Rehabilitation nursing planning system and method for burn and plastic surgery of patient
By combining regional segmentation and physiological parameters with a generative adversarial network model to optimize burn and plastic surgery rehabilitation planning, the problems of insufficient data and inadequate regional segmentation were solved, resulting in more accurate rehabilitation and nursing planning and improved patient satisfaction.
Patent Information
- Application Number
- CN202511750080.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-26
- Publication Date
- 2026-02-24
AI Technical Summary
In existing burn and plastic surgery rehabilitation programs, the applicability of models is insufficient due to privacy and limited data volume in burn images. The models do not take into account the situation of patients and medical staff, and the regions are not divided, resulting in inaccurate data generation, high error rate in automated design, and impact on rehabilitation and nursing outcomes.
By collecting image data of burn sites and dividing them into regions, a generative adversarial network model is used to generate regional burn image data. Combined with physiological parameters and nursing staff status characteristics, a rehabilitation nursing planning model is constructed, and the generated rehabilitation nursing methods are optimized.
It improves the accuracy of automated design for burn and plastic surgery rehabilitation planning, reduces the complexity of image data acquisition, increases the satisfaction of rehabilitation care, and ensures the adaptability and accuracy of rehabilitation methods.
Smart Images
Figure CN121565374A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical information planning technology, and in particular relates to a rehabilitation nursing planning system and method for patients with burns and plastic surgery. Background Technology
[0002] Burns are a common and serious accidental injury that not only causes immense physical pain but can also lead to long-term disability and psychological trauma. In traditional burn treatment, the attending physician mainly relies on clinical observation and experience to determine the severity of the patient's burns when the patient seeks medical attention, and then provides an appropriate treatment plan based on the assessment.
[0003] However, the complexity of burn and plastic surgery rehabilitation lies in its slow and uncertain recovery process, requiring real-time adjustments to the treatment plan based on the patient's progress. For patients with moderate to severe burns, the burn wounds are large and deep, the skin's barrier function is severely impaired, and their immunity decreases significantly after the burn, making them highly susceptible to pathogens such as bacteria and fungi. Therefore, the rehabilitation planning methods differ greatly depending on the location of the burn.
[0004] Currently, medical institutions primarily rely on specialized nursing staff to regularly monitor the wound healing of inpatients undergoing burn and plastic surgery rehabilitation (regularly changing dressings, keeping the wound clean, and promoting wound healing, etc.). However, wound infections are often difficult to detect with the naked eye in their early stages, and their progression can be extremely rapid. If not detected and treated promptly, these infections can worsen and deepen the wound, prolong hospital stays, and even endanger the patient's life. This makes burn patient care heavily dependent on the experience and expertise of nursing staff. This method of relying on manual observation has limitations. On the one hand, nursing staff cannot be present with patients at all times for continuous monitoring, which can lead to gaps in the monitoring process. On the other hand, differences in the experience and expertise of different nursing staff can result in inaccurate assessments of wound infections, thus affecting the patient's treatment outcome.
[0005] Current burn and plastic surgery nursing care is merely a simple routine care method for nursing staff, and the planning of nursing methods does not take into account the patient, nursing staff, and the burn condition.
[0006] Based on the above analysis, the problems and shortcomings of the existing technology are as follows:
[0007] In existing burn and plastic surgery rehabilitation planning, there are issues such as insufficient model applicability due to patient privacy and low data sample size. Furthermore, the lack of simultaneous consideration of patient and medical staff circumstances, and the absence of regional segmentation of burn areas for targeted image generation, leads to inaccurate data. Consequently, automated design of burn and plastic surgery rehabilitation plans suffers from high error rates and significant deviations from reality, failing to effectively assist rehabilitation nurses in adaptive planning of rehabilitation methods.
[0008] How can we generate images with high adaptability to burn and plastic surgery sites, and use physiological parameters related to patients and medical staff that influence rehabilitation planning, to derive patient-friendly rehabilitation plans? How can we obtain rehabilitation planning methods that are more suitable for patients, not only reducing the complexity of image data acquisition, but also improving the satisfaction of burn patients with rehabilitation care? Summary of the Invention
[0009] To address the aforementioned technical problems, this invention proposes a rehabilitation nursing planning system and method for patients with burns and plastic surgery.
[0010] In a first aspect of the invention, a method for rehabilitation nursing planning for patients with burns and plastic surgery is provided, the method comprising:
[0011] The system collects image data of the burn sites of patients and divides them into regions. Based on the region-corrected generative adversarial network model, it generates regional burn image data and collects physiological parameters of burn patients and the state characteristics of nursing staff.
[0012] Based on the processing of regional burn image data, physiological parameters, and state characteristics, the characteristics of plastic surgery rehabilitation planning are obtained, and the first rehabilitation nursing methods issued by doctors for the burn site at this time are collected.
[0013] A rehabilitation nursing planning model is constructed based on the characteristics of the plastic surgery rehabilitation plan and the first rehabilitation nursing method. The second rehabilitation nursing method, derived from the image of the burn site before processing based on the rehabilitation nursing planning model, is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model.
[0014] Furthermore, the image data of the burn sites were divided into face, hands, feet, and chest and back based on the region.
[0015] Furthermore, the region-based modified generative adversarial network model uses generation modification parameters calculated based on the proportion of burn area in the divided parts, burn color, and disordered eschar texture to modify the generative adversarial network model.
[0016] Furthermore, the region-based modified generative adversarial network model specifically uses generation modification parameters to modify the decision maker of the generative adversarial network model.
[0017] Furthermore, the physiological parameters include blood glucose, blood pressure, and weight, and the nursing staff's status characteristics include the nursing staff's heart rate, respiratory rate, and extreme values of skin conductance signals.
[0018] Furthermore, the second rehabilitation nursing method derived from the burn site image before generation based on the rehabilitation nursing planning model is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model. Specifically, the optimized model is determined by calculating the range similarity of the values obtained from the burn site image before generation and the burn site image after generation processed by the rehabilitation nursing planning model.
[0019] Furthermore, the rehabilitation nursing planning model adopts a multi-classification model based on Fisher's criterion.
[0020] A rehabilitation nursing planning system for patients with burns and plastic surgery is also provided, including a burn site image data collection and site segmentation module, a region-corrected generative adversarial network model module, a patient and nursing staff physiological state feature collection module, a plastic surgery rehabilitation planning feature processing module, a rehabilitation nursing planning model construction and optimization module, and a rehabilitation nursing planning module, characterized in that:
[0021] The burn site image data collection and site division module is used to collect image data of the patient's burn sites and divide them into face, hands, feet and chest and back according to the region.
[0022] The region-corrected generative adversarial network model module generates a region-corrected generative adversarial network model, which uses generation correction parameters to correct the decision-maker.
[0023] The patient and nursing staff physiological state characteristic collection module is used to collect physiological parameters of burn patients and state characteristics of nursing staff.
[0024] The plastic surgery rehabilitation planning feature processing module: obtains plastic surgery rehabilitation planning features based on regional burn image data, physiological parameters and state features, and collects the first rehabilitation nursing method issued by the doctor for the burn site at this time;
[0025] The rehabilitation nursing planning model construction and optimization module: Based on the plastic surgery rehabilitation planning characteristics and the first rehabilitation nursing method, a rehabilitation nursing planning model is constructed. Based on the second rehabilitation nursing method obtained from the burn site image before processing and generation of the rehabilitation nursing planning model, the first rehabilitation nursing method is compared with the second rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model.
[0026] The rehabilitation nursing planning module utilizes an optimized rehabilitation nursing planning model to process the plastic surgery rehabilitation planning features obtained from subsequent patient treatment, resulting in a third rehabilitation nursing method. Based on this third rehabilitation nursing method, rehabilitation nursing planning is carried out for burn plastic surgery patients.
[0027] Furthermore, based on the region-corrected generative adversarial network model, the generative correction parameters calculated using the proportion of burn area in the divided parts, burn color, and disordered eschar texture are used to correct the generative adversarial network model.
[0028] In existing burn and plastic surgery rehabilitation planning, a generative adversarial network model based on regional feature parameters to generate correction coefficients addresses the issue of insufficient model applicability due to patient privacy and low sample size in burn images. Furthermore, by considering both patient and medical staff circumstances and dividing the burn area into regions for targeted image generation, the generated data becomes more accurate, improving the accuracy of automated burn and plastic surgery rehabilitation planning, reducing actual deviations, and effectively assisting rehabilitation nurses in adaptive planning of rehabilitation methods.
[0029] This invention generates highly adaptable images of burn and plastic surgery sites and references physiological parameters related to patients and medical staff that influence rehabilitation planning. Furthermore, it combines image data generated by an adaptive data generation network with the status characteristics of patients and relevant medical staff to derive patient-friendly rehabilitation planning, obtaining a rehabilitation planning method that is more suitable for patients. This not only reduces the complexity of image data acquisition but also improves the satisfaction of burn patients with rehabilitation care.
[0030] Further embodiments and improvements of the present invention will be described in conjunction with the accompanying drawings and specific examples. Attached Figure Description
[0031] Figure 1 This is a flowchart of a rehabilitation nursing planning method for patients with burns and plastic surgery according to the present invention;
[0032] Figure 2 This is a schematic diagram of a rehabilitation nursing planning system for patients with burns and plastic surgery according to the present invention;
[0033] Figure 3 This is a schematic diagram of the generative adversarial network principle in an embodiment of the present invention;
[0034] Figure 4 This is a schematic diagram of the Fisher criterion multi-classifier in the embodiments of the present invention;
[0035] Figure 5 This is a schematic diagram of an electronic device structure for implementing the method of the present invention in an embodiment of the present invention. Detailed Implementation
[0036] The invention will now be further described in conjunction with the accompanying drawings and specific embodiments. All images acquired in this invention are color images.
[0037] like Figure 2 As shown, the system of this invention belongs to the medical information technology consulting service of medical information planning, and therefore belongs to the information planning information technology consulting service system.
[0038] In a first aspect of the invention, a method for rehabilitation nursing planning for patients with burns and plastic surgery is provided, the method comprising:
[0039] The system collects image data of the burn sites of patients and divides them into regions. Based on the region-corrected generative adversarial network model, it generates regional burn image data and collects physiological parameters of burn patients and the state characteristics of nursing staff.
[0040] Based on the processing of regional burn image data, physiological parameters, and state characteristics, the characteristics of plastic surgery rehabilitation planning are obtained, and the first rehabilitation nursing methods issued by doctors for the burn site at this time are collected.
[0041] A rehabilitation nursing planning model is constructed based on the characteristics of the plastic surgery rehabilitation plan and the first rehabilitation nursing method. The second rehabilitation nursing method, derived from the image of the burn site before processing based on the rehabilitation nursing planning model, is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model.
[0042] Furthermore, the burn site image data is divided into face, hands, feet, and chest and back based on the region. Due to the different burn sites, the model may encounter applicability issues during the image generation process. Therefore, the feature parameters of the burn site are used to modify the generative adversarial network model to generate burn image data.
[0043] Furthermore, the region-based modified generative adversarial network model uses generation correction parameters calculated based on the proportion of burn area in different parts of the burn zone, burn color, and the disorder of eschar texture to correct the generative adversarial network model. Since the burn conditions vary in different parts of the body, the proportion of burn area can effectively represent the burn location. At the same time, the burn color and the degree of disorder of eschar texture in different parts of the burn significantly affect the determination of the burn location and subsequent rehabilitation care. Therefore, the calculation method for the generation correction parameters is as follows:
[0044]
[0045] In the formula, To generate correction parameters, To define the burn area of part x, To define the area of region x in the image, the images acquired in this invention are all three-channel image data. Therefore, the burn color is characterized using the maximum value of each channel to reduce the amount of data processing. n represents the number of image pixels. This represents the maximum grayscale value among the three channels of the i-th image pixel. This represents the average maximum value of the three gray channels where the darkest color is visible in the image indicating a burn. This indicates the disorder of eschar texture. The more disordered the eschar in the image, the more severe the burn, and the more closely the generated burn image data needs to resemble this type of image data. m represents the number of image pixels corresponding to the eschar in the image. x represents the probability of the gray value of the j-th eschar image pixel appearing, where x can be arbitrarily chosen as one of the face, hand, foot, or chest and back.
[0046] Furthermore, the region-based modified generative adversarial network model specifically uses generation modification parameters to modify the decision maker of the generative adversarial network model. The decision maker calculation formula is as follows:
[0047]
[0048] In the formula, The maximum error rate of the decision maker. Image distribution of burn sites. The distribution of burn site images generated by the region-modified generative adversarial network model. To generate corrected parameters, this invention addresses the issue that current generative adversarial network (GAN) models rely on conventional decision-making methods to determine the usability of burn image data, neglecting the impact of burn site image features on the decision-maker in subsequent classification samples. This invention, recognizing the differential influence of burn site feature parameters on burn image data, employs a fusion of feature data coefficients based on burn area proportions, burn color, and chaotic eschar texture to correct the generated burn site image distribution data. This avoids excessive use of client privacy image features, making the images more universally applicable. It also considers the significant impact of burn sites on rehabilitation care and effectively addresses the limited data available for the same burn site, providing higher-quality input data for subsequent rehabilitation care planning.
[0049] Furthermore, the physiological parameters include blood glucose, blood pressure, and weight, and the physiological parameter features are represented by a feature vector composed of the patient's blood glucose, blood pressure, and weight (blood glucose, blood pressure, weight). The nursing staff's state features include the nursing staff's heart rate, respiratory rate, and extreme values of electrodermal signals, and the nursing staff's state features are represented by a feature vector composed of these features (heart rate, respiratory rate, extreme values of electrodermal signals).
[0050] Furthermore, the second rehabilitation nursing method derived from the burn site image before generation based on the rehabilitation nursing planning model is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model. Specifically, this optimized model is determined by calculating the range similarity of the values obtained from the burn site image before and after generation processed by the rehabilitation nursing planning model. The method for calculating the range similarity of the determination values is as follows:
[0051]
[0052] In the formula, For range similarity, The values are obtained by processing the burn site image generated using the region-modified generative adversarial network model with the rehabilitation care planning model. The values obtained from the burn site images before the generation of the rehabilitation nursing planning model are processed. The rehabilitation nursing planning model is selected based on the deviation of the values input into the model after generating the burn image data. This is to avoid excessive model generalization and ensure that the output rehabilitation nursing method is controlled within a reasonable range. In this embodiment, [the model is selected]. A rehabilitation nursing planning model with a success rate of 20% or less is considered an optimized rehabilitation nursing planning model.
[0053] Furthermore, the rehabilitation nursing planning model adopts a multi-classification model based on Fisher's criterion.
[0054] Furthermore, the calculation formula for the multi-class classification model based on the Fisher criterion is as follows:
[0055]
[0056] in , These are the normal vector and intercept of the hyperplane, respectively, obtained through training using features of orthopedic rehabilitation planning and the first rehabilitation care methods issued by doctors for burn sites. The rehabilitation and nursing methods to be output. The features of the plastic surgery rehabilitation plan are described above.
[0057] It also provides a rehabilitation nursing planning system for patients with burns and plastic surgery, including a burn site image data collection and site segmentation module, a region correction generative adversarial network model module, a patient and nursing staff physiological state feature collection module, a plastic surgery rehabilitation planning feature processing module, a rehabilitation nursing planning model construction and optimization module, and a rehabilitation nursing planning module;
[0058] The burn site image data collection and site division module is used to collect image data of the patient's burn sites and divide them into face, hands, feet and chest and back according to the region.
[0059] The region-corrected generative adversarial network model module generates a region-corrected generative adversarial network model, which uses generation correction parameters to correct the decision-maker.
[0060] The patient and nursing staff physiological state characteristic collection module is used to collect physiological parameters of burn patients and state characteristics of nursing staff.
[0061] The plastic surgery rehabilitation planning feature processing module: obtains plastic surgery rehabilitation planning features based on regional burn image data, physiological parameters and state features, and collects the first rehabilitation nursing method issued by the doctor for the burn site at this time;
[0062] The rehabilitation nursing planning model construction and optimization module: Based on the plastic surgery rehabilitation planning characteristics and the first rehabilitation nursing method, a rehabilitation nursing planning model is constructed. Based on the second rehabilitation nursing method obtained from the burn site image before processing and generation of the rehabilitation nursing planning model, the first rehabilitation nursing method is compared with the second rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model.
[0063] The rehabilitation nursing planning module utilizes an optimized rehabilitation nursing planning model to process the plastic surgery rehabilitation planning features obtained from subsequent patient treatment, resulting in a third rehabilitation nursing method. Based on this third rehabilitation nursing method, rehabilitation nursing planning is carried out for burn plastic surgery patients.
[0064] Furthermore, based on the region-corrected generative adversarial network (GAN) model, the GAN model is corrected using generation correction parameters calculated from the proportion of burn area in different burn locations, burn color, and the disorder of eschar texture. Since the burn conditions vary across different locations, the proportion of burn area effectively reflects the burn location. Meanwhile, the burn color and the degree of disorder in eschar texture significantly influence the determination of burn locations and subsequent rehabilitation care. Therefore, the calculation method for the generation correction parameters is as follows:
[0065]
[0066] In the formula, To generate correction parameters, To define the burn area of part x, To define the area of region x in the image, the images acquired in this invention are all three-channel image data. Therefore, the burn color is characterized using the maximum value of each channel to reduce the amount of data processing. n represents the number of image pixels. This represents the maximum grayscale value among the three channels of the i-th image pixel. This represents the average maximum value of the three gray channels where the darkest color is visible in the image indicating a burn. This indicates the disorder of eschar texture. The more disordered the eschar in the image, the more severe the burn, and the more closely the generated burn image data needs to resemble this type of image data. m represents the number of image pixels corresponding to the eschar in the image. x represents the probability of the gray value of the j-th eschar image pixel appearing, where x can be arbitrarily chosen as one of the face, hand, foot, or chest and back.
[0067] In existing burn and plastic surgery rehabilitation planning, a generative adversarial network model based on regional feature parameters to generate correction coefficients addresses the issue of insufficient model applicability due to patient privacy and low sample size in burn images. Furthermore, by considering both patient and medical staff circumstances and dividing the burn area into regions for targeted image generation, the generated data becomes more accurate, improving the accuracy of automated burn and plastic surgery rehabilitation planning, reducing actual deviations, and effectively assisting rehabilitation nurses in adaptive planning of rehabilitation methods.
[0068] This invention generates highly adaptable images of burn and plastic surgery sites and references physiological parameters related to patients and medical staff that influence rehabilitation planning. Furthermore, it combines image data generated by an adaptive data generation network with the status characteristics of patients and relevant medical staff to derive patient-friendly rehabilitation planning, obtaining a rehabilitation planning method that is more suitable for patients. This not only reduces the complexity of image data acquisition but also improves the satisfaction of burn patients with rehabilitation care.
[0069] Of course, it is understood that each embodiment of the present invention can achieve one of the effects on its own, and the combination of multiple embodiments of the present invention can achieve all the above effects. However, it is not required that each embodiment of the present invention achieve all the above advantages and effects, because each embodiment of the present invention can constitute a separate technical solution and make one or more contributions to the prior art.
[0070] For any module structures not specifically defined in this invention, the existing technical specifications shall prevail. The existing technical specifications mentioned in the foregoing background and specific embodiments sections are considered part of this invention and are used to understand the meaning of certain technical features or parameters. The scope of protection of this invention is determined by the actual contents of the claims.
Claims
1. A rehabilitation nursing plan method for patients with burns and plastic surgery, characterized in that, The method includes: The system collects image data of the burn sites of patients and divides them into regions. Based on the region-corrected generative adversarial network model, it generates regional burn image data and collects physiological parameters of burn patients and the state characteristics of nursing staff. Based on the processing of regional burn image data, physiological parameters, and state characteristics, the characteristics of plastic surgery rehabilitation planning are obtained, and the first rehabilitation nursing methods issued by doctors for the burn site at this time are collected. A rehabilitation nursing planning model is constructed based on the characteristics of the plastic surgery rehabilitation plan and the first rehabilitation nursing method. The second rehabilitation nursing method, derived from the image of the burn site before processing based on the rehabilitation nursing planning model, is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model.
2. The rehabilitation nursing planning method for patients with burns and plastic surgery as described in claim 1, characterized in that: The image data of the burn sites are divided into face, hands, feet, and chest and back based on the region.
3. A rehabilitation nursing planning method for patients with burns and plastic surgery as described in claim 1 or 2, characterized in that: The region-based modified generative adversarial network model uses generation correction parameters calculated based on the proportion of burn area in the divided parts, burn color, and disordered eschar texture to correct the generative adversarial network model.
4. A rehabilitation nursing plan method for patients with burns and plastic surgery as described in claim 3, characterized in that: The region-modified generative adversarial network model specifically uses generation modification parameters to modify the decision maker of the generative adversarial network model.
5. A rehabilitation nursing plan method for patients with burns and plastic surgery as described in claim 1 or 4, characterized in that: The physiological parameters include blood glucose, blood pressure, and weight, and the nursing staff's status characteristics include heart rate, respiratory rate, and extreme values of skin conductance signals.
6. A rehabilitation nursing plan method for patients with burns and plastic surgery as described in claim 5, characterized in that: The second rehabilitation nursing method, derived from the burn site image before generation based on the rehabilitation nursing planning model, is compared with the first rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model. Specifically, the optimized model is determined by calculating the range similarity of the values obtained from the burn site image before generation and the burn site image after generation processed by the rehabilitation nursing planning model.
7. A rehabilitation nursing plan method for patients with burns and plastic surgery as described in claim 6, characterized in that: The rehabilitation nursing planning model adopts a multi-classification model based on Fisher's criterion.
8. A rehabilitation nursing planning system for patients with burns and plastic surgery, the system implementing the method as described in claim 1, comprising a burn site image data collection and site segmentation module, a region-modified generative adversarial network model module, a patient and nursing staff physiological state feature collection module, a plastic surgery rehabilitation planning feature processing module, a rehabilitation nursing planning model construction and optimization module, and a rehabilitation nursing planning module, characterized in that: The burn site image data collection and site division module is used to collect image data of the patient's burn sites and divide them into face, hands, feet and chest and back according to the region. The region-corrected generative adversarial network model module generates a region-corrected generative adversarial network model, which uses generation correction parameters to correct the decision-maker. The patient and nursing staff physiological state characteristic collection module is used to collect physiological parameters of burn patients and state characteristics of nursing staff. The plastic surgery rehabilitation planning feature processing module: obtains plastic surgery rehabilitation planning features based on regional burn image data, physiological parameters and state features, and collects the first rehabilitation nursing method issued by the doctor for the burn site at this time; The rehabilitation nursing planning model construction and optimization module: Based on the plastic surgery rehabilitation planning characteristics and the first rehabilitation nursing method, a rehabilitation nursing planning model is constructed. Based on the second rehabilitation nursing method obtained from the burn site image before processing and generation of the rehabilitation nursing planning model, the first rehabilitation nursing method is compared with the second rehabilitation nursing method to obtain an optimized rehabilitation nursing planning model. The rehabilitation nursing planning module utilizes an optimized rehabilitation nursing planning model to process the plastic surgery rehabilitation planning features obtained from subsequent patient treatment, resulting in a third rehabilitation nursing method. Based on this third rehabilitation nursing method, rehabilitation nursing planning is carried out for burn plastic surgery patients.
9. A rehabilitation nursing planning system for patients with burns and plastic surgery as described in claim 8, characterized in that: Based on the region-corrected generative adversarial network model, the generative correction parameters calculated using the proportion of burn area in the divided parts, burn color, and disordered eschar texture are used to correct the generative adversarial network model.