Computer-aided wound surface pulling and closing management method and computer-aided wound surface pulling and closing management system
By real-time monitoring of multi-dimensional perception parameters of wounds, using computer systems to identify and predict wound characteristics, and building personalized healing plans, the problem of lack of personalized plans for wound healing management is solved, and personalized nursing control of wound healing is achieved.
Patent Information
- Application Number
- CN202510968419.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-14
- Publication Date
- 2025-10-17
AI Technical Summary
The existing technology lacks personalized solutions for wound healing management, making it difficult to formulate personalized nursing strategies based on different wound types and patient physical signs.
By real-time monitoring of multi-dimensional perception parameters such as wound tension, surface temperature, and tissue fluid exudation, the central computer system is used to call the wound characteristic identification network and the healing tension prediction network to build a wound healing case library and generate personalized wound healing plans.
It realizes personalized wound pulling and closing nursing management and control, and improves the quality of wound healing and the controllability of the recovery cycle.
Smart Images
Figure CN120809277A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of wound closure management, in particular to a computer-aided wound closure management method and system. BACKGROUND
[0002] In medical processes such as surgical operation, burn repair, chronic wound treatment, effective closure and nursing of wounds are key factors affecting healing quality and recovery period. Traditional wound closure treatment mainly relies on the clinical experience of doctors for judgment and manual operation, lacks real-time sensing means for wound physiological state, and it is difficult to develop individualized nursing strategies according to different wound types and patient signs. In addition, although some medical devices have tried to introduce tension sensors, temperature measurement and other functions to assist in judgment, a systematic and intelligent wound closure management mechanism has not yet been formed, and there are still significant deficiencies in wound feature recognition, tension control prediction and nursing scheme optimization. SUMMARY
[0003] The present application provides a computer-aided wound closure management method and system, which solves the technical problem of lack of individualized scheme in wound closure management in the prior art.
[0004] In a first aspect, the present application provides a computer-aided wound closure management method, which comprises:
[0005] Real-time monitoring of a set of multi-dimensional sensing key parameters of a target wound, the set of multi-dimensional sensing key parameters comprising wound tension, surface temperature and tissue fluid exudation; wirelessly transmitting the set of multi-dimensional sensing key parameters to a central computer system, calling a wound feature recognition network and a wound healing tension prediction network through the central computer system; using the wound feature recognition network to identify the features of the set of multi-dimensional sensing key parameters to obtain wound feature parameters; based on the wound healing tension prediction network, predicting the tension of the wound feature parameters, and outputting the wound required tension; constructing a wound closure case library, using the wound closure case library to analyze the wound feature parameters and the wound required tension, determining a target wound closure scheme, and based on the target wound closure scheme, performing wound closure nursing control.
[0006] In a second aspect, the present application provides a computer-aided wound closure management system, which comprises:
[0007] The data monitoring module: real-time monitoring of the multi-dimensional perception key parameter set of the target wound surface, the multi-dimensional perception key parameter set including wound tension, surface temperature and tissue fluid exudation; the data transmission module: wirelessly transmitting the multi-dimensional perception key parameter set to the central computer system, calling the wound characteristic identification network and the wound healing tension prediction network through the central computer system; the characteristic identification module: identifying the characteristics of the multi-dimensional perception key parameter set by using the wound characteristic identification network to obtain the wound characteristic parameters; the tension prediction module: predicting the tension of the wound characteristic parameters based on the wound healing tension prediction network to output the wound demand tension; the adjustment analysis module: constructing the wound closure case library, adjusting and analyzing the wound characteristic parameters and the wound demand tension by using the wound closure case library, determining the target wound closure scheme, and performing wound closure nursing management and control based on the target wound closure scheme.
[0008] The one or more technical solutions provided in the present application have at least the following technical effects or advantages:
[0009] First, the multi-dimensional perception key parameter set of the target wound surface is monitored in real time, and the multi-dimensional perception key parameter set includes wound tension, surface temperature and tissue fluid exudation. Then, the multi-dimensional perception key parameter set is wirelessly transmitted to the central computer system, and the wound characteristic identification network and the wound healing tension prediction network are called through the central computer system. Further, the wound characteristic identification network is used to identify the characteristics of the multi-dimensional perception key parameter set to obtain the wound characteristic parameters; the wound healing tension prediction network is used to predict the tension of the wound characteristic parameters to output the wound demand tension. Finally, the wound closure case library is constructed, the wound characteristic parameters and the wound demand tension are adjusted and analyzed by using the wound closure case library, the target wound closure scheme is determined, and the wound closure nursing management and control are performed based on the target wound closure scheme. The technical problem of lack of personalized scheme in the wound closure management in the prior art is solved, and the technical effect of personalized wound closure nursing management and control is achieved. BRIEF DESCRIPTION OF DRAWINGS
[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.
[0011] Figure 1 A computer-aided wound closure management method flowchart is provided for the embodiments of the present application.
[0012] Figure 2 A computer-aided wound closure management system structure diagram is provided for the embodiments of the present application.
[0013] Legend: data monitoring module 11, data transmission module 12, characteristic identification module 13, tension prediction module 14, adjustment analysis module 15. DETAILED DESCRIPTION
[0014] The present application provides a computer-aided wound contraction management method and system, which solves the technical problem of lack of personalized scheme in wound contraction management in the prior art.
[0015] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.
[0016] It should be noted that the terms "comprise" and "have" are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server comprising a series of steps or units does not have to be limited to those clearly listed steps or units, but can include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.
[0017] Embodiment one, as shown in the present application provides a computer-aided wound contraction management method, wherein the method comprises: Figure 1
[0018] Real-time monitoring of multi-dimensional perception key parameter set of target wound, the multi-dimensional perception key parameter set includes wound tension, surface temperature and tissue fluid exudation.
[0019] The physiological state parameters of the wound surface are continuously collected and recorded by multiple types of sensor modules arranged on the wound surface and periphery. The sensor modules at least include: a micro-strain sensor for measuring the tension of the wound, a thermistor or infrared temperature sensor for measuring the surface temperature of the wound, and a conductivity sensor or exudate volume sensing device for detecting the amount of tissue fluid exudation. The micro-strain sensor is fixedly arranged on the suture line or the tensioning area of the wound dressing, for real-time detection of the tension change during the tensioning process of the wound, and obtaining the dynamic curve of the tension value over time; the surface temperature sensor senses the local skin temperature by closely adhering to the wound area, for reflecting the risk of inflammatory response or infection; the tissue fluid sensor is embedded in the dressing layer, for detecting the volume change and exudation rate of the wound exudate. All sensors are integrated into a portable wearable monitoring terminal, and the collected multi-dimensional perception data set is transmitted in real time to a central computer system through a low-power wireless communication module (such as BLE, ZigBee or NB-IoT), to realize continuous, remote and high-precision monitoring of the target wound state, and to provide data support for subsequent characteristic identification and personalized tensioning decision-making.
[0020] The multi-dimensional perception key parameter set is wirelessly transmitted to a central computer system, and the wound characteristic identification network and the wound healing tension prediction network are called by the central computer system.
[0021] After the multi-dimensional perception key parameter set is collected, the data is packaged and transmitted in real time through the wireless communication module integrated in the perception terminal. The wireless communication module can use Bluetooth Low Energy (BLE), ZigBee, LoRa or NB-IoT communication protocols to adapt to the transmission distance and power consumption requirements in different scenarios. The data packets transmitted include multi-source parameters such as wound tension, surface temperature and tissue fluid exudation marked by time stamp. These data are received by the central computer system (such as a hospital nursing management server or a local intelligent analysis workstation) on the receiving end and stored in a designated data buffer area.
[0022] The central computer system is configured with an intelligent analysis engine for processing the above perception data. After receiving the real-time data stream, the wound characteristic identification network and the wound healing tension prediction network preloaded in the system are automatically called. The wound characteristic identification network is used to identify the current physiological structure state and evolution characteristics of the wound, and the wound healing tension prediction network is used to predict the optimal tension interval of the current wound state. Both networks can be called through the deep learning inference module inside the system, and the input is the above multi-dimensional perception key parameter set, and the output is the wound characteristic parameter and the wound required tension parameter respectively.
[0023] Further, the wound characteristic identification network and the wound healing tension prediction network are called by the central computer system, including:
[0024] The wound contraction historical data set is collected, the wound contraction historical data set is standardized to obtain a standard wound contraction historical data set, the standard wound contraction historical data set is subjected to characteristic sample identification training to generate a wound characteristic identification network, and the wound characteristic output sample set of the wound characteristic identification network is subjected to tension prediction training to construct a wound healing tension prediction network. The wound characteristic identification network and the wound healing tension prediction network are stored in the central computer system.
[0025] Preferably, the wound contraction historical data set covering different wound types, healing stages, contraction strategies and patient sign characteristics is collected, the wound contraction historical data set includes wound tension change curve, surface temperature change data, tissue fluid exudation record, and corresponding nursing measures, healing time and effect evaluation results. Through standardization processing of the wound contraction historical data set, including data missing repair, outlier rejection, scale normalization, time alignment and other operations, a standard wound contraction historical data set with unified structure and can be used for model training is generated. Then, the standard wound contraction historical data set is subjected to characteristic sample identification training to identify the feature parameters strongly related to the wound state, and a wound index characteristic sample set is constructed. The above sample set is trained and verified by using a deep neural network with multi-layer feature extraction capability (such as convolutional neural network CNN or multi-layer perceptron MLP), to generate a wound characteristic identification network for automatically outputting wound structure and state characteristics after inputting perception parameters. Then, the existing standard wound contraction historical data set is subjected to batch reasoning processing based on the above wound characteristic identification network, and a wound characteristic parameter sample set is output. The wound characteristic parameter sample set is arranged and integrated in time sequence or treatment cycle order to form a wound characteristic output sample sequence. Then, the sequence sample set is used as input, and a long short-term memory network (LSTM) or a gated recurrent unit (GRU) sequence modeling algorithm is used for supervised learning training to construct a wound healing tension prediction network for predicting the required contraction tension interval at different stages of the wound. Finally, the wound characteristic identification network and the wound healing tension prediction network trained and evaluated to reach the set precision threshold by the validation set are stored in the local database of the central computer system, and are used as the core computing module for subsequent processing of real-time perception parameters and generation of individualized contraction nursing scheme.
[0026] Further, the wound characteristic identification network is generated, including:
[0027] The key characteristic index set of the wound is extracted from the standard wound contraction historical data set, and the sample set of the wound characteristic index is obtained by sample identification according to the wound key characteristic index set. The wound characteristic recognition network is generated by using a deep neural network structure to recognize, train and verify and optimize the wound characteristic index sample set.
[0028] Preferably, the standard wound contraction historical data set is subjected to multidimensional analysis, and the key characteristic indexes highly related to the wound healing state are extracted, including but not limited to tension change rate, surface temperature gradient, total amount of tissue fluid exudation, exudation rate, wound area change rate, wound margin contraction ratio, color change distribution, etc. The original features are reduced and selected by a feature engineering algorithm (such as principal component analysis PCA, correlation coefficient analysis, information gain method, etc.), and a wound key characteristic index set with representativeness and discrimination is constructed. Subsequently, based on the wound key characteristic index set, each group of data samples in the standard wound contraction historical data set is identified for characteristics, a mapping relationship is established, different wound types, contraction strategies and healing stages are coded into labeled feature samples, and a wound characteristic index sample set is formed. Then, a deep neural network based on a multi-layer perception structure (such as MLP, ResNet variant or nested convolution structure CNN) is used to train the wound characteristic index sample set. During the training process, the cross-entropy loss function and the Adam optimizer are combined for back propagation and iterative optimization of the network weight. The model performance is monitored by introducing a validation set, and the early stopping mechanism is used to prevent overfitting. Finally, a wound characteristic recognition network with good generalization ability and high recognition accuracy is obtained. The wound characteristic recognition network can receive any group of perception parameter input and quickly output the matching wound state classification result.
[0029] Further, the wound healing tension prediction network is constructed, including:
[0030] The wound characteristic recognition network is used to identify the characteristics of the standard wound contraction historical data set, and a wound characteristic output sample set is output. The wound characteristic output sample set is arranged and integrated according to the time sequence information, and a wound characteristic sequence sample set is obtained. The long short-term memory network is used to train the wound characteristic sequence sample set for tension prediction, and the wound healing tension prediction network is constructed.
[0031] Preferably, the trained wound characteristic recognition network is called to perform feature recognition on each data record in the standard wound contraction history data set, and a corresponding wound characteristic output sample set is output. The wound characteristic output sample set is a high-dimensional feature vector formed after deep learning network processing, representing the comprehensive characteristic state of the wound under multi-parameter perception at each time. Subsequently, the wound characteristic output sample set is sorted according to the corresponding time stamp, treatment cycle or nursing stage to form a wound characteristic sequence sample set with time sequence logic, which reflects the state evolution trajectory of the wound in the entire healing cycle. Then, a long short-term memory network with time-dependent modeling capability is used to train the above wound characteristic sequence sample set. The training target is to enable the network to predict the optimal contraction tension value required at the current or future time according to the wound state characteristics within a period of time. During the training process, mean square error (MSE) is used as the loss function to minimize the error between the predicted tension value and the historical actual tension label. Finally, after evaluating the network performance through the validation set, the trained LSTM model is saved, i.e., a wound healing tension prediction network is formed. The wound healing tension prediction network can receive the current feature vector sequence output by the wound characteristic recognition network in real time and predict the range of contraction tension that should be applied to the current wound.
[0032] The wound characteristic recognition network is used to perform characteristic recognition on the multi-dimensional perception key parameter set to obtain wound characteristic parameters.
[0033] By calling the wound characteristic recognition network, multi-layer nonlinear mapping and feature extraction are performed on the multi-dimensional perception key parameter set to generate wound characteristic parameters, which include but are not limited to wound tension response type, healing stage classification result, tension sensitivity index, exudation activity level, temperature fluctuation amplitude feature, etc.
[0034] The wound healing tension prediction network is used to perform tension prediction on the wound characteristic parameters to output wound demand tension.
[0035] By calling the wound healing tension prediction network, the wound characteristic parameters are subjected to tension prediction to output wound demand tension, which reflects the optimal contraction force of the wound under the premise of ensuring healing effect and avoiding excessive stretching.
[0036] A wound contraction case library is constructed, and the wound characteristic parameters and the wound demand tension are analyzed by using the wound contraction case library to determine a target wound contraction scheme, and the wound contraction nursing control is performed based on the target wound contraction scheme.
[0037] A wound contraction case library is constructed by collecting a large number of historical wound contraction cases. The wound contraction case library contains multi-dimensional information such as different patient wound types, contraction schemes, body characteristic parameters, healing effects and nursing records.
[0038] Based on the wound characteristic parameters and the wound demand tension, cases in the wound contraction case library are screened to obtain applicable wound contraction cases. Specifically, by calculating the similarity between the target wound characteristic parameters and the corresponding parameters in the case library, the case that best matches the target wound state is selected. Further, in combination with the predicted wound demand tension, the contraction scheme in the selected case is dynamically adjusted and optimized to form a personalized target contraction scheme for the current wound.
[0039] Based on the target wound contraction scheme, the system generates specific nursing control instructions, including contraction force adjustment, dressing replacement timing, nursing frequency, and auxiliary treatment recommendations, to enable clinical medical staff to implement scientific and reasonable wound contraction nursing.
[0040] Further, determining the target wound contraction scheme includes:
[0041] Based on the case information in the set of available wound contraction cases, similarity calculations are performed on the wound characteristic parameters and the wound demand tension to obtain a set of wound contraction case similarities. The set of wound contraction case similarities is used to optimize the set of available wound contraction cases to obtain a matching wound contraction scheme. The matching wound contraction scheme is adjusted and corrected to determine the target wound contraction scheme.
[0042] Based on the case information in the set of available wound contraction cases obtained through screening, similarity calculations are performed on the wound characteristic parameters and the wound demand tension of the current target wound. The similarity calculations can use multi-dimensional feature matching methods such as Euclidean distance, cosine similarity, or weighted Manhattan distance to comprehensively consider tension characteristics, temperature parameters, exudate, and patient individual parameters, generating a set of wound contraction case similarities that reflect the matching degree of cases and the target wound.
[0043] Based on the obtained set of wound contraction case similarities, the set of available cases is optimized and screened. The optimization strategy includes selecting the top several cases with the highest similarity or setting a threshold to remove low-similarity cases, ultimately obtaining a set of matching wound contraction schemes that best match the target wound state and demand tension. Finally, the selected matching wound contraction schemes are carefully adjusted and corrected, and after adjustment and correction, the personalized target wound contraction scheme for the target wound is ultimately determined.
[0044] Further, adjusting and correcting the matching wound contraction scheme to determine the target wound contraction scheme includes:
[0045] Based on the matching wound contraction scheme, nursing simulation and healing effect evaluation are performed to obtain wound healing effect prediction parameters. The matching wound contraction scheme is adjusted and corrected based on the wound healing effect prediction parameters to determine the target wound contraction scheme.
[0046] Based on the matching wound contraction scheme, combined with the current wound characteristic parameters and the target patient body characteristics, a nursing simulation simulation is performed in the central computer system. The simulation simulation process adopts a digital twin model or a rule-driven physiological model to simulate the influence process of different contraction intensities, frequencies, and fitting methods on wound healing speed, tissue tension distribution, exudation control, etc.
[0047] In the nursing simulation simulation process, the system dynamically generates wound healing effect prediction parameters, including but not limited to: predicted healing time, tension stability score, tissue growth trend, infection risk index, and nursing intervention responsiveness score, etc. Based on the above-mentioned healing effect prediction parameters, the system adjusts and corrects the preliminary matched wound contraction scheme, including the initial tension setting of the contraction device, the adjustment frequency, the selection of the dressing type, the matching of the auxiliary anti-infection scheme, etc. The adjustment process is iteratively solved according to the preset efficacy objective function or optimization rule (such as maximizing the healing speed and minimizing the tissue tension fluctuation), to ensure that the finally generated target wound contraction scheme has better healing effect and clinical adaptability.
[0048] In summary, the embodiments of the present application have at least the following technical effects:
[0049] First, the multi-dimensional perception key parameter set of the target wound is monitored in real time, including wound tension, surface temperature, and tissue fluid exudation. Then, the multi-dimensional perception key parameter set is wirelessly transmitted to the central computer system, and the wound characteristic identification network and the wound healing tension prediction network are called by the central computer system. Further, the wound characteristic identification network is used to identify the characteristics of the multi-dimensional perception key parameter set to obtain the wound characteristic parameters; the wound healing tension prediction network is used to predict the wound characteristic parameters to output the wound demand tension. Finally, a wound contraction case library is constructed, and the wound contraction case library is used to adjust and analyze the wound characteristic parameters and the wound demand tension to determine the target wound contraction scheme, and the wound contraction nursing management is performed based on the target wound contraction scheme. The technical problem of lack of personalized scheme in the wound contraction management in the prior art is solved, and the technical effect of personalized wound contraction nursing management is achieved.
[0050] Embodiment two, based on the same inventive concept as the computer-aided wound contraction management method in the foregoing embodiments, as shown in Figure 2 The present application provides a computer-aided wound contraction management system, wherein the system comprises:
[0051] The data monitoring module 11 monitors a multi-dimensional sensing key parameter set of the target wound surface in real time, the multi-dimensional sensing key parameter set including wound tension, surface temperature and tissue fluid exudation; the data transmission module 12 wirelessly transmits the multi-dimensional sensing key parameter set to a central computer system, and calls a wound characteristic identification network and a wound healing tension prediction network through the central computer system; the characteristic identification module 13 identifies the characteristics of the multi-dimensional sensing key parameter set by using the wound characteristic identification network to obtain wound characteristic parameters; the tension prediction module 14 predicts the tension of the wound characteristic parameters based on the wound healing tension prediction network to output wound demand tension; the adjustment analysis module 15 constructs a wound contraction case library, and adjusts and analyzes the wound characteristic parameters and the wound demand tension by using the wound contraction case library to determine a target wound contraction scheme, and performs wound contraction nursing and control based on the target wound contraction scheme.
[0052] Further, the data transmission module 12 is used to execute the following method:
[0053] The wound contraction historical data set is collected, and the wound contraction historical data set is standardized to obtain a standard wound contraction historical data set; the standard wound contraction historical data set is subjected to characteristic sample identification training to generate a wound characteristic identification network; the wound characteristic output sample set of the wound characteristic identification network is subjected to tension prediction training to construct a wound healing tension prediction network; and the wound characteristic identification network and the wound healing tension prediction network are stored in the central computer system.
[0054] Further, the data transmission module 12 is used to execute the following method:
[0055] The standard wound contraction historical data set is subjected to key characteristic index extraction to obtain a wound key characteristic index set; the standard wound contraction historical data set is subjected to sample identification according to the wound key characteristic index set to obtain a wound index characteristic sample set; the wound index characteristic sample set is subjected to identification training and verification optimization by using a deep neural network structure to generate the wound characteristic identification network.
[0056] Further, the data transmission module 12 is used to execute the following method:
[0057] The standard wound contraction historical data set is subjected to characteristic identification based on the wound characteristic identification network to output a wound characteristic output sample set; the wound characteristic output sample set is arranged and integrated according to time sequence information to obtain a wound characteristic sequence sample set; the wound characteristic sequence sample set is subjected to tension prediction training by using a long short-term memory network to construct the wound healing tension prediction network.
[0058] Further, the adjustment analysis module 15 is configured to perform the following method:
[0059] obtaining a body characteristic parameter of the target wound object; dividing the wound contraction case library based on the body characteristic parameter to obtain a set of available wound contraction cases; and performing adjustment analysis on the wound characteristic parameter and the wound demand tension by using the set of available wound contraction cases to determine the target wound contraction scheme.
[0060] Further, the adjustment analysis module 15 is configured to perform the following method:
[0061] performing similarity calculation on the wound characteristic parameter and the wound demand tension based on each case information in the set of available wound contraction cases to obtain a set of wound contraction case similarities; performing optimization on the set of available wound contraction cases by using the set of wound contraction case similarities to obtain a matching wound contraction scheme; and performing adjustment and correction on the matching wound contraction scheme to determine the target wound contraction scheme.
[0062] Further, the adjustment analysis module 15 is configured to perform the following method:
[0063] performing nursing simulation and healing effect evaluation based on the matching wound contraction scheme to obtain a wound healing effect prediction parameter; and performing adjustment and correction on the matching wound contraction scheme by using the wound healing effect prediction parameter to determine the target wound contraction scheme.
[0064] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.
[0065] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
[0066] The present specification and drawings are merely exemplary of the present application, and any and all modifications, variations, combinations or equivalents that fall within the scope of the present application are considered to be covered by the present application. Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.
Claims
1. A computer-assisted wound closure management method, characterized in that: The method comprises: Real-time monitoring of a target wound surface's multi-dimensional sensing key parameter set, including wound surface tension, surface temperature, and tissue fluid exudate; Wirelessly transmitting the multi-dimensional sensing key parameter set to a central computer system, and invoking a wound surface characteristic recognition network and a wound surface healing tension prediction network through the central computer system; Using the wound surface characteristic recognition network to perform characteristic recognition on the multi-dimensional perception key parameter set to obtain wound surface characteristic parameters; Performing tension prediction on the characteristic parameters of the wound based on the wound healing tension prediction network, and outputting the required tension of the wound; A wound closure case library is constructed, and the wound closure case library is used to adjust and analyze the wound characteristic parameters and the wound tension requirement, determine the target wound closure plan, and perform wound closure nursing management and control based on the target wound closure plan.
2. A computer-assisted wound closure management method according to claim 1, characterized in that: The calling of the wound surface characteristic identification network and the wound surface healing tension prediction network by the central computer system includes: Collecting a historical data set of wound surface pulling and closing, and performing standardization processing on the historical data set of wound surface pulling and closing to obtain a standard historical data set of wound surface pulling and closing; Performing characteristic sample identification training on the standard wound surface pulling and closing historical data set to generate a wound surface characteristic recognition network; Performing tension prediction training based on the wound surface characteristic output sample set of the wound surface characteristic recognition network to construct a wound healing tension prediction network; The wound surface characteristic recognition network and the wound surface healing tension prediction network are stored in the central computer system.
3. A computer-assisted wound closure management method according to claim 2, characterized in that: The generating of the wound surface characteristic recognition network comprises: Extracting key characteristic indicators from the standard wound surface pulling historical data set to obtain a wound surface key characteristic indicator set; Performing sample identification on the standard wound surface pull-in historical data set according to the wound surface key characteristic indicator set to obtain a wound surface indicator characteristic sample set; A deep neural network structure is used to perform recognition training and verification optimization on the wound surface indicator characteristic sample set to generate the wound surface characteristic recognition network.
4. A computer-assisted wound closure management method according to claim 2, characterized in that: The method of constructing a wound healing tension prediction network includes: Based on the wound surface characteristic recognition network, characteristic recognition is performed on the standard wound surface pulling and closing historical data set, and a wound surface characteristic output sample set is output; Arranging and integrating the wound surface characteristic output sample set according to time sequence information to obtain a wound surface characteristic sequence sample set; A long short-term memory network is used to perform tension prediction training on the wound characteristic sequence sample set to construct the wound healing tension prediction network.
5. A computer-assisted wound closure management method according to claim 1, characterized in that: Determining the target wound surface closing plan includes: Obtaining physical characteristic parameters of the target wound surface object; Dividing the wound surface closing case library based on the body characteristic parameters to obtain an available wound surface closing case set; The available wound closing case set is used to adjust and analyze the wound characteristic parameters and the required wound tension to determine a target wound closing solution.
6. A computer-assisted wound closure management method according to claim 5, characterized in that: Determining the target wound surface closing plan includes: performing similarity calculation on the wound characteristic parameters and the required wound tension based on information of each case in the available wound closing case set, to obtain a wound closing case similarity set; The available wound closing case set is optimized using the wound closing case similarity set to obtain a matching wound closing solution; The matching wound surface closing scheme is adjusted and corrected to determine a target wound surface closing scheme.
7. A computer-assisted wound closure management method according to claim 6, characterized in that: The adjusting and correcting the matching wound surface closing scheme to determine a target wound surface closing scheme includes: Perform nursing simulation and wound healing effect evaluation based on the matching wound stretching and closing scheme to obtain wound healing effect prediction parameters; The matching wound healing scheme is adjusted and corrected using the wound healing effect prediction parameters to determine the target wound healing scheme.
8. A computer-aided wound closing management system, characterized in that: A computer-assisted wound closure management method for implementing any one of claims 1 to 7, the system comprising: Data monitoring module: real-time monitoring of the target wound surface's multi-dimensional perception key parameter set, including wound surface tension, surface temperature, and tissue fluid exudate; Data transmission module: wirelessly transmits the multi-dimensional sensing key parameter set to a central computer system, and calls a wound surface characteristic recognition network and a wound surface healing tension prediction network through the central computer system; Characteristic recognition module: using the wound surface characteristic recognition network to perform characteristic recognition on the multi-dimensional perception key parameter set to obtain wound surface characteristic parameters; Tension prediction module: performs tension prediction on the characteristic parameters of the wound based on the wound healing tension prediction network, and outputs the required tension of the wound; Adjustment and analysis module: Build a wound closure case library, use the wound closure case library to adjust and analyze the wound characteristic parameters and the wound tension requirement, determine the target wound closure plan, and perform wound closure nursing management and control based on the target wound closure plan.