Optimal burn treatment massage path planning method based on graph theory and SAC algorithm

Abstract

The invention relates to a burn treatment path planning technology, and discloses a burn treatment optimal massage path planning method based on a graph theory and an SAC algorithm. The system comprises a data acquisition and initialization module, a path modeling module, a reinforcement learning module, an optimization and generation module and a closed-loop management module. The data acquisition and initialization module is used for generating accurate point cloud data through a binocular camera in combination with an IMU (Inertial Measurement Unit), and calculating burn average degree and uniformity to provide a basis for subsequent processing; the path modeling module adopts a graph theory method to model a burn area into nodes and edges, and multi-objective optimization is carried out by integrating treatment efficiency and patient comfort; the reinforcement learning module designs a state space, an action space and a reward function based on an SAC algorithm, and realizes optimization of a path and a strength strategy; the optimization and generation module generates a final treatment path through initial exploration and local optimization, and introduces an adaptive force adjustment function to prevent secondary injury; the closed-loop management module supports off-line learning and real-time feedback, and ensures high efficiency and individuation of a therapeutic scheme through dynamic parameter adjustment and report generation. According to the method, the accuracy and efficiency of burn treatment are effectively improved, and the method has wide clinical application potential.

Description

Technical field:

[0001] The present invention relates to the field of medical technology and intelligent device control, and specifically to an optimal massage path planning method designed for burn treatment through graph theory and SAC (Soft Actor-Critic) algorithm, aiming to achieve accurate treatment path and strength control. Background technology:

[0002] In burn treatment, reasonable massage path and intensity planning are the key to promoting recovery. However, the current manual operation mainly relies on the doctor's experience, which is easily affected by subjectivity, resulting in unstable treatment effect and low efficiency. In addition, the characteristics of the burn area of ​​each patient are different, and a single treatment method is difficult to meet personalized needs.

[0003] The rapid development of artificial intelligence technology has brought new opportunities for burn treatment. In recent years, reinforcement learning algorithms have been widely used in path planning problems, but traditional methods often have slow convergence speed and poor strategy adaptability. The SAC algorithm has efficient exploration capabilities and stable strategy optimization characteristics, which is very suitable for solving complex path planning problems. Combining graph theory to model the burn area as a network structure of nodes and edges can provide a more efficient calculation method for treatment path planning. Summary of the invention:

[0004] The present invention aims to provide an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm to solve the problems of low treatment efficiency and inaccurate path planning in the prior art.

[0005] In order to achieve the above objectives, the present invention designs a path planning method comprising the following modules: a data acquisition and initialization module, a path modeling module, a reinforcement learning module, an optimization and generation module and a closed-loop management module.

[0006] The data acquisition and initialization module is responsible for collecting the three-dimensional point cloud data of the burn area and dividing the area into multiple treatment units through the image segmentation algorithm. Each unit records the degree of burn and the intensity required, providing basic data for subsequent optimization. At the same time, the module can integrate historical treatment data and establish a dynamic data update mechanism.

[0007] The path modeling module uses graph theory to model each treatment unit as a node. Node attributes include burn severity and intensity requirements. The edge weights between nodes are determined by burn severity and physical distance, and nodes with high weights are processed first. At the same time, the model considers multi-objective optimization methods and combines treatment efficiency and patient comfort into path planning.

[0008] The reinforcement learning module adopts the SAC algorithm, which specifically includes the design of state space, action space and reward function. The state space covers the burn grade, massage intensity and current path state; the action space involves path selection and intensity adjustment. The reward function integrates treatment uniformity, efficiency and patient comfort to achieve the best treatment effect.

[0009] The optimization and generation module outputs the optimal path and force strategy through reinforcement learning training. In the initial exploration stage, the model focuses on the exploration of the global path to find the potential best solution; in the later optimization stage, the strategy focuses on improving local accuracy to ensure that key areas are treated first. Furthermore, the module introduces an adaptive force adjustment function to adjust the massage force according to real-time data to prevent secondary damage to sensitive areas.

[0010] The closed-loop management module establishes an offline learning and real-time feedback system on the server side. It optimizes the learning strategy through large-scale offline simulation and deploys the optimization results to the actual device for execution. This module can continuously collect device operation data and dynamically adjust model parameters.

[0011] Furthermore, the data acquisition and initialization module generates accurate point cloud data through a binocular camera combined with an IMU, and performs image segmentation processing on the burn area. On this basis, the module calculates the average degree of burns by weighted average of the burn degree of each treatment unit, and calculates the uniformity of burns by square difference. Combined with the actual treatment needs of the patient, the initial treatment data is generated. The module supports real-time updating of the collected data to dynamically adapt to the patient's recovery progress.

[0012] Furthermore, the pathway modeling module supports a dynamic weight adjustment mechanism to update the priority of pathway planning based on real-time monitoring of treatment progress to ensure that the treatment strategy matches the patient's needs. Furthermore, the module can also integrate the patient's specific personalized preferences to generate a pathway planning solution that better meets the patient's needs.

[0013] Furthermore, in order to enhance the stability and adaptability of the strategy, the reinforcement learning module introduces an entropy regularization mechanism. By adding entropy rewards to the strategy optimization, the exploration efficiency is improved to avoid the suboptimal solution problem caused by premature convergence. Entropy regularization can effectively motivate the agent to explore more possible paths in the early stage and provide a more comprehensive decision-making basis for the treatment plan. In addition, through multiple rounds of training iterations, the module can dynamically adjust the strategy parameters to achieve more accurate adaptation to complex scenarios.

[0014] Furthermore, the reinforcement learning module supports an online learning mechanism to obtain patient feedback data in real time during the treatment process, including the immediate response of the treatment area and the patient's subjective comfort evaluation. Based on these feedbacks, the module can quickly optimize path selection and intensity adjustment, making the treatment process more intelligent and personalized. This dynamic update capability ensures that the treatment strategy can be continuously optimized as the patient's recovery status changes, thereby improving the overall treatment effect and patient experience.

[0015] Furthermore, in order to meet personalized needs, the optimization and generation module also uses historical data and simulation scenarios to enhance the pertinence of model training, so that it can adapt to the complex burn conditions of different patients. At the same time, the module supports multi-device collaborative optimization, and plans paths for multiple burn areas at the same time to improve overall treatment efficiency. Furthermore, the closed-loop management module not only supports periodic strategy updates, but also has a dynamic adaptation function, which can intelligently optimize device parameters based on long-term monitoring data, thereby ensuring that the treatment plan always remains efficient, accurate and personalized. At the same time, the module can also automatically generate detailed treatment reports, including key data, trend analysis and effect evaluation during the treatment process, to provide comprehensive data support for the medical team, assisting them in formulating more scientific and accurate follow-up treatment plans, and further improving the overall treatment quality and patient satisfaction. Beneficial effects:

[0017] The present invention significantly improves the efficiency and accuracy of burn treatment path planning by innovatively combining graph theory and SAC algorithm, breaks through the limitation of traditional treatment relying on doctor's experience, and realizes high automation and personalization of treatment plan.

[0018] The present invention can greatly improve treatment efficiency, reduce redundant operations and resource waste, make the treatment process more streamlined and efficient, and improve the utilization rate of medical resources.

[0019] The present invention ensures timely and effective treatment of lesions by giving priority to key areas, thereby significantly improving the patient's recovery effect, shortening the rehabilitation period, and improving the overall treatment quality and patient satisfaction.

[0020] The present invention realizes intelligent and automated treatment path planning, and significantly reduces the errors caused by manual intervention through precise calculation and dynamic adjustment of treatment plans, making treatment more accurate and controllable.

[0021] The present invention can provide personalized treatment plans according to the specific conditions of patients, fully consider the special needs and individual differences of patients, ensure the pertinence and scientific nature of treatment, and provide patients with better medical services.

[0022] The present invention is not only widely applicable in the field of burn treatment, but can also be extended to other medical application scenarios that require precise path planning and force control, such as surgical navigation, precision injection, etc. It provides a new technical solution for the research and development of intelligent medical equipment and promotes the development and popularization of intelligent medical technology. Description of the drawings:

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0024] Figure 1 A schematic diagram of the system structure of an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm of the present invention;

[0025] Figure 2 This is a data collection and initialization flow chart of an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm of the present invention;

[0026] Figure 3 A flow chart of path modeling of an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm of the present invention;

[0027] Figure 4 This is a design diagram of a reinforcement learning module for an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm of the present invention. Specific implementation method:

[0028] The present invention provides an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm, including a data acquisition and initialization module, a path modeling module, a reinforcement learning module, an optimization and generation module and a closed-loop management module. In order to clearly and completely describe the technical solution in the embodiment of the present invention, the following is a detailed description in conjunction with the drawings in the embodiment.

[0029] It should be noted that the described embodiments are only a part of the embodiments of the present invention, not all of them. Without doing creative work, all other embodiments that can be implemented by ordinary technicians in this field based on the content of the present invention are within the protection scope of the present invention.

[0030] Example embodiments will be described more fully below with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, the purpose of providing these embodiments is to make the present disclosure thorough and complete and to enable those skilled in the art to fully understand the scope of the present disclosure.

[0031] In the absence of conflict, the various embodiments of the present disclosure and the various features therein may be combined with each other.

[0032] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0033] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements and / or components is specified, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof is not excluded.

[0034] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless explicitly defined as such herein.

[0035] See also Figure 1 The present invention relates to a method for planning the optimal massage path for burn treatment based on graph theory and SAC algorithm. This method realizes personalized planning and dynamic optimization of the burn treatment path through the collaborative work of multiple modules. The specific process is as follows:

[0036] Step 1: The data acquisition and initialization module uses a binocular camera combined with an IMU to obtain 3D point cloud data of the burn area, and uses an image segmentation algorithm to divide the area into multiple treatment units. This module calculates the average degree and uniformity of the burn, generates initial treatment data, and supports real-time updates to dynamically adapt to the patient's recovery progress, providing an accurate data basis for subsequent path planning.

[0037] Step 2: The path modeling module uses graph theory to model the treatment units as nodes, and uses burn severity and physical distance as edge weights. Through a multi-objective optimization method, the optimization objectives suitable for the reinforcement learning module are constructed by comprehensively considering treatment efficiency and patient comfort.

[0038] Step 3: The reinforcement learning module introduces the SAC algorithm to design the path planning strategy. The module defines the state space, action space and reward function. The state space covers the burn grade, massage intensity and current path state; the action space includes path selection and intensity adjustment; the reward function integrates treatment uniformity, efficiency and patient comfort to ensure that the path planning meets the treatment needs.

[0039] Step 4: The optimization and generation module generates the final treatment path and intensity adjustment plan based on the strategy output by the reinforcement learning module. In the early stage, the global path exploration is used to find the potential optimal solution, and in the later stage, it focuses on local optimization, giving priority to treating key areas, and adjusting the intensity in real time to avoid secondary damage to sensitive areas.

[0040] Step 5: The closed-loop management module provides offline learning and real-time feedback functions. The module optimizes the strategy through large-scale simulation and deploys it to actual equipment, collects treatment data in real time to update model parameters. At the same time, the module generates detailed treatment reports to provide decision support for the medical team and ensure the dynamic adaptability and intelligence level of the treatment plan.

[0041] The multi-module collaborative structure of the present invention realizes the automation and personalization of burn treatment path planning, significantly improves treatment efficiency and accuracy, and reduces dependence on manual experience. The real-time optimization capability of the closed-loop management module further enhances the adaptability and robustness of the solution, providing reliable technical support for complex burn treatment.

[0042] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for planning an optimal massage path for burn treatment, characterized in that: include: Data collection and initialization module, path modeling module, reinforcement learning module, optimization and generation module and closed-loop management module; The data acquisition and initialization module is used to generate three-dimensional point cloud data of the burn area through a binocular camera combined with an IMU, perform image segmentation on the burn area, calculate the average degree and uniformity of the burn, generate initial treatment data and support real-time updates; The path modeling module models the treatment units as nodes based on graph theory, establishes edge weights between nodes according to burn severity and physical distance, and uses a multi-objective optimization method to comprehensively consider treatment efficiency and patient comfort to generate a path planning model; The reinforcement learning module designs the state space, action space and reward function based on the SAC algorithm, optimizes the path selection and intensity adjustment strategy through reinforcement learning, and realizes the intelligent planning of treatment path and intensity; The optimization and generation module is responsible for generating the final treatment path based on the strategy output by reinforcement learning, combining the initial global path exploration with the local optimization of key areas, and preventing secondary damage to sensitive areas through adaptive force adjustment; The closed-loop management module dynamically adjusts model parameters through offline learning and real-time feedback mechanisms, and generates treatment reports to provide an optimization basis for subsequent treatment.

2. The method for planning the optimal massage path for burn treatment according to claim 1, characterized in that: The data acquisition and initialization module includes: a binocular camera unit, an IMU unit and an image processing unit; The binocular camera unit and the IMU unit work together to generate accurate three-dimensional point cloud data; The image processing unit divides the burn area into a plurality of treatment units by using an image segmentation algorithm, and calculates the average degree and uniformity of the burns of the treatment units.

3. The method for planning the optimal massage path for burn treatment according to claim 1, characterized in that: The path modeling module further includes: a node construction unit and a weight calculation unit; The node construction unit models each treatment unit as a node according to the segmentation result of the burn area; The weight calculation unit calculates the edge weights between nodes according to the physical distance between treatment units and the difference in burn severity.

4. The method for planning the optimal massage path for burn treatment according to claim 1, characterized in that: The reinforcement learning module includes: a state space design unit, an action space design unit and a reward function design unit; The state space design unit is used to define the burn grade, massage intensity requirement and path state of the treatment unit; The action space design unit is used for designing path selection and force adjustment strategies; The reward function design unit calculates the reward value by comprehensively considering treatment uniformity, efficiency and patient comfort.

5. The method for planning the optimal massage path for burn treatment according to claim 1, characterized in that: The optimization and generation module includes: a global exploration unit, a local optimization unit and a strength adjustment unit; The global exploration unit explores the optimal solution of the treatment path in the initial stage; The local optimization unit improves the path accuracy in key areas; The intensity adjustment unit dynamically adjusts the treatment intensity based on real-time data to prevent damage to sensitive areas.

6. The method for planning the optimal massage path for burn treatment according to claim 1, characterized in that: The closed-loop management module includes: an offline learning unit, a real-time feedback unit, a dynamic optimization unit and a report generation unit; The offline learning unit generates an initial path planning strategy through large-scale reinforcement learning training; The real-time feedback unit is used to collect the equipment operation status and patient feedback data; The dynamic optimization unit adjusts model parameters according to real-time data; The report generation unit automatically generates a detailed report containing key treatment data, trend analysis and effect evaluation.

7. The method for planning the optimal massage path for burn treatment according to any one of claims 1 to 6, characterized in that: The closed-loop management module supports periodic strategy updates and optimizes device parameters in combination with long-term monitoring data to adapt to the patient's recovery progress.

8. An implementation step of an optimal massage path planning method for burn treatment based on graph theory and SAC algorithm, comprising the following steps: S1. Collect 3D point cloud data of the burn area through binocular camera combined with IMU to record the degree of burn and patient treatment needs; S2. Processing the burn area using an image segmentation algorithm, dividing it into multiple treatment units, and calculating the average degree and uniformity of the burn in each unit; S3, construct a path model based on graph theory, model the treatment units as nodes, and the edge weights between nodes are determined by the severity of burns and physical distance; S4. Use the SAC algorithm to design the state space, action space and reward function, and optimize the path selection and intensity adjustment strategy through reinforcement learning; S5. Combine global exploration and local optimization in the optimization and generation module to generate the final treatment path, and dynamically adjust the intensity to avoid secondary damage; S6. Collect real-time data and patient feedback through the closed-loop management module, dynamically optimize treatment pathways and strategies, and generate reports containing treatment effect analysis.

Images