support; at the same time, traditional methods often ignore the continuity and dynamics of rehabilitation data; for example, the patient's wound healing and functional recovery progress are usually not monitored in real time, resulting in the rehabilitation training plan cannot be optimized in a timely manner according to the patient's recovery status. Summary of the invention

[0004] The present invention can predict the patient's recovery trend during the postoperative training period by comprehensively analyzing the patient's postoperative data, including wound healing status and functional recovery progress, and generate a scientific initial functional recovery training plan to provide clear guidance for the patient's rehabilitation training; during the actual training process, the system further dynamically adjusts the rehabilitation training plan in real time by comparing the difference between the actual recovery status and the predicted trend; this method ensures that the rehabilitation training plan is always consistent with the patient's actual recovery status, thereby maximizing the rehabilitation effect, reducing the rehabilitation time, and reducing the risk of complications.

[0005] Postoperative rehabilitation management system for hand and foot surgery patients, including: The data monitoring module includes a resting period data acquisition unit and a training period data acquisition unit; the resting period data acquisition unit is used to acquire the patient's wound healing data, functional recovery data and wound infection index once a day during the patient's postoperative resting period, wherein the patient's postoperative resting period is after the completion of the operation, and when the patient's wound healing data meets the preset first wound healing condition, the postoperative resting period ends and enters the postoperative training period; the training period data acquisition unit is used to acquire the patient's wound healing data and functional recovery data once a day during the patient's postoperative training period, wherein when the patient's wound healing data meets the preset second wound healing condition, the postoperative training period ends; The rehabilitation plan prediction module includes a training sample acquisition unit, a rehabilitation decision set acquisition unit, an individual feature set acquisition unit and a rehabilitation plan prediction unit; the training sample acquisition unit is used to acquire a training sample set; the rehabilitation decision set acquisition unit is used to, at the beginning of the postoperative training period, sort the wound healing data, functional recovery data and wound infection indicators acquired during the postoperative rest period for several days in time to form a rehabilitation decision set; the individual feature set acquisition unit is used to form an individual feature set from the patient's age data, weight data and basic disease data; the rehabilitation plan prediction unit is used to train a rehabilitation plan prediction model using the training sample set, use the acquired rehabilitation decision set and individual feature set as inputs to the rehabilitation plan prediction model, and output a predicted rehabilitation data set for the current patient; The dynamic adjustment module of the rehabilitation plan includes a weight calculation unit and a functional recovery training adjustment unit; the weight calculation unit is used to calculate the wound healing priority weight and the functional recovery priority weight of the day on any day during the patient's postoperative training period, according to the number of days from the start of the postoperative training period to that day, as well as the patient's age data, basic disease data and the complexity of the operation; the functional recovery training adjustment unit is used to use the wound healing data, functional recovery data, expected wound healing data and expected functional recovery data of the day, combined with the application of wound healing priority weight and functional recovery priority weight, to adjust the actual functional recovery training intensity and actual functional recovery training time of the next day.

[0006] Preferably, in the dynamic adjustment module of the rehabilitation plan, the wound healing priority weight and the function recovery priority weight of the day are calculated based on the number of days from the start of the postoperative training period to the day, the patient's age data, the underlying disease data and the complexity of the operation. The specific operation is as follows: Obtain the complexity of the current patient's surgery, normalize the obtained patient age data and surgery complexity, and obtain the patient's age score A and surgery complexity score C; at the same time, based on the patient's underlying disease data, obtain the underlying disease score B. If the patient has an underlying disease, the value of the underlying disease score B is 1; if the patient does not have an underlying disease, the value of the underlying disease score B is 0; Set minimum wound healing priority weight ; Maximum wound healing priority weight ; Using the formula Calculate the wound healing priority weight of the current patient on the first day of postoperative training ; Then use the formula Calculate the priority weight of obtaining function recovery ; For any day other than the first day during the postoperative training period of the current patient, use the formula Calculate the wound healing priority weight for the day , =1, 2, ..., n, where n represents the number of days of the patient's postoperative training period, is the attenuation control parameter, which is used to control the speed of wound healing priority weight attenuation; then the formula is used Calculate the priority weight of function recovery for the day .

[0007] Preferably, in the dynamic adjustment module of the rehabilitation plan, the wound healing data, functional recovery data, expected wound healing data and expected functional recovery data of the day are used, combined with the wound healing priority weight and the functional recovery priority weight, to adjust the actual functional recovery training intensity and the actual functional recovery training time of the next day. The specific operations are as follows: The acquired predicted rehabilitation data set includes the number of days of the current patient's postoperative training period and the predicted rehabilitation data for each day, and the predicted rehabilitation data includes the expected functional recovery training intensity, the expected functional recovery training time, the expected wound healing data, and the expected functional recovery data; the expected functional recovery training intensity and the expected functional recovery training time for each day are used as the initial functional recovery training plan for the current patient; during the patient's postoperative training period, the patient applies the initial functional recovery training plan for functional recovery training; Setting wound healing deviation tolerance threshold and functional recovery deviation tolerance threshold ; For any day during the current patient's postoperative training period, obtain the patient's wound healing data for that day and functional recovery data , based on the patient's expected wound healing data for the day and expected function recovery data , using the formula Calculate the wound healing progress deviation of the day ; At the same time, using the formula Calculate the functional recovery progress deviation for the day ; When wound healing progress deviates Greater than And the functional recovery progress deviation Less than Based on the patient's expected functional recovery time the next day and expected functional recovery training intensity , using the formula Calculate the actual functional recovery training time one day after acquisition ,in, is the training time adjustment coefficient, Prioritize the wound healing for the day; use the formula Calculate the actual functional recovery training intensity one day after acquisition ,in, is the training intensity adjustment factor; When wound healing progress deviates Less than And the functional recovery progress deviation Greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ,in, Restore the priority weights for the functions; at the same time, use the formula Calculate the actual functional recovery training intensity one day after acquisition ; When wound healing progress deviates Greater than And the functional recovery progress deviation At the same time greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ; At the same time, using the formula Calculate the actual functional recovery training intensity one day after acquisition ; Prompt the patient to use the actual functional recovery training time obtained on the next day and actual functional recovery training intensity Carry out functional recovery training the next day; if the wound healing progress is deviated Less than And the functional recovery progress deviation At the same time less than , the patient is prompted to apply the expected functional recovery training time of the next day and expected functional recovery training intensity Carry out functional recovery training the next day.

[0008] Preferably, in the rehabilitation plan prediction module, the rehabilitation plan prediction model is established based on the Transformer model, including: an input layer, a first fully connected layer, a Transformer encoder, a Transformer decoder and an output layer; the input layer is used to receive a rehabilitation decision set and an individual feature set; the first fully connected layer is used to extract the non-temporal features of the individual feature set; the Transformer encoder is used to extract the temporal features of the rehabilitation decision set; the Transformer decoder is used to generate the predicted number of days of the patient's postoperative training period and the expected functional recovery training intensity, expected functional recovery training time, expected wound healing data and expected functional recovery data for each day based on the non-temporal features of the individual feature set and the temporal features of the rehabilitation decision set; the output layer is used to output the generation result of the Transformer decoder.

[0009] Preferably, in the rehabilitation plan prediction module, the Transformer encoder in the rehabilitation plan prediction model includes an embedding layer, a multi-head self-attention mechanism layer and a feedforward network layer; the embedding layer is used to map the rehabilitation decision set to a high-dimensional feature space and add position information to form an embedded sequence; the multi-head self-attention mechanism layer is used to capture the dependencies between different time steps in the embedded sequence; the feedforward network layer is used to perform an independent nonlinear transformation on each time step output by the multi-head self-attention mechanism layer, and output the temporal characteristics of the rehabilitation decision set.

[0010] Preferably, in the rehabilitation plan prediction module, the Transformer decoder in the rehabilitation plan prediction model includes an input preparation layer, a second fully connected layer and a multi-step decoding loop layer; the input preparation layer is used to splice the temporal features of the rehabilitation decision set and the non-temporal features of the individual feature set to form global features, and generate initial inputs at the same time; the second fully connected layer is used to generate the predicted number of days of the patient's postoperative training period based on the global features; the multi-step decoding loop layer is used to gradually generate daily predicted rehabilitation data based on the predicted number of days of the patient's postoperative training period and the initial input generated by the input preparation layer, wherein the number of cycles is the predicted number of days of the patient's postoperative training period, and the predicted rehabilitation data for the day is generated in each cycle, and the generated predicted rehabilitation data and global features are used as inputs for the next cycle.

[0011] Preferably, in the rehabilitation plan prediction module, the specific operation of obtaining the training sample set is as follows: Based on the type of surgery of the current patient, the historical rehabilitation decision set of patients with the same type of surgery during the postoperative rest period and the individual feature set of each patient were selected from the historical medical database. At the same time, the number of days of the postoperative training period, the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data of all selected patients were obtained. Set a functional recovery progress threshold, and for any selected patient, determine whether the functional recovery data of the current patient at the end of the postoperative training period is greater than the functional recovery progress threshold. If so, the current patient is classified as a preferred patient; if not, no operation is performed; The data of all the preferred patients obtained are combined into a training sample set. Each training sample contains a historical rehabilitation decision set and individual feature set of a preferred patient in the postoperative rest period, as well as the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data during the postoperative training period.

[0012] Preferably, in the rehabilitation plan prediction module, the rehabilitation plan prediction model is obtained by training using the training sample set, and the specific operations are as follows: All training sample sets are divided into training data sets and validation data sets. For any training sample in the training data set, the patient's historical rehabilitation decision set and individual feature set are used as the input of the rehabilitation plan prediction model, and the number of days of the patient's postoperative training period and the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data are used as the target output of the rehabilitation plan prediction model to train the rehabilitation plan prediction model; then the validation data set is input into the rehabilitation plan prediction model to obtain the verification result; the training conditions are set to determine whether the obtained verification result meets the training conditions. If so, the trained rehabilitation plan prediction model is output; if not, the training data set is continued to be used to train the rehabilitation plan prediction model.

[0013] The present invention has the following advantages: The present invention can predict the patient's recovery trend during the postoperative training period by comprehensively analyzing the patient's postoperative data, including wound healing status and functional recovery progress, and generate a scientific initial functional recovery training plan to provide clear guidance for the patient's rehabilitation training; during the actual training process, the system further dynamically adjusts the rehabilitation training plan in real time by comparing the difference between the actual recovery status and the predicted trend; this method ensures that the rehabilitation training plan is always consistent with the patient's actual recovery status, thereby maximizing the rehabilitation effect, reducing the rehabilitation time, and reducing the risk of complications.

[0014] The present invention assigns personalized wound healing priority weights and functional recovery priority weights to each patient by considering factors such as the patient's age, weight, underlying disease and complexity of surgery; these weights change dynamically as the rehabilitation process progresses, reflecting the changes in rehabilitation focus at different stages, ensuring that the rehabilitation training plan can adapt to the individual differences of patients and the natural evolution of the rehabilitation process; in this way, the system can manage all aspects of rehabilitation training more finely, ensuring that appropriate attention is paid to key factors at critical moments, thereby optimizing the overall rehabilitation effect; this dynamic adjustment mechanism makes the present invention more flexible and effective in dealing with complex postoperative rehabilitation situations, and can provide patients with more comprehensive and detailed rehabilitation management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the structure of a postoperative rehabilitation management system for hand and foot surgery patients adopted in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0017] In view of the defects of traditional postoperative rehabilitation training management methods, the embodiment proposes a postoperative rehabilitation management system for hand and foot surgery patients, which aims to realize the personalized formulation and dynamic adjustment of rehabilitation training plans through advanced data monitoring and analysis technology; the system can timely identify the changes in key indicators in the rehabilitation process, predict rehabilitation trends, and adjust the intensity and content of rehabilitation training according to the actual situation of the patient by collecting and analyzing the patient's postoperative data in real time, as follows: Postoperative rehabilitation management system for hand and foot surgery patients, such as Figure 1 As shown, including: The data monitoring module includes a resting period data acquisition unit and a training period data acquisition unit; the resting period data acquisition unit is used to obtain the patient's wound healing data, functional recovery data and wound infection indicators once a day during the patient's postoperative resting period. The patient's postoperative resting period is after the operation is completed. This stage is mainly for the primary healing of the wound and reduces external interference, such as wound dehiscence or infection that may be caused by excessive activity. Through resting period monitoring, the medical team can ensure that the wound heals in the best environment and lay a good foundation for subsequent rehabilitation training. This stage of monitoring is to ensure that the wound can heal naturally without external interference, while monitoring potential infection risks. When the patient's wound healing data meets the preset first wound healing When the conditions are met, the postoperative rest period ends and the postoperative training period begins; the training period data acquisition unit is used to acquire the patient's wound healing data and functional recovery data once a day during the patient's postoperative training period. The monitoring at this stage is to evaluate the patient's performance and progress in rehabilitation training, while ensuring that the wound continues to heal under safe conditions. The postoperative training period ends when the patient's wound healing data meets the preset second wound healing condition; the reason why functional recovery training is performed while the wound is still healing is that modern rehabilitation concepts emphasize early intervention, that is, mild functional recovery training is performed as the wound begins to heal. The benefits of doing so include reducing the risk of muscle atrophy and joint stiffness, promoting blood circulation, and accelerating wound healing; The rehabilitation plan prediction module includes a training sample acquisition unit, a rehabilitation decision set acquisition unit, an individual feature set acquisition unit and a rehabilitation plan prediction unit; the training sample acquisition unit is used to acquire training sample sets, which include the postoperative rehabilitation data of previous patients, and are used to train the rehabilitation plan prediction model. By learning these data, the model can identify key patterns and trends in the rehabilitation process, providing a basis for prediction; the rehabilitation decision set acquisition unit is used to, at the beginning of the postoperative training period, sort the wound healing data, functional recovery data and wound infection indicators obtained during the postoperative rest period for several days in chronological order to form a rehabilitation decision set. These data reflect the patient's recovery without external interference, and provide a basis for formulating a personalized rehabilitation plan; the individual feature set acquisition unit is used to sort the patient's age and Age data, weight data and basic disease data constitute an individual feature set, where if the patient has basic diseases, the basic disease data is the type of the patient's disease; if the patient does not have basic diseases, the basic disease data is no disease. These individual characteristics are crucial to the formulation of a rehabilitation plan because they may affect the patient's recovery speed and response during the rehabilitation process; the rehabilitation plan prediction unit is used to train a rehabilitation plan prediction model using a training sample set, use the acquired rehabilitation decision set and individual feature set as inputs to the rehabilitation plan prediction model, and output a predicted rehabilitation data set for the current patient. This predicted rehabilitation data set includes predictions of the patient's expected functional recovery training intensity, expected functional recovery training time, expected wound healing data, and expected functional recovery data every day during the postoperative training period; The dynamic adjustment module of the rehabilitation plan includes a weight calculation unit and a functional recovery training adjustment unit; the weight calculation unit is used to calculate the wound healing priority weight and functional recovery priority weight of the day on any day during the patient's postoperative training period, based on the number of days from the start of the postoperative training period to that day, as well as the patient's age data, basic disease data and the complexity of the operation. The purpose of this unit is to dynamically calculate the priority weights of wound healing and functional recovery according to the patient's specific situation and the different stages of postoperative recovery. These weights reflect the relative importance of wound healing and functional recovery at a specific point in time. By considering factors such as the patient's age, basic disease and the complexity of the operation, the weight calculation unit can provide each patient with personalized weights, thereby more accurately guiding the adjustment of rehabilitation training; the functional recovery training adjustment unit is used to utilize the wound healing data of the day. The data, functional recovery data, expected wound healing data and expected functional recovery data are combined with the application of wound healing priority weights and functional recovery priority weights to adjust the actual functional recovery training intensity and actual functional recovery training time of the next day. Postoperative rehabilitation is a dynamic process. The patient's condition changes every day. A fixed rehabilitation training plan may not be able to adapt to these changes, resulting in poor rehabilitation effect or delayed rehabilitation process. Therefore, it is necessary to dynamically adjust the rehabilitation training plan based on the difference between the patient's real-time data and predicted data, as well as the individual differences of the patient. This adjustment is to ensure that rehabilitation training can both promote wound healing and maximize the effect of functional recovery. Inappropriate training intensity or excessive time may lead to delayed wound healing or hindered functional recovery. Dynamic adjustment can avoid these problems and ensure the safe and effective recovery of patients.

[0018] In the dynamic adjustment module of the rehabilitation plan, the wound healing priority weight and function recovery priority weight of the day are calculated based on the number of days from the start of the postoperative training period to the day, as well as the patient's age data, basic disease data, and the complexity of the operation. The specific operations are as follows: Obtain the complexity of the current patient's surgery, normalize the obtained patient age data and surgery complexity, and obtain the patient's age score A and surgery complexity score C; at the same time, based on the patient's underlying disease data, obtain the underlying disease score B. If the patient has an underlying disease, the value of the underlying disease score B is 1; if the patient does not have an underlying disease, the value of the underlying disease score B is 0; Set minimum wound healing priority weight ; Maximum wound healing priority weight ; Using the formula Calculate the wound healing priority weight of the current patient on the first day of postoperative training ; Then use the formula Calculate the priority weight of obtaining function recovery ; For any day other than the first day during the postoperative training period of the current patient, use the formula Calculate the wound healing priority weight for the day , =1, 2, ..., n, where n represents the number of days of the patient's postoperative training period, is the attenuation control parameter, which is used to control the speed of wound healing priority weight attenuation; then the formula is used Calculate the priority weight of function recovery for the day .

[0019] In the dynamic adjustment module of the rehabilitation plan, the wound healing data, functional recovery data, expected wound healing data and expected functional recovery data of the day are used, combined with the wound healing priority weight and functional recovery priority weight, to adjust the actual functional recovery training intensity and actual functional recovery training time of the next day. The specific operations are as follows: The acquired predicted rehabilitation data set includes the number of days of the current patient's postoperative training period and the predicted rehabilitation data for each day, and the predicted rehabilitation data includes the expected functional recovery training intensity, the expected functional recovery training time, the expected wound healing data, and the expected functional recovery data; the expected functional recovery training intensity and the expected functional recovery training time for each day are used as the initial functional recovery training plan for the current patient; during the patient's postoperative training period, the patient applies the initial functional recovery training plan for functional recovery training; Setting wound healing deviation tolerance threshold and functional recovery deviation tolerance threshold ; For any day during the current patient's postoperative training period, obtain the patient's wound healing data for that day and functional recovery data , based on the patient's expected wound healing data for the day and expected function recovery data , using the formula Calculate the wound healing progress deviation of the day ; At the same time, using the formula Calculate the functional recovery progress deviation for the day ; When wound healing progress deviates Greater than And the functional recovery progress deviation Less than Based on the patient's expected functional recovery time the next day and expected functional recovery training intensity , using the formula Calculate the actual functional recovery training time one day after acquisition ,in, is the training time adjustment coefficient, Prioritize the wound healing for the day; use the formula Calculate the actual functional recovery training intensity one day after acquisition ,in, is the training intensity adjustment factor; When wound healing progress deviates Less than And the functional recovery progress deviation Greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ,in, Restore the priority weights for the functions; at the same time, use the formula Calculate the actual functional recovery training intensity one day after acquisition ; When wound healing progress deviates Greater than And the functional recovery progress deviation At the same time greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ; At the same time, using the formula Calculate the actual functional recovery training intensity one day after acquisition ; Prompt the patient to use the actual functional recovery training time obtained on the next day and actual functional recovery training intensity Carry out functional recovery training the next day; if the wound healing progress is deviated Less than And the functional recovery progress deviation At the same time less than , the patient is prompted to apply the expected functional recovery training time of the next day and expected functional recovery training intensity Carry out functional recovery training the next day.

[0020] In the rehabilitation plan prediction module, the rehabilitation plan prediction model is established based on the Transformer model, including: an input layer, a first fully connected layer, a Transformer encoder, a Transformer decoder and an output layer; the input layer is used to receive the rehabilitation decision set and the individual feature set; the first fully connected layer is used to extract the non-temporal features of the individual feature set; the Transformer encoder is used to extract the temporal features of the rehabilitation decision set; the Transformer decoder is used to generate the predicted number of days of the patient's postoperative training period and the expected functional recovery training intensity, expected functional recovery training time, expected wound healing data and expected functional recovery data for each day based on the non-temporal features of the individual feature set and the temporal features of the rehabilitation decision set; the output layer is used to output the generation results of the Transformer decoder.

[0021] The rehabilitation plan prediction model of this system is improved and optimized on the basis of the conventional Transformer model to meet the unique needs of postoperative rehabilitation management of hand and foot surgery patients. Specifically, the model not only integrates the Transformer encoder to capture the temporal features in the rehabilitation decision set, but also introduces the first fully connected layer to extract the non-temporal features in the individual feature set, realizing the accurate consideration of individual differences such as patient age, weight, and underlying diseases. In addition, the decoder part of the model combines the second fully connected layer and the multi-step decoding cycle layer, which not only generates the predicted number of days of the patient's postoperative training period, but also gradually refines the expected functional recovery training intensity, time and wound healing data every day. This structure enables the model to comprehensively consider the individual characteristics of the patient and the dynamic changes of postoperative recovery, and generate a more accurate and personalized rehabilitation plan; In the rehabilitation plan prediction module, the Transformer encoder in the rehabilitation plan prediction model includes an embedding layer, a multi-head self-attention mechanism layer and a feedforward network layer; the embedding layer is used to map the rehabilitation decision set to a high-dimensional feature space and add position information to form an embedded sequence; the multi-head self-attention mechanism layer is used to capture the dependencies between different time steps in the embedded sequence; the feedforward network layer is used to perform an independent nonlinear transformation on each time step output by the multi-head self-attention mechanism layer, and output the temporal characteristics of the rehabilitation decision set.

[0022] In the rehabilitation plan prediction module, the Transformer decoder in the rehabilitation plan prediction model includes an input preparation layer, a second fully connected layer and a multi-step decoding loop layer; the input preparation layer is used to concatenate the temporal features of the rehabilitation decision set and the non-temporal features of the individual feature set to form global features and generate initial inputs at the same time; the second fully connected layer is used to generate the predicted number of days of the patient's postoperative training period based on the global features; the multi-step decoding loop layer is used to gradually generate daily predicted rehabilitation data based on the predicted number of days of the patient's postoperative training period and the initial input generated by the input preparation layer, where the number of cycles is the predicted number of days of the patient's postoperative training period, and the predicted rehabilitation data of the day is generated in each cycle, and the generated predicted rehabilitation data and global features are used as inputs for the next cycle.

[0023] In the rehabilitation plan prediction module, the specific operations for obtaining the training sample set are as follows: Based on the type of surgery of the current patient, the historical rehabilitation decision set of patients with the same type of surgery during the postoperative rest period and the individual feature set of each patient were selected from the historical medical database. At the same time, the number of days of the postoperative training period, the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data of all selected patients were obtained. Set the functional recovery progress threshold. For any selected patient, determine whether the functional recovery data of the current patient at the end of the postoperative training period is greater than the functional recovery progress threshold. If so, the current patient is classified as a preferred patient; if not, no operation is performed. The purpose of this step is to screen out patients with good recovery effects, namely "preferred patients", through the functional recovery progress threshold. The data of these patients will be used to train the rehabilitation plan prediction model because they represent successful rehabilitation cases. The data of all the preferred patients obtained are combined into a training sample set. Each training sample contains a historical rehabilitation decision set and individual feature set of a preferred patient during the postoperative rest period, as well as the functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data of each day during the postoperative training period. By integrating the data of the preferred patients into a training sample set, high-quality learning materials can be provided for the model. The data in these sample sets reflect the effective rehabilitation process and successful recovery results, which helps the model learn effective rehabilitation strategies.

[0024] In the rehabilitation plan prediction module, the rehabilitation plan prediction model is trained using the training sample set. The specific operations are as follows: All training sample sets are divided into training data sets and validation data sets. For any training sample in the training data set, the patient's historical rehabilitation decision set and individual feature set are used as the input of the rehabilitation plan prediction model, and the number of days of the patient's postoperative training period and the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data are used as the target output of the rehabilitation plan prediction model to train the rehabilitation plan prediction model; then the validation data set is input into the rehabilitation plan prediction model to obtain the verification result; the training conditions are set to determine whether the obtained verification result meets the training conditions. If so, the trained rehabilitation plan prediction model is output; if not, the training data set is continued to be used to train the rehabilitation plan prediction model.

[0025] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. Postoperative rehabilitation management system for hand and foot surgery patients, characterized by: include: The data monitoring module includes a resting period data acquisition unit and a training period data acquisition unit; the resting period data acquisition unit is used to acquire the patient's wound healing data, functional recovery data and wound infection index once a day during the patient's postoperative resting period, wherein the patient's postoperative resting period is after the completion of the operation, and when the patient's wound healing data meets the preset first wound healing condition, the postoperative resting period ends and enters the postoperative training period; the training period data acquisition unit is used to acquire the patient's wound healing data and functional recovery data once a day during the patient's postoperative training period, wherein when the patient's wound healing data meets the preset second wound healing condition, the postoperative training period ends; The rehabilitation plan prediction module includes a training sample acquisition unit, a rehabilitation decision set acquisition unit, an individual feature set acquisition unit and a rehabilitation plan prediction unit; the training sample acquisition unit is used to acquire a training sample set; the rehabilitation decision set acquisition unit is used to, at the beginning of the postoperative training period, sort the wound healing data, functional recovery data and wound infection indicators acquired during the postoperative rest period for several days in time to form a rehabilitation decision set; the individual feature set acquisition unit is used to form an individual feature set from the patient's age data, weight data and basic disease data; the rehabilitation plan prediction unit is used to train a rehabilitation plan prediction model using the training sample set, use the acquired rehabilitation decision set and individual feature set as inputs to the rehabilitation plan prediction model, and output a predicted rehabilitation data set for the current patient; The dynamic adjustment module of the rehabilitation plan includes a weight calculation unit and a functional recovery training adjustment unit; the weight calculation unit is used to calculate the wound healing priority weight and the functional recovery priority weight of the day on any day during the patient's postoperative training period, according to the number of days from the start of the postoperative training period to that day, as well as the patient's age data, basic disease data and the complexity of the operation; the functional recovery training adjustment unit is used to use the wound healing data, functional recovery data, expected wound healing data and expected functional recovery data of the day, combined with the application of wound healing priority weight and functional recovery priority weight, to adjust the actual functional recovery training intensity and actual functional recovery training time of the next day.

2. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 1, characterized in that: In the dynamic adjustment module of the rehabilitation plan, the wound healing priority weight and function recovery priority weight of the day are calculated based on the number of days from the start of the postoperative training period to the day, as well as the patient's age data, basic disease data, and the complexity of the operation. The specific operations are as follows: Obtain the complexity of the current patient's surgery, normalize the obtained patient age data and surgery complexity, and obtain the patient's age score A and surgery complexity score C; at the same time, based on the patient's underlying disease data, obtain the underlying disease score B. If the patient has an underlying disease, the value of the underlying disease score B is 1; if the patient does not have an underlying disease, the value of the underlying disease score B is 0; Set minimum wound healing priority weight ; Maximum wound healing priority weight ; Using the formula Calculate the wound healing priority weight of the current patient on the first day of postoperative training ; Then use the formula Calculate the priority weight of obtaining function recovery ; For any day other than the first day during the postoperative training period of the current patient, use the formula Calculate the wound healing priority weight for the day , =1, 2, ..., n, where n represents the number of days of the patient's postoperative training period, is the attenuation control parameter, which is used to control the speed of wound healing priority weight attenuation; then the formula is used Calculate the priority weight of function recovery for the day .

3. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 2, characterized in that: In the dynamic adjustment module of the rehabilitation plan, the wound healing data, functional recovery data, expected wound healing data and expected functional recovery data of the day are used, combined with the wound healing priority weight and functional recovery priority weight, to adjust the actual functional recovery training intensity and actual functional recovery training time of the next day. The specific operations are as follows: The acquired predicted rehabilitation data set includes the number of days of the current patient's postoperative training period and the predicted rehabilitation data for each day, and the predicted rehabilitation data includes the expected functional recovery training intensity, the expected functional recovery training time, the expected wound healing data, and the expected functional recovery data; the expected functional recovery training intensity and the expected functional recovery training time for each day are used as the initial functional recovery training plan for the current patient; during the patient's postoperative training period, the patient applies the initial functional recovery training plan for functional recovery training; Setting wound healing deviation tolerance threshold and functional recovery deviation tolerance threshold ; For any day during the current patient's postoperative training period, obtain the patient's wound healing data for that day and functional recovery data , based on the patient's expected wound healing data for the day and expected function recovery data , using the formula Calculate the wound healing progress deviation of the day ; At the same time, using the formula Calculate the functional recovery progress deviation for the day ; When wound healing progress deviates Greater than And the functional recovery progress deviation Less than Based on the patient's expected functional recovery time the next day and expected functional recovery training intensity , using the formula Calculate the actual functional recovery training time one day after acquisition ,in, is the training time adjustment coefficient, Prioritize the wound healing for the day; use the formula Calculate the actual functional recovery training intensity one day after acquisition ,in, is the training intensity adjustment factor; When wound healing progress deviates Less than And the functional recovery progress deviation Greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ,in, Restore the priority weights for the functions; at the same time, use the formula Calculate the actual functional recovery training intensity one day after acquisition ; When wound healing progress deviates Greater than And the functional recovery progress deviation At the same time greater than When using the formula Calculate the actual functional recovery training time one day after acquisition ; At the same time, using the formula Calculate the actual functional recovery training intensity one day after acquisition ; Prompt the patient to use the actual functional recovery training time obtained on the next day and actual functional recovery training intensity Carry out functional recovery training the next day; if the wound healing progress is deviated Less than And the functional recovery progress deviation At the same time less than , the patient is prompted to apply the expected functional recovery training time of the next day and expected functional recovery training intensity Carry out functional recovery training the next day.

4. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 3, characterized in that: In the rehabilitation plan prediction module, the rehabilitation plan prediction model is established based on the Transformer model, including: an input layer, a first fully connected layer, a Transformer encoder, a Transformer decoder and an output layer; the input layer is used to receive the rehabilitation decision set and the individual feature set; the first fully connected layer is used to extract the non-temporal features of the individual feature set; the Transformer encoder is used to extract the temporal features of the rehabilitation decision set; the Transformer decoder is used to generate the predicted number of days of the patient's postoperative training period and the expected functional recovery training intensity, expected functional recovery training time, expected wound healing data and expected functional recovery data for each day based on the non-temporal features of the individual feature set and the temporal features of the rehabilitation decision set; the output layer is used to output the generation results of the Transformer decoder.

5. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 4, characterized in that: In the rehabilitation plan prediction module, the Transformer encoder in the rehabilitation plan prediction model includes an embedding layer, a multi-head self-attention mechanism layer and a feedforward network layer; the embedding layer is used to map the rehabilitation decision set to a high-dimensional feature space and add position information to form an embedded sequence; the multi-head self-attention mechanism layer is used to capture the dependencies between different time steps in the embedded sequence; the feedforward network layer is used to perform an independent nonlinear transformation on each time step output by the multi-head self-attention mechanism layer, and output the temporal characteristics of the rehabilitation decision set.

6. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 5, characterized in that: In the rehabilitation plan prediction module, the Transformer decoder in the rehabilitation plan prediction model includes an input preparation layer, a second fully connected layer, and a multi-step decoding loop layer; the input preparation layer is used to concatenate the temporal features of the rehabilitation decision set and the non-temporal features of the individual feature set to form a global feature and generate the initial input at the same time; The second fully connected layer is used to generate the predicted number of days of the patient's postoperative training period based on the global features; the multi-step decoding cycle layer is used to gradually generate the predicted rehabilitation data for each day based on the predicted number of days of the patient's postoperative training period and the initial input generated by the input preparation layer, where the number of cycles is the predicted number of days of the patient's postoperative training period, and the predicted rehabilitation data for the day is generated in each cycle, and the generated predicted rehabilitation data and global features are used as input for the next cycle.

7. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 6, characterized in that: In the rehabilitation plan prediction module, the specific operations for obtaining the training sample set are as follows: Based on the type of surgery of the current patient, the historical rehabilitation decision set of patients with the same type of surgery during the postoperative rest period and the individual feature set of each patient were selected from the historical medical database. At the same time, the number of days of the postoperative training period, the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data of all selected patients were obtained. Set a functional recovery progress threshold, and for any selected patient, determine whether the functional recovery data of the current patient at the end of the postoperative training period is greater than the functional recovery progress threshold. If so, the current patient is classified as a preferred patient; if not, no operation is performed; The data of all the preferred patients obtained are combined into a training sample set. Each training sample contains a historical rehabilitation decision set and individual feature set of a preferred patient in the postoperative rest period, as well as the daily functional recovery training intensity, functional recovery training time, wound healing data and functional recovery data during the postoperative training period.

8. The postoperative rehabilitation management system for hand and foot surgery patients according to claim 7, characterized in that: In the rehabilitation plan prediction module, the rehabilitation plan prediction model is trained using the training sample set. The specific operations are as follows: All training sample sets are divided into training data sets and validation data sets. For any training sample in the training data set, the patient's historical rehabilitation decision set and individual feature set are used as inputs of the rehabilitation plan prediction model. The number of days of the patient's postoperative training period, the daily functional recovery training intensity, functional recovery training time, wound healing data, and functional recovery data are used as target outputs of the rehabilitation plan prediction model to train the rehabilitation plan prediction model. The validation data set is then input into the rehabilitation plan prediction model to obtain the validation results. The training conditions are set to determine whether the obtained verification results meet the training conditions. If so, the trained rehabilitation plan prediction model is output; if not, the training data set is continued to be used to train the rehabilitation plan prediction model.

Images