Intelligent pressure regulating type lower limb lymphedema treatment method and device based on dynamic grading

Abstract

The invention discloses an intelligent pressure regulating type lower limb lymphedema treatment method and device based on dynamic grading. The method comprises the steps that standardized multi-modal data of pressure distribution, tissue impedance, surface temperature and limb volume of lower limbs of a patient are collected in real time; inputting the standardized multi-modal data into a trained DCNN edema grade evaluation model, and outputting the current edema grade of the patient; wherein the DCNN edema level evaluation model comprises an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a Region proposal network layer and an ROI pooling network layer; an initial pressure treatment scheme is selected from a preset pressure treatment model according to the current edema level, and an intelligent pneumatic pressurization module is controlled to conduct self-adaptive dynamic adjustment on the pressure of different parts of the lower limb to generate pressure adjusting parameters. Through intelligent pressure regulation control and real-time feedback, the intelligent level and the treatment effect of lower limb lymphedema treatment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical auxiliary treatment, and in particular to an intelligent pressure-regulating lower limb lymphedema treatment method based on dynamic grading, an apparatus and a computing device. Background Art

[0002] Lower limb lymphedema is limb swelling caused by abnormal accumulation of lymph fluid in the tissue spaces due to dysfunction of the lymphatic system. It not only affects the patient's limb function and appearance, but also causes complications such as pain, infection, and skin ulcers, seriously affecting the patient's quality of life.

[0003] Currently, treatments for lower limb lymphedema primarily include conservative and surgical therapies. Conservative treatments include manual lymphatic drainage, pressure therapy (elastic stockings, bandages), exercise therapy, and skin care, with pressure therapy being the most commonly used. Surgical treatments include lymphatic venous anastomosis, lymph node transplantation, and liposuction. Among these treatments, pressure therapy typically utilizes fixed-pressure elastic stockings or bandages. However, this pressure cannot be individually adjusted based on the patient's specific needs. Excessive pressure may not achieve therapeutic effects, while excessive pressure may cause discomfort or even injury. Manual lymphatic drainage requires specialized therapists, is time-consuming and labor-intensive, and its effectiveness is significantly affected by the therapist's experience. Surgical treatment carries certain risks and complications and is not suitable for all patients. Furthermore, traditional compression devices (such as air-pump-driven compression sleeves) typically only provide a preset fixed or cyclic pressure, making dynamic adjustment based on the patient's real-time physiological feedback difficult.

[0004] To solve the above problems, the present invention proposes an intelligent pressure-regulating lower limb lymphedema treatment method based on dynamic grading, which improves the intelligence level and treatment effect of lower limb lymphedema treatment through intelligent pressure regulation control and real-time feedback. Summary of the Invention

[0005] In view of the above problems, the present invention provides an intelligent pressure-regulating lower limb lymphedema treatment method, device, and computing equipment based on dynamic grading.

[0006] According to one aspect of the present invention, a method for treating lower limb lymphedema based on intelligent pressure regulation and dynamic grading is provided, comprising:

[0007] Using a multimodal sensor array to collect standardized multimodal data of pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limb in real time; the multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner;

[0008] Inputting the standardized multimodal data into a trained DCNN edema grade assessment model to output the patient's current edema grade; wherein the DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a Region proposal network layer, and a ROIpooling network layer;

[0009] According to the current edema level, an initial pressure treatment plan is selected from a preset pressure treatment model, and the intelligent pneumatic pressurization module is controlled to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

[0010] In an optional manner, the DCNN edema grade assessment model is trained using a dynamic weighted multi-task loss function, which is:

[0011]

[0012] Among them, L Dice is the Dice similarity coefficient loss; L CE is the cross entropy loss; α, β, γ are dynamic weight coefficients; I(t) is the bioimpedance distribution field at time t; D is the diffusion coefficient matrix; is the divergence operator; T is the duration of a single treatment cycle.

[0013] In an optional manner, the optimization objective function of the pressure gradient of the intelligent pneumatic pressurization module is:

[0014]

[0015] Where V(t) is the limb volume at time t; V target is the target volume; T is the treatment time; represents minimizing the objective function J by adjusting the pressure gradient ΔP; ΔP(t) is the pressure gradient at time t.

[0016] In an optional manner, the energy function for surface reconstruction using the 3D scanner limb volume measurement is:

[0017]

[0018] Where f is the implicit surface function; g is the original scan data; κ is the mean curvature; H0 is the target curvature; λ and μ are regularization parameters; Ω is the scan space; The surface boundary of the limb.

[0019] In an optional manner, the multimodal sensor array adopts a six-axis spatial layout scheme;

[0020] Among them, a 4×4 flexible pressure sensor matrix is ​​arranged 5 cm above the ankle joint, with a sampling frequency of ≥200 Hz;

[0021] Two sets of bioimpedance electrode pairs were placed on the front and back sides of the lower leg, and the four-electrode method was used for measurement with an excitation frequency range of 1kHz-1MHz.

[0022] The infrared temperature sensors are arranged in a spiral pattern with a spacing of 1 cm, covering the entire calf area;

[0023] The 3D scanner uses binocular structured light, with a baseline length of 150mm, a working distance of 300-500mm, and a single scanning accuracy of ≤0.5mm.

[0024] In an optional manner, the intelligent pneumatic pressurization module includes a three-level pressure regulation system consisting of a basic pressure layer, a dynamic compensation layer and a safety protection layer;

[0025] The basic pressure layer consists of 12 independent air chambers, each of which is equipped with a proportional pressure valve and a flow sensor;

[0026] The dynamic compensation layer adopts a microvalve array driven by a shape memory alloy, and the response time is ≤20ms;

[0027] The safety protection layer integrates a pressure relief valve and an overpressure alarm circuit. The pressure control algorithm uses a model predictive control framework with a prediction time domain of N = 15 and a control time domain of M = 5. The optimization objective function includes three dimensions: pressure tracking error, pressure change rate, and chamber volume change. Constraints include a maximum allowable pressure gradient of 5 kPa / s, a minimum safety pressure of 3 kPa, and a maximum operating frequency of 0.5 Hz.

[0028] When the tissue impedance change rate Z>0.2Ω / s is detected, the pressure hold mode is automatically triggered and the current pressure value is maintained until Z<0.05Ω / s. The volume change compensation during the pressure hold stage is achieved through a PID controller with a proportional coefficient Kp=0.8, an integral time Ti=30s, and a differential time Td=5s.

[0029] In an optional manner, the training process of the DCNN edema grade assessment model further includes:

[0030] Extracting shallow features of multimodal data through the initial DCNN network layer, the shallow features include local patterns of pressure distribution, spatial gradients of tissue impedance, texture information of surface temperature, and geometric contours of limb volume;

[0031] By enhancing the DCNN network layer and using residual connections and attention mechanisms, the network depth is deepened and the feature expression capability is enhanced;

[0032] The offset field network layer predicts the offset of each pixel to align the shape and position of the limbs to resolve measurement errors caused by changes in limb posture;

[0033] The RPN network layer is used to generate candidate regions. The RPN network layer generates multiple anchor boxes of different sizes and scales on the feature map through a sliding window and predicts the foreground / background probability and bounding box regression parameters of each anchor box.

[0034] The Region Proposal network layer selects candidate regions with higher confidence based on the output of the RPN as the input of the ROI pooling network layer;

[0035] The ROI pooling network layer maps candidate regions of different sizes to feature vectors of fixed size.

[0036] In an optional manner, the step of solving the optimization objective function of the pressure gradient further includes:

[0037] Divide the treatment time T into N time steps, each time step is Δt = T / N;

[0038] The volume change rate is approximated by the difference method, and the discrete expression of the volume change rate is obtained;

[0039] The objective function is converted into a discrete summation expression, which includes the square of the volume change rate at each time step, the square of the pressure gradient, and the square of the deviation between the volume at the final moment and the target volume;

[0040] The discretized objective function is solved by a quadratic programming solver to obtain the optimal pressure gradient sequence. The solution process includes the constraints of the maximum allowable pressure gradient, minimum safety pressure and maximum operating frequency of the pneumatic pressurization module.

[0041] According to another aspect of the present invention, there is provided an intelligent pressure-regulating lower limb lymphedema treatment device based on dynamic grading, comprising:

[0042] A multimodal data acquisition module for collecting standardized multimodal data of pressure distribution, tissue impedance, surface temperature, and limb volume of the patient's lower limb in real time using a multimodal sensor array; the multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner;

[0043] An edema grade assessment module is configured to input the standardized multimodal data into a trained DCNN edema grade assessment model and output the patient's current edema grade; wherein the DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROIpooling network layer;

[0044] An intelligent pressure control module is used to select an initial pressure treatment plan from a preset pressure treatment model according to the current edema level, and control the intelligent pneumatic pressurization module to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

[0045] According to another aspect of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;

[0046] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned intelligent pressure-regulating lower limb lymphedema treatment method based on dynamic grading.

[0047] According to the solution provided by the present invention, a multimodal sensor array is used to collect standardized multimodal data on pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limbs in real time. The multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner. The standardized multimodal data is input into a trained DCNN edema grade assessment model to output the patient's current edema grade. The DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROI pooling network layer. Based on the current edema grade, an initial pressure treatment plan is selected from a preset pressure treatment model, and an intelligent pneumatic pressurization module is controlled to adaptively and dynamically adjust the pressure at different parts of the lower limb to generate pressure adjustment parameters. The state space of the preset pressure treatment model uses the real-time pressure value feedback from the pressure sensor and the tissue impedance change rate feedback from the bioimpedance sensor, while the action space uses the pressure gradient, pressure rhythm, and duration of the pneumatic pressurization module. Through intelligent pressure regulation control and real-time feedback, the present invention improves the intelligent level and therapeutic effect of lower limb lymphedema treatment.

[0048] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0050] Figure 1 A schematic diagram showing a process of a method for treating lower limb lymphedema based on intelligent pressure regulation and dynamic grading according to an embodiment of the present invention is shown;

[0051] Figure 2 A schematic diagram of the framework of an intelligent pressure-regulating lower limb lymphedema treatment device based on dynamic grading according to an embodiment of the present invention is shown;

[0052] Figure 3 A schematic structural diagram of a computing device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0053] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0054] Figure 1 The flowchart of the intelligent pressure-regulating lower limb lymphedema treatment method based on dynamic classification according to an embodiment of the present invention is shown. Figure 1 As shown, the following steps are included:

[0055] Step S101, using a multimodal sensor array to collect standardized multimodal data of pressure distribution, tissue impedance, surface temperature and limb volume of the patient's lower limbs in real time; the multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor and a 3D scanner.

[0056] In this embodiment, pressure sensors quantify pressure distribution on the limb surface, reflecting changes in tissue tension caused by lymphatic fluid accumulation. Bioimpedance sensors measure interstitial fluid content, reflecting the degree of edema. Infrared temperature sensors detect inflammatory responses, reflecting the metabolic state of the edematous area. A 3D scanner measures changes in limb volume as an indicator of overall edema. All sensors are non-invasive or minimally invasive, avoiding additional pain or injury to the patient.

[0057] In an optional manner, the energy function for surface reconstruction using the 3D scanner limb volume measurement is:

[0058]

[0059] Where f is the implicit surface function; g is the original scan data; κ is the mean curvature; H0 is the target curvature; λ and μ are regularization parameters; Ω is the scan space; The surface boundary of the limb.

[0060] In this embodiment, high-precision limb surface reconstruction is achieved through the implicit surface function and the error term of the original scan data, reducing the noise and error in the scan data. The smoothness and continuity of the reconstructed surface are guaranteed by the constraint terms of the mean curvature and the target curvature, avoiding unnatural mutations and sharp angles in the surface reconstruction process. By constraining the limb surface boundary, the geometric shape of the limb is captured more accurately, and the accuracy of the reconstructed surface is improved. For example, a 3D scanner is used to scan the patient's calf, the original scan data is obtained, and an initial implicit surface function (such as an ellipsoid) is selected. The implicit surface function is iteratively updated using the gradient descent method, and the mean curvature and target curvature of the current surface are calculated in each iteration. The smoothness of the surface and the data fit are ensured by calculating the regularization term. The surface of the calf is reconstructed by the optimized implicit surface function.

[0061] In an optional manner, the multimodal sensor array adopts a six-axis spatial layout scheme;

[0062] Among them, a 4×4 flexible pressure sensor matrix is ​​arranged 5 cm above the ankle joint, with a sampling frequency of ≥200 Hz;

[0063] Two sets of bioimpedance electrode pairs were placed on the front and back sides of the lower leg, and the four-electrode method was used for measurement with an excitation frequency range of 1kHz-1MHz.

[0064] The infrared temperature sensors are arranged in a spiral pattern with a spacing of 1 cm, covering the entire calf area;

[0065] The 3D scanner uses binocular structured light, with a baseline length of 150mm, a working distance of 300-500mm, and a single scanning accuracy of ≤0.5mm.

[0066] In this embodiment, a multimodal sensor array with a six-axis spatial layout is used to simultaneously collect pressure, bioimpedance, temperature and 3D shape information, provide multi-angle data support, and improve the accuracy of lower limb lymphedema treatment.

[0067] Step S102: Input the standardized multimodal data into a trained DCNN edema grade assessment model to output the patient's current edema grade; wherein the DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a Region proposal network layer, and a ROI pooling network layer.

[0068] In this embodiment, the Offset Field network layer predicts pixel offsets to correct measurement errors caused by changes in limb posture. The initial DCNN network layer extracts shallow features, while the enhanced DCNN network layer deepens the network through residual connections and attention mechanisms, thereby improving feature expression capabilities. The RPN and ROIpooling network layers focus on key areas of the lower limbs (such as areas with significant edema), improving the ability to recognize key pathological features.

[0069] In an optional manner, the DCNN edema grade assessment model is trained using a dynamic weighted multi-task loss function, which is:

[0070]

[0071] Among them, L Dice is the Dice similarity coefficient loss; L CE is the cross entropy loss; α, β, γ are dynamic weight coefficients; I(t) is the bioimpedance distribution field at time t; D is the diffusion coefficient matrix; is the divergence operator; T is the duration of a single treatment cycle.

[0072] In this embodiment, the loss function combines Dice similarity coefficient loss, cross entropy loss, and a regularization term based on the bioimpedance distribution field. The model simultaneously learns the segmentation of edema areas and the spatiotemporal variation characteristics of bioimpedance. Among them, Dice similarity coefficient loss is specifically used to handle segmentation tasks, especially in the case of class imbalance, which can better optimize the segmentation of small targets. It is very important for the segmentation of edema areas because the size of edema areas can vary greatly. Cross entropy loss is a standard pixel-level classification loss function that encourages the model to accurately classify the edema level of each pixel.

[0073] In this embodiment, the training process of the DCNN edema grade assessment model further includes:

[0074] Extracting shallow features of multimodal data through the initial DCNN network layer, the shallow features include local patterns of pressure distribution, spatial gradients of tissue impedance, texture information of surface temperature, and geometric contours of limb volume;

[0075] By enhancing the DCNN network layer and using residual connections and attention mechanisms, the network depth is deepened and the feature expression capability is enhanced;

[0076] The offset field network layer predicts the offset of each pixel to align the shape and position of the limbs to resolve measurement errors caused by changes in limb posture;

[0077] The RPN network layer is used to generate candidate regions. The RPN network layer generates multiple anchor boxes of different sizes and scales on the feature map through a sliding window and predicts the foreground / background probability and bounding box regression parameters of each anchor box.

[0078] The Region Proposal network layer selects candidate regions with higher confidence based on the output of the RPN as the input of the ROI pooling network layer;

[0079] The ROI pooling network layer maps candidate regions of different sizes to feature vectors of fixed size.

[0080] Step S103: Select an initial pressure treatment plan from a preset pressure treatment model according to the current edema level, and control the intelligent pneumatic pressurization module to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

[0081] In this embodiment, pressure sensors and bioimpedance sensors are used to monitor changes in pressure and tissue impedance in the patient's lower extremities in real time. This data is then input into a pre-set pressure therapy model as a state space, enabling dynamic adjustment of the treatment plan. An intelligent pneumatic pressurization module precisely controls pressure at different locations in the lower extremities, enabling adaptive dynamic adjustment of the pressure gradient to ensure treatment comfort and safety. Dynamic adjustment of the pressure gradient, pressure rhythm, and duration can more effectively promote lymphatic return and thus reduce edema.

[0082] In an optional manner, the optimization objective function of the pressure gradient of the intelligent pneumatic pressurization module is:

[0083]

[0084] Where V(t) is the limb volume at time t; V target is the target volume; T is the treatment time; represents minimizing the objective function J by adjusting the pressure gradient ΔP; ΔP(t) is the pressure gradient at time t.

[0085] In this embodiment, the objective function The term directly reflects the square of the rate of change of limb volume. Minimizing this term means reducing limb volume as smoothly and efficiently as possible, thereby optimizing lymphatic return and avoiding the situation where only focusing on pressure itself ignores the actual treatment effect. ||ΔP(t)|| 2 The term is a penalty term for the pressure gradient, which can avoid drastic pressure changes and ensure the comfort and safety of treatment, because too rapid pressure changes can cause vasoconstriction or patient discomfort.

[0086] In an optional manner, the step of solving the optimization objective function of the pressure gradient further includes:

[0087] Divide the treatment time T into N time steps, each time step is Δt = T / N;

[0088] The volume change rate is approximated by the difference method, and the discrete expression of the volume change rate is obtained;

[0089] The objective function is converted into a discrete summation expression, which includes the square of the volume change rate at each time step, the square of the pressure gradient, and the square of the deviation between the volume at the final moment and the target volume;

[0090] The discretized objective function is solved by a quadratic programming solver to obtain the optimal pressure gradient sequence. The solution process includes the constraints of the maximum allowable pressure gradient, minimum safety pressure and maximum operating frequency of the pneumatic pressurization module.

[0091] In this embodiment, by discretizing the continuous treatment process into multiple time steps, the pressure gradient of each time step can be optimized, achieving more precise control of the pressurization process, thereby more effectively promoting lymphatic return and reducing edema.

[0092] Real-time optimization: The objective function is converted into a discrete form using the difference method to approximate the volume change rate, facilitating rapid solution using a quadratic programming solver. This allows for dynamic adjustment of the pressure gradient based on real-time data, better adapting to individual patient differences and treatment progress. Physical constraints of the pneumatic pressurization module (such as the maximum allowable pressure gradient, minimum safe pressure, and maximum operating frequency) are introduced during the optimization process to ensure that the optimized pressure gradient sequence achieves both therapeutic efficacy and safety during the treatment process.

[0093] In an optional manner, the intelligent pneumatic pressurization module includes a three-level pressure regulation system consisting of a basic pressure layer, a dynamic compensation layer and a safety protection layer;

[0094] The basic pressure layer consists of 12 independent air chambers, each of which is equipped with a proportional pressure valve and a flow sensor;

[0095] The dynamic compensation layer adopts a microvalve array driven by a shape memory alloy, and the response time is ≤20ms;

[0096] The safety protection layer integrates a pressure relief valve and an overpressure alarm circuit. The pressure control algorithm uses a model predictive control framework with a prediction time domain of N = 15 and a control time domain of M = 5. The optimization objective function includes three dimensions: pressure tracking error, pressure change rate, and chamber volume change. Constraints include a maximum allowable pressure gradient of 5 kPa / s, a minimum safety pressure of 3 kPa, and a maximum operating frequency of 0.5 Hz.

[0097] When the tissue impedance change rate Z>0.2Ω / s is detected, the pressure hold mode is automatically triggered and the current pressure value is maintained until Z<0.05Ω / s. The volume change compensation during the pressure hold stage is achieved through a PID controller with a proportional coefficient Kp=0.8, an integral time Ti=30s, and a differential time Td=5s.

[0098] In this embodiment, the base pressure layer provides a stable baseline pressure, the dynamic compensation layer rapidly responds to changes in tissue state, and the safety protection layer ensures the safety of the treatment process. The dynamic compensation layer utilizes a shape memory alloy (SMA)-driven microvalve array with a fast response speed (≤20ms), enabling rapid pressure adjustment based on changes in tissue impedance, better accommodating individual patient differences and treatment needs. The safety protection layer integrates a pressure release valve and overpressure alarm circuit, enabling timely pressure relief and alarming when pressure rises abnormally, preventing harm to the patient. The pressure control algorithm utilizes a model predictive control (MPC) framework, which predicts pressure changes over time and optimizes control in advance, resulting in smoother and more stable pressure control. The rate of change in tissue impedance, used as a trigger for pressure-hold mode, automatically identifies sudden changes in tissue state (possibly caused by overpressurization) and allows for timely adjustments to the treatment plan, ensuring both safety and effectiveness. During the pressure-hold phase, a PID controller compensates for volume changes, ensuring that effective compression of the limb is maintained by adjusting the air chamber volume while maintaining constant pressure, continuously promoting lymphatic return.

[0099] According to the solution provided by the present invention, a multimodal sensor array is used to collect standardized multimodal data on pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limbs in real time. The multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner. The standardized multimodal data is input into a trained DCNN edema grade assessment model to output the patient's current edema grade. The DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROI pooling network layer. Based on the current edema grade, an initial pressure treatment plan is selected from a preset pressure treatment model, and an intelligent pneumatic pressurization module is controlled to adaptively and dynamically adjust the pressure at different parts of the lower limb to generate pressure adjustment parameters. The state space of the preset pressure treatment model uses the real-time pressure value feedback from the pressure sensor and the tissue impedance change rate feedback from the bioimpedance sensor, while the action space uses the pressure gradient, pressure rhythm, and duration of the pneumatic pressurization module. Through intelligent pressure regulation control and real-time feedback, the present invention improves the intelligent level and therapeutic effect of lower limb lymphedema treatment.

[0100] Figure 2 The schematic diagram of the framework of the intelligent pressure-regulating lower limb lymphedema treatment device based on dynamic grading according to an embodiment of the present invention is shown. The intelligent pressure-regulating lower limb lymphedema treatment device based on dynamic grading comprises:

[0101] A multimodal data acquisition module 210 is configured to acquire standardized multimodal data of pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limb in real time using a multimodal sensor array comprising a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner;

[0102] The edema level assessment module 220 is configured to input the standardized multimodal data into a trained DCNN edema level assessment model and output the patient's current edema level; wherein the DCNN edema level assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROIpooling network layer;

[0103] The intelligent pressure control module 230 is used to select an initial pressure treatment plan from a preset pressure treatment model according to the current edema level, and control the intelligent pneumatic pressurization module to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

[0104] Figure 3 The schematic diagram of the structure of the computing device embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computing device.

[0105] like Figure 3 As shown, the computing device may include: a processor 302 , a communications interface 304 , a memory 306 , and a communication bus 308 .

[0106] Processor 302, communication interface 304, and memory 306 communicate with each other via communication bus 308. Communication interface 304 is used to communicate with other devices, such as client devices or other server network elements. Processor 302 is used to execute program 310, which specifically performs the steps described in the embodiment of the intelligent pressure-regulating treatment method for lower limb lymphedema based on dynamic grading.

[0107] Specifically, the program 310 may include program codes, which include computer operation instructions.

[0108] Processor 302 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a computing device may be processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0109] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0110] According to the solution provided by the present invention, a multimodal sensor array is used to collect standardized multimodal data on pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limbs in real time. The multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor, and a 3D scanner. The standardized multimodal data is input into a trained DCNN edema grade assessment model to output the patient's current edema grade. The DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROI pooling network layer. Based on the current edema grade, an initial pressure treatment plan is selected from a preset pressure treatment model, and an intelligent pneumatic pressurization module is controlled to adaptively and dynamically adjust the pressure at different parts of the lower limb to generate pressure adjustment parameters. The state space of the preset pressure treatment model uses the real-time pressure value feedback from the pressure sensor and the tissue impedance change rate feedback from the bioimpedance sensor, while the action space uses the pressure gradient, pressure rhythm, and duration of the pneumatic pressurization module. Through intelligent pressure regulation control and real-time feedback, the present invention improves the intelligent level and therapeutic effect of lower limb lymphedema treatment.

[0111] Those skilled in the art will appreciate that modules in the devices of the embodiments may be adaptively modified and deployed in one or more devices different from the embodiments. Modules, units, or components in the embodiments may be combined into a single module, unit, or component, and furthermore, they may be divided into multiple submodules, subunits, or subcomponents. All features disclosed in this specification (including the accompanying claims, abstract, and drawings), as well as all processes or units of any method or device disclosed therein, may be combined in any combination, except where at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that provides the same, equivalent, or similar purpose. Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims below, any of the claimed embodiments may be used in any combination. The present invention may be implemented using hardware comprising a number of different elements and using a suitably programmed computer. In a unit claim that lists several means, several of these means may be embodied by the same hardware item. Unless otherwise specified, the steps in the above embodiments should not be understood as limiting the order of execution.

Claims

1. A method for treating lower limb lymphedema based on intelligent pressure regulation and dynamic grading, characterized in that: include: A multimodal sensor array is used to collect standardized multimodal data of pressure distribution, tissue impedance, surface temperature, and limb volume of the patient's lower limbs in real time; The multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor and a 3D scanner; Inputting the standardized multimodal data into a trained DCNN edema grade assessment model to output the patient's current edema grade; wherein the DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a Region proposal network layer, and a ROI pooling network layer; According to the current edema level, an initial pressure treatment plan is selected from a preset pressure treatment model, and the intelligent pneumatic pressurization module is controlled to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

2. The intelligent pressure-regulating method for treating lower limb lymphedema based on dynamic grading according to claim 1 is characterized in that: The DCNN edema grade assessment model is trained using a dynamic weighted multi-task loss function, which is: Among them, L Dice is the Dice similarity coefficient loss; L CE is the cross entropy loss; α, β, γ are dynamic weight coefficients; I(t) is the bioimpedance distribution field at time t; D is the diffusion coefficient matrix; is the divergence operator; T is the duration of a single treatment cycle.

3. The intelligent pressure-regulating method for treating lower limb lymphedema based on dynamic grading according to claim 1 is characterized in that: The optimization objective function of the pressure gradient of the intelligent pneumatic pressurization module is: Where V(t) is the limb volume at time t; V target is the target volume; T is the treatment time; represents minimizing the objective function J by adjusting the pressure gradient ΔP; ΔP(t) is the pressure gradient at time t.

4. The method for treating lower limb lymphedema based on dynamic grading and intelligent pressure regulation according to claim 1 is characterized in that: The energy function for surface reconstruction of the 3D scanner limb volume measurement is: Where f is the implicit surface function; g is the original scan data; κ is the mean curvature; H0 is the target curvature; λ and μ are regularization parameters; Ω is the scan space; The surface boundary of the limb.

5. The method for treating lower limb lymphedema based on dynamic grading and intelligent pressure regulation according to claim 1 is characterized in that: The multimodal sensor array adopts a six-axis spatial layout scheme; Among them, a 4×4 flexible pressure sensor matrix is ​​arranged 5 cm above the ankle joint, with a sampling frequency of ≥200 Hz; Two sets of bioimpedance electrode pairs were placed on the front and back sides of the lower leg, and the four-electrode method was used for measurement with an excitation frequency range of 1kHz-1MHz. The infrared temperature sensors are arranged in a spiral pattern with a spacing of 1 cm, covering the entire calf area; The 3D scanner uses binocular structured light, with a baseline length of 150mm, a working distance of 300-500mm, and a single scanning accuracy of ≤0.5mm.

6. The method for treating lower limb lymphedema based on dynamic grading and intelligent pressure regulation according to claim 1, characterized in that: The intelligent pneumatic pressurization module includes a three-level pressure regulation system consisting of a basic pressure layer, a dynamic compensation layer and a safety protection layer; The basic pressure layer consists of 12 independent air chambers, each of which is equipped with a proportional pressure valve and a flow sensor; The dynamic compensation layer adopts a microvalve array driven by a shape memory alloy, and the response time is ≤20ms; The safety protection layer integrates a pressure relief valve and an overpressure alarm circuit. The pressure control algorithm uses a model predictive control framework with a prediction time domain of N = 15 and a control time domain of M = 5. The optimization objective function includes three dimensions: pressure tracking error, pressure change rate, and chamber volume change. Constraints include a maximum allowable pressure gradient of 5 kPa / s, a minimum safety pressure of 3 kPa, and a maximum operating frequency of 0.5 Hz. When the tissue impedance change rate Z>0.2Ω / s is detected, the pressure hold mode is automatically triggered and the current pressure value is maintained until Z<0.05Ω / s. The volume change compensation during the pressure hold stage is achieved through a PID controller with a proportional coefficient Kp=0.8, an integral time Ti=30s, and a differential time Td=5s.

7. The method for treating lower limb lymphedema based on dynamic grading and intelligent pressure regulation according to claim 1 is characterized in that: The training process of the DCNN edema grade assessment model further includes: Extracting shallow features of multimodal data through the initial DCNN network layer, the shallow features include local patterns of pressure distribution, spatial gradients of tissue impedance, texture information of surface temperature, and geometric contours of limb volume; By enhancing the DCNN network layer and using residual connections and attention mechanisms, the network depth is deepened and the feature expression capability is enhanced; The offset field network layer predicts the offset of each pixel to align the shape and position of the limbs to resolve measurement errors caused by changes in limb posture; The RPN network layer is used to generate candidate regions. The RPN network layer generates multiple anchor boxes of different sizes and scales on the feature map through a sliding window and predicts the foreground / background probability and bounding box regression parameters of each anchor box. The Region Proposal network layer selects candidate regions with higher confidence based on the output of the RPN as the input of the ROI pooling network layer; The ROI pooling network layer maps candidate regions of different sizes to feature vectors of fixed size.

8. The intelligent pressure-regulating method for treating lower limb lymphedema based on dynamic grading according to claim 3 is characterized in that: The step of solving the optimization objective function of the pressure gradient further includes: Divide the treatment time T into N time steps, each time step is Δt = T / N; The volume change rate is approximated by the difference method, and the discrete expression of the volume change rate is obtained; The objective function is converted into a discrete summation expression, which includes the square of the volume change rate at each time step, the square of the pressure gradient, and the square of the deviation between the volume at the final moment and the target volume; The discretized objective function is solved by a quadratic programming solver to obtain the optimal pressure gradient sequence. The solution process includes the constraints of the maximum allowable pressure gradient, minimum safety pressure and maximum operating frequency of the pneumatic pressurization module.

9. An intelligent pressure-regulating lower limb lymphedema treatment device based on dynamic grading, characterized in that: include: a multimodal data acquisition module for acquiring standardized multimodal data of pressure distribution, tissue impedance, surface temperature, and limb volume of a patient's lower limbs in real time using a multimodal sensor array; The multimodal sensor array includes a flexible pressure sensor, a bioimpedance sensor, an infrared temperature sensor and a 3D scanner; An edema grade assessment module is configured to input the standardized multimodal data into a trained DCNN edema grade assessment model and output the patient's current edema grade; wherein the DCNN edema grade assessment model includes an initial DCNN network layer, an enhanced DCNN network layer, an offset field network layer, an RPN network layer, a region proposal network layer, and a ROIpooling network layer; An intelligent pressure control module is used to select an initial pressure treatment plan from a preset pressure treatment model according to the current edema level, and control the intelligent pneumatic pressurization module to adaptively and dynamically adjust the pressure of different parts of the lower limbs to generate pressure regulation parameters; wherein, the state space of the preset pressure treatment model adopts the real-time pressure value fed back by the pressure sensor and the tissue impedance change rate fed back by the bioimpedance sensor; the action space adopts the pressure gradient, pressure rhythm and duration of the pneumatic pressurization module.

10. A computing device comprising: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned intelligent pressure-regulating lower limb lymphedema treatment method based on dynamic grading.

Images