Method and device for optimizing non-invasive physical intervention parameters of shoulder involved pain after laparoscopic surgery

Abstract

The invention relates to a method and a device for optimizing non-invasive physical intervention parameters of shoulder involved pain after a laparoscopic surgery. The method comprises the following steps: acquiring individualized data of a patient; constructing a personalized pain-involved neural pathway mapping model according to the personalized data; according to the three-dimensional nerve stimulation target point candidate areas and the priority ranking thereof, using a finite element analysis algorithm to optimize the forms, positions and pressure distribution of adjustable support belts and support pads in the wearable shoulder support structure; according to the probability incidence relation, the shapes, positions and pressure distribution of the supporting belt and the supporting pad can be adjusted to dynamically explore the optimal stimulation parameter combination. According to the invention, a novel highly-individualized, accurate and noninvasive physical intervention method is realized, and the treatment effect on the shoulder involved pain after the laparoscopic surgery and the patient experience are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of non-invasive physical intervention technology, specifically to a method and device for optimizing parameters of non-invasive physical intervention for referred pain in the shoulder after laparoscopic surgery, and computer equipment. Background Technology

[0002] Laparoscopic surgery is widely used in clinical practice due to its minimally invasive nature and rapid recovery. However, postoperative complications remain a major concern for patients. Among these, referred pain in the shoulder after laparoscopic surgery is a common complication with a high incidence, causing significant pain and severely impacting postoperative recovery and quality of life. It typically manifests as dull or radiating pain in the shoulder or neck, primarily related to the stimulation of the diaphragm by the infusion of carbon dioxide gas into the abdominal cavity during surgery. This stimulation excites the phrenic nerve, which then transmits pain signals through the spinal cord to the cerebral cortex, where the pain is mistakenly identified as originating from the shoulder.

[0003] In existing technologies, interventions for referred shoulder pain after laparoscopic surgery mainly include pharmacological and non-pharmacological treatments. Pharmacological treatment primarily relieves pain through analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs), and local anesthesia. However, pharmacological treatment has side effects (such as gastrointestinal reactions and liver and kidney damage), and long-term use can lead to dependence or drug tolerance. Non-pharmacological treatments include physical therapy, nerve blocks, and postural interventions. While physical therapy, such as transcutaneous electrical nerve stimulation (TENS), can relieve pain to some extent, its stimulation parameters are usually set based on experience, lacking individualized customization, resulting in significant differences in effectiveness and potential stimulation of non-target areas. Nerve blocks, although providing rapid pain relief, are invasive procedures with risks of infection and bleeding. Postural interventions, while reducing diaphragmatic traction to some extent, have limited effectiveness and cannot completely eliminate pain.

[0004] To address the aforementioned issues, this invention proposes a method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery, aiming to improve the treatment efficacy and patient experience for shoulder referred pain after laparoscopic surgery. Summary of the Invention

[0005] In view of the above problems, the present invention provides a method and device for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery, as well as a computer device.

[0006] According to one aspect of the present invention, a method for optimizing non-invasive physical intervention parameters for referred shoulder pain after laparoscopic surgery is provided, comprising:

[0007] Acquire individualized patient data, including medical imaging data, neurophysiological data, pain scale data, and physiological indicator data;

[0008] A personalized referred pain neural pathway mapping model is constructed based on the individualized data. The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them. Furthermore, a probabilistic correlation is established between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions.

[0009] Based on the candidate regions of the three-dimensional neural stimulation targets and their priority ranking, the morphology, position and pressure distribution of the adjustable support straps and support pads in the wearable shoulder support structure are optimized using the finite element analysis algorithm.

[0010] The optimal combination of stimulation parameters is dynamically explored based on the probabilistic correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

[0011] In one alternative approach, the medical imaging data includes MRI, CT, or ultrasound images of the shoulder region after laparoscopic surgery;

[0012] The neurophysiological data includes surface electromyography, event-related potentials, or evoked potential data.

[0013] The pain scale data includes the Visual Analogue Scale (VAS) and the Numerical Rating Scale (NRS);

[0014] The physiological data include the patient's heart rate variability (HRV), skin conductance (EDA), and respiratory rate.

[0015] In one alternative approach, the personalized referred pain neural pathway mapping model, which generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them, further includes:

[0016] The medical imaging data is segmented at the voxel level to extract key anatomical structures, including potential sources of referred pain, related peripheral nerves, and related brain regions. The related peripheral nerves include the phrenic nerve, branches of the vagus nerve, the brachial plexus, and related spinal nerve roots. The related brain regions include the primary sensory cortex, the secondary sensory cortex, the anterior cingulate cortex, the insula, and the thalamus.

[0017] Each potential pain signal pathway is obtained by simulating the triangular mesh model or voxel label map of the key anatomical structure based on the fiber tracing algorithm of diffusion tensor imaging. The pathway is a pain signal transmission path that starts from the potential referred pain source, transmits along the relevant peripheral nerve to the spinal cord gray matter, and then reaches the relevant brain region through the ascending pathway of the spinothalamic tract.

[0018] Calculate the efficiency weight map of each potential pain signaling pathway in pain signal transmission, wherein the efficiency weight map includes the length, diameter, degree of myelination, number of synapses, and excitability threshold of neurons along the pathway;

[0019] Identify the primary pain signaling pathway that is most efficient and best matches the patient’s current pain perception pattern from the efficiency weighting graph;

[0020] Key ganglia, nerve fiber tract segments, and projection regions in the cerebral cortex of the main pain signaling pathway are mapped as initial three-dimensional neural stimulation target candidate regions; the initial three-dimensional neural stimulation target candidate regions are prioritized and stimulation weights are assigned to each target.

[0021] In one alternative approach, establishing probabilistic associations between the patient's pain perception patterns, physiological responses, neuroanatomical structures, and cerebral cortex regions further includes:

[0022] The pain scale data is used to quantify the patient's current pain intensity, pain type, and pain distribution as the initial input for the pain perception pattern; the physiological index data is used to analyze the patient's autonomic nervous system activity, stress level, and breathing pattern as the initial input for the physiological response; and the neuroelectrophysiological data is used to analyze the latency, amplitude, and spatial distribution of the response of different regions of the cerebral cortex to pain stimuli as the initial input for the activity of the cerebral cortex regions.

[0023] The initial input of the pain perception pattern, the initial input of the physiological response, the initial input of the activity of the cerebral cortex, the key ganglia, nerve fiber tract segments and projection areas in the cerebral cortex of the main pain signal pathway are correlated and analyzed to obtain a multimodal correlation feature set.

[0024] A nonlinear probabilistic association model is established based on the multimodal association feature set to connect the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region.

[0025] In one alternative approach, optimizing the shape, position, and pressure distribution of the adjustable support straps and support pads in a wearable shoulder support structure using finite element analysis algorithms further includes:

[0026] A three-dimensional surface model of the patient's shoulder and torso is generated based on the medical imaging data, serving as the geometric basis for the design of the wearable shoulder support structure. The initial three-dimensional neural stimulation target candidate regions and their priorities are sorted and mapped onto the three-dimensional surface model of the patient's shoulder and torso, serving as the target regions and priorities for which the support structure needs to apply pressure. An initial CAD model of the wearable shoulder support structure is established, and its material properties and contact properties are defined.

[0027] The three-dimensional surface model of the patient's shoulder and trunk, the mapped three-dimensional neural stimulation target candidate region and its priority, and the initial CAD model of the wearable shoulder support structure are imported into the finite element analysis software to simulate the pressure distribution of the support structure and the patient's skin contact surface under different support band tensions, support pad positions and shapes, and calculate the average pressure, pressure gradient and stress-strain relationship with the surrounding tissues of each stimulation target region.

[0028] Based on the priority of the candidate stimulation target regions, a preset effective stimulation pressure range is achieved in the priority target regions, while minimizing the pressure in non-target regions and maximizing patient comfort; the shape, position, and material parameters of the adjustable support strap and support pad are adjusted according to the topology optimization iterative algorithm, and the optimized shape, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are output.

[0029] In one alternative approach, adjusting the shape, position, and material parameters of the adjustable support strip and support pad according to a topology optimization iterative algorithm further includes:

[0030] Construct an objective optimization function that aims to maximize the pressure efficiency of the target area and minimize the overall structural flexibility; wherein the objective optimization function is subject to equilibrium equation constraints, volume constraints, and manufacturing constraints;

[0031] Solving the objective optimization function outputs the optimal material distribution map to determine the optimal shape, position, and material parameter combination of the adjustable support strip and support pad.

[0032] In one alternative approach, the objective optimization function is:

[0033]

[0034] in, To support the relative density of each finite element in the structure; The number of target regions; Let be the target pressure value for the i-th target area; The average pressure in the region of the i-th target point is calculated through finite element analysis. This is the tolerance parameter; These are the weighting coefficients; This is the global displacement vector; This is the global stiffness matrix;

[0035] The equilibrium equations are constrained as follows:

[0036]

[0037] in, This is a load vector function originating from the tension of the support belt;

[0038] The volume constraint is:

[0039]

[0040] in, Let x be the volume of the x-th finite element; The total number of finite element elements in the design domain; For the maximum permissible material volume, , It is the volume fraction. The total volume of the design domain;

[0041] The manufacturing constraints are:

[0042]

[0043] in, The relative density of the x-th finite element after smoothing; Let be the volume of the j-th finite element; Let be the original relative density of the j-th finite element; For linear weighting functions, , To control the minimum feature size parameter, Center point With the center point The Euclidean distance between them.

[0044] In an alternative approach, establishing a nonlinear probabilistic association model between the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region based on the multimodal association feature set further includes:

[0045] The multimodal association feature set is defined as a high-dimensional feature vector F, wherein the high-dimensional feature vector contains d features extracted from pain perception patterns, physiological responses, cerebral cortex activity, and neuroanatomical structures.

[0046] Construct a probabilistic graphical model based on conditional random fields, wherein the Gibbs distribution of the probabilistic graphical model is:

[0047]

[0048] in, These are implicit state variables, representing the activation state of neuroanatomical structures and cerebral cortical regions; It is the partition function; Let be the set of all cliques in the graph; Let be the potential function defined on the clique c; ; This is a univariate feature function used to model the relationship between the activation probability of a single neural node and the multimodal feature F; It is a binary feature function used to model the coactivation relationship between peripheral nerves and cortical regions; , represents the model parameters to be learned; i and j represent the i-th and j-th neural nodes, respectively;

[0049] The optimal parameters are obtained by maximizing the log-likelihood function of the model on the training data, and then iteratively updated using the stochastic gradient ascent method to obtain a nonlinear probabilistic correlation model.

[0050] According to another aspect of the present invention, a non-invasive physical intervention parameter optimization device for shoulder referred pain after laparoscopic surgery is provided, comprising:

[0051] The individualized data acquisition module is used to acquire the patient's individualized data, which includes medical imaging data, neuroelectrophysiological data, pain scale data, and physiological indicator data.

[0052] The personalized referred pain neural pathway mapping module is used to construct a personalized referred pain neural pathway mapping model based on the individualized data. The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them; and establishes a probabilistic correlation between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions.

[0053] The wearable shoulder support structure optimization module is used to optimize the shape, position and pressure distribution of the adjustable support band and support pad in the wearable shoulder support structure based on the candidate regions of the three-dimensional neural stimulation target and their priority ranking, using the finite element analysis algorithm.

[0054] The optimal stimulation parameter optimization module is used to dynamically explore the optimal combination of stimulation parameters based on the probability correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein, the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

[0055] According to another aspect of the present invention, a computer device is provided, 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 through the communication bus;

[0056] The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the above-described method for optimizing parameters of non-invasive physical intervention for shoulder referred pain after laparoscopic surgery.

[0057] According to the solution provided by the present invention, individualized patient data is acquired, including medical imaging data, neurophysiological data, pain scale data, and physiological index data; a personalized referred pain neural pathway mapping model is constructed based on the individualized data, wherein the personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on the imaging data and prioritizes them; and a probabilistic correlation is established between the patient's pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region; based on the three-dimensional neural stimulation target candidate region and its priority ranking, the morphology, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are optimized using a finite element analysis algorithm; and the optimal combination of stimulation parameters is dynamically explored based on the probabilistic correlation, the morphology, position, and pressure distribution of the adjustable support strap and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode. This invention utilizes neural pathway mapping and probabilistic correlation modeling, combined with finite element analysis and multi-objective reward functions for dynamic stimulation parameter adjustment, to achieve a highly individualized, precise, and non-invasive physical intervention method, significantly improving the treatment effect and patient experience for referred pain in the shoulder after laparoscopic surgery. Specifically, a three-dimensional neural stimulation target candidate region is constructed based on the patient's own imaging data and prioritized to accurately locate pain signal pathways. Finite element analysis algorithms are used to optimize the shape, position, and pressure distribution of the adjustable support straps and pads in the wearable shoulder support structure, ensuring that the support structure can achieve the optimal mechanical stimulation effect according to the patient's individual anatomy and target needs, while maximizing patient comfort. Through voxel-level segmentation and fiber tracing of medical imaging data, the main pain signal pathways in peripheral nerves (phrenic nerve, vagal nerve branches, brachial plexus, spinal nerve roots) and related brain regions (primary sensory cortex, secondary sensory cortex, anterior cingulate cortex, insula, thalamus) associated with referred pain are accurately identified.

[0058] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0059] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0060] Figure 1 A flowchart illustrating the method for optimizing parameters of non-invasive physical intervention for shoulder referred pain after laparoscopic surgery according to an embodiment of the present invention is shown.

[0061] Figure 2 A schematic diagram of the frame of the non-invasive physical intervention parameter optimization device for shoulder referred pain after laparoscopic surgery according to an embodiment of the present invention is shown.

[0062] Figure 3 A schematic diagram of the structure of a computer device according to an embodiment of the present invention is shown. Detailed Implementation

[0063] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0064] Figure 1 This diagram illustrates a flowchart of a method for optimizing parameters of non-invasive physical intervention for shoulder referred pain after laparoscopic surgery, according to an embodiment of the present invention. Specifically, as... Figure 1 As shown, it includes the following steps:

[0065] Step S101: Obtain the patient's individualized data, which includes medical imaging data, neurophysiological data, pain scale data, and physiological indicator data.

[0066] In this embodiment, the medical imaging data includes MRI, CT, or ultrasound images of the shoulder region after laparoscopic surgery; the neurophysiological data includes surface electromyography, event-related potentials, or evoked potential data; the pain scale data includes the Visual Analogue Scale (VAS) and the Numerical Rating Scale (NRS); and the physiological indicators include the patient's heart rate variability (HRV), skin conductance (EDA), and respiratory rate. For example, a 35-year-old female patient developed persistent dull pain in her right shoulder on the third day after laparoscopic cholecystectomy, which worsened especially during deep breathing. Shoulder MRI showed mild edema of the right diaphragm, with no obvious structural abnormalities or compression around the phrenic nerve, and no obvious lesions in the shoulder joint itself. Neck ultrasound examination of the right phrenic nerve revealed that it was in slightly closer contact with surrounding tissues during certain respiratory movements, but without obvious signs of compression. Multiple myofascial trigger points were also found in the upper part of the right trapezius muscle, with local tissue echo changes visible on ultrasound. sEMG recording of the electromyographic activity of the right trapezius and rhomboid muscles revealed a slightly elevated resting potential in the upper part of the right trapezius muscle, suggesting muscle tension. During deep breathing, diaphragmatic electromyographic activity was normal, but compensatory increases were observed in the right trapezius and levator scapulae muscles. Mild electrical stimulation of the myofascial trigger points in the patient's right shoulder using PREP revealed significantly increased PREP amplitude and slightly shortened latency in the primary sensory cortex (S1) and anterior cingulate cortex (ACC), suggesting increased pain sensitivity in these areas. Pain scale data (VAS: patient-reported current pain intensity 6 / 10; NRS: patient-reported current pain intensity 5) showed pain primarily located below the right acromion and along the medial border of the scapula, extending to the neck during deep breathing. An elevated LF / HF ratio, decreased SDNN, relatively enhanced sympathetic nerve activity, and impaired autonomic nervous system balance may be related to pain and stress. EDA skin conductance was slightly above the normal range and fluctuated considerably.

[0067] Step S102: Construct a personalized referred pain neural pathway mapping model based on the individualized data, wherein the personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them; and establishes a probabilistic correlation between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions.

[0068] In this embodiment, not only are potential stimulation areas identified, but they are also prioritized. Intervention can preferentially target the most critical areas that contribute the most to pain transmission, thereby achieving better results with less stimulation intensity and shorter time, avoiding ineffective or inefficient stimulation of non-critical areas, and improving treatment efficiency. A probabilistic link is established between the patient's subjective pain perception, objective physiological response, and specific neuroanatomical structures and cerebral cortex activity, capturing the complexity and dynamics of pain, rather than just static anatomical structures, thus more accurately predicting which areas intervention will effectively relieve pain. Through probabilistic correlation models, the potential impact of different stimulation targets on the patient's pain perception and physiological response can be predicted. When the patient's pain pattern changes, a reassessment can be conducted to guide adjustments to stimulation parameters and supporting structures, achieving truly adaptive intervention.

[0069] In one alternative approach, the personalized referred pain neural pathway mapping model, which generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them, further includes:

[0070] The medical imaging data is segmented at the voxel level to extract key anatomical structures, including potential sources of referred pain, related peripheral nerves, and related brain regions. The related peripheral nerves include the phrenic nerve, branches of the vagus nerve, the brachial plexus, and related spinal nerve roots. The related brain regions include the primary sensory cortex, the secondary sensory cortex, the anterior cingulate cortex, the insula, and the thalamus.

[0071] Each potential pain signal pathway is obtained by simulating the triangular mesh model or voxel label map of the key anatomical structure based on the fiber tracing algorithm of diffusion tensor imaging. The pathway is a pain signal transmission path that starts from the potential referred pain source, transmits along the relevant peripheral nerve to the spinal cord gray matter, and then reaches the relevant brain region through the ascending pathway of the spinothalamic tract.

[0072] Calculate the efficiency weight map of each potential pain signaling pathway in pain signal transmission, wherein the efficiency weight map includes the length, diameter, degree of myelination, number of synapses, and excitability threshold of neurons along the pathway;

[0073] Identify the primary pain signaling pathway that is most efficient and best matches the patient’s current pain perception pattern from the efficiency weighting graph;

[0074] Key ganglia, nerve fiber tract segments, and projection regions in the cerebral cortex of the main pain signaling pathway are mapped as initial three-dimensional neural stimulation target candidate regions; the initial three-dimensional neural stimulation target candidate regions are prioritized and stimulation weights are assigned to each target.

[0075] In this embodiment, individualized medical imaging data of patients is acquired and segmented at the voxel level to identify the unique neuroanatomical structures of each patient (including potential pain sources, related peripheral nerves, and brain regions), thereby avoiding the limitations of standardized treatment. Using diffusion tensor imaging (DTI) fiber tracing algorithms, the actual pain signal pathways within the patient's body are simulated and identified. Based on this, three-dimensional, prioritized candidate regions for neural stimulation targets are generated, making intervention no longer blind or based on experience, but rather based on precise localization of physiological pathways, significantly improving the effectiveness and safety of intervention. Not only are pathways identified, but the efficiency weighting of pain signals in each pathway is further calculated (considering multiple physiological parameters such as nerve fiber length, diameter, degree of myelination, number of synapses, and excitability thresholds of neurons along the pathway), making the identified main pain signal pathways closer to the patient's actual pain transmission mechanism. This best matches the patient's current pain perception pattern, allowing target selection to be based not only on anatomical and physiological parameters but also on the patient's subjective pain experience, achieving an organic combination of objective data and subjective feelings, thus enabling more precise intervention in the key pathways leading to the current pain perception.

[0076] For example, a 45-year-old female patient presented with referred pain in her right shoulder following cholecystectomy (laparoscopic surgery), with a VAS pain score of 6. Imaging revealed no structural damage to the shoulder joint, suggesting referred pain caused by diaphragmatic irritation. The patient underwent MRI of the right shoulder and cervicothoracic region, obtaining DTI sequences. A voxel-level segmentation algorithm was used to precisely extract the potential source of referred pain, related peripheral nerves (outlining the path of the right phrenic nerve descending from the neck down the thoracic cavity to the diaphragm, as well as the C3-C5 nerve roots and branches of the brachial plexus), and related brain regions (identifying the right primary sensory cortex (corresponding to the shoulder region), thalamus, anterior cingulate cortex, and insula) from the MRI data. Starting from the diaphragmatic irritation area, a DTI fiber tracing algorithm was used to simulate the nerve fiber path. Tracing path 1: Phrenic nerve → C3-C5 spinal cord segments → spinothalamic tract → thalamus → S1 / S2 / ACC / Insula. Pathway 2: Through visceral-somatic reflexes, signals are transmitted via the vagus nerve or other visceral nerves, passing through the spinal cord or other brainstem nuclei, ultimately converging at the brain's pain processing area. Hundreds, even thousands, of potential pain signal transmission pathways are generated, and the efficiency weight of each simulated pathway is calculated. For example, the phrenic nerve pathway is assumed to have a large fiber diameter and high degree of myelination, but a long path. The brachial plexus pathway is assumed to directly participate in somatic pain, with a fast conduction velocity, but may not be the primary mechanism of referred pain. The efficiency score of each pathway is calculated by quantifying the FA value (reflecting myelination), fiber length, etc., combined with the number of synapses and the neuronal excitability threshold. The efficiency scores of all pathways are ranked, and combined with the patient's reported right shoulder pain (VAS 6) and possible descriptions of pain characteristics (e.g., dull pain, persistent pain), the pathway with the highest efficiency and best matching the characteristics of referred pain is selected. The most efficient and suitable pathway is the phrenic nerve afferent pathway, passing through the C3-C5 spinal cord segments and then ascending to the brain, because it highly matches the classic mechanism of referred pain in the shoulder caused by diaphragmatic stimulation after gallbladder surgery. Key nodes on the identified main pain signal pathway (phrenic nerve → C3-C5 spinal cord segments → thalamus → S1) are mapped as target points: Target 1 (high priority): the superficial portion of the right phrenic nerve in the neck or supraclavicular fossa region, as this is a location easily stimulated non-invasively. Target 2 (medium priority): the skin projection area corresponding to the C3-C5 spinal cord segments (such as the paravertebral region of the cervical spine). Target 3 (low priority, only auxiliary): the area in the right primary sensory cortex corresponding to the shoulder (but directly acting on this area through non-invasive physical stimulation is difficult; it is more about modulating its plasticity). Priority ranking: Target 1 (cervical segment of the phrenic nerve) is given the highest priority and the largest stimulation weight because it is a key nerve for diaphragmatic afferent signals and is relatively superficial and easy to intervene in. Target 2 (spinal cord segment) is the next highest priority. Output a list of three-dimensional coordinates, including the cervical segment of the phrenic nerve and the C3-C5 spinal cord projection areas, along with their respective priorities and suggested stimulation weights.For example, [(X1, Y1, Z1, Priority: High, Weight: 0.7), (X2, Y2, Z2, Priority: Medium, Weight: 0.2),...].

[0077] In one alternative approach, establishing probabilistic associations between the patient's pain perception patterns, physiological responses, neuroanatomical structures, and cerebral cortex regions further includes:

[0078] The pain scale data is used to quantify the patient's current pain intensity, pain type, and pain distribution as the initial input for the pain perception pattern; the physiological index data is used to analyze the patient's autonomic nervous system activity, stress level, and breathing pattern as the initial input for the physiological response; and the neuroelectrophysiological data is used to analyze the latency, amplitude, and spatial distribution of the response of different regions of the cerebral cortex to pain stimuli as the initial input for the activity of the cerebral cortex regions.

[0079] The initial input of the pain perception pattern, the initial input of the physiological response, the initial input of the activity of the cerebral cortex, the key ganglia, nerve fiber tract segments and projection areas in the cerebral cortex of the main pain signal pathway are correlated and analyzed to obtain a multimodal correlation feature set.

[0080] A nonlinear probabilistic association model is established based on the multimodal association feature set to connect the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region.

[0081] In this embodiment, by associating multimodal features with key ganglia, nerve fiber tracts, and projection areas in the cerebral cortex along the main pain signaling pathways, it is possible to better predict which neural structures and brain regions play a dominant role in the current pain experience, thereby guiding the selection of stimulation targets. The probabilistic association model can not only pinpoint "where it hurts," but also, to some extent, explain "why it hurts."

[0082] In an alternative approach, establishing a nonlinear probabilistic association model between the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region based on the multimodal association feature set further includes:

[0083] The multimodal association feature set is defined as a high-dimensional feature vector F, wherein the high-dimensional feature vector contains d features extracted from pain perception patterns, physiological responses, cerebral cortex activity, and neuroanatomical structures.

[0084] Construct a probabilistic graphical model based on conditional random fields, wherein the Gibbs distribution of the probabilistic graphical model is:

[0085]

[0086] in, These are implicit state variables, representing the activation state of neuroanatomical structures and cerebral cortical regions; It is the partition function; Let be the set of all cliques in the graph; Let be the potential function defined on the clique c; ; This is a univariate feature function used to model the relationship between the activation probability of a single neural node and the multimodal feature F; It is a binary feature function used to model the coactivation relationship between peripheral nerves and cortical regions; , represents the model parameters to be learned; i and j represent the i-th and j-th neural nodes, respectively;

[0087] The optimal parameters are obtained by maximizing the log-likelihood function of the model on the training data, and then iteratively updated using the stochastic gradient ascent method to obtain a nonlinear probabilistic correlation model.

[0088] In this embodiment, through the potential function Univariate characteristic function and binary characteristic functions This method can not only assess the activation probability of individual neural nodes (such as a peripheral ganglion or cortical region), but also explicitly model the co-activation relationships and spatial / functional dependencies between different neural nodes (e.g., phrenic nerve activation is often accompanied by a response in a specific cortical region). By analyzing the learned parameters... , This allows us to infer the strength of the influence of different features (F) on the activation (Y) of neural nodes and the strength of the synergistic effect between different neural nodes.

[0089] Step S103: Based on the candidate regions of the three-dimensional neural stimulation targets and their priority ranking, the morphology, position and pressure distribution of the adjustable support straps and support pads in the wearable shoulder support structure are optimized using a finite element analysis algorithm.

[0090] In this embodiment, finite element analysis can accurately simulate the interaction between the support structure and human tissue, allowing for fine-tuning of the shape, position, and pressure distribution of the support bands and pads. This not only ensures that the preset effective stimulation pressure is achieved in the target area but also minimizes pressure on non-target areas and optimizes patient comfort, reducing the risk of decreased compliance due to discomfort.

[0091] In one alternative approach, optimizing the shape, position, and pressure distribution of the adjustable support straps and support pads in a wearable shoulder support structure using finite element analysis algorithms further includes:

[0092] A three-dimensional surface model of the patient's shoulder and torso is generated based on the medical imaging data, serving as the geometric basis for the design of the wearable shoulder support structure. The initial three-dimensional neural stimulation target candidate regions and their priorities are sorted and mapped onto the three-dimensional surface model of the patient's shoulder and torso, serving as the target regions and priorities for which the support structure needs to apply pressure. An initial CAD model of the wearable shoulder support structure is established, and its material properties and contact properties are defined.

[0093] The three-dimensional surface model of the patient's shoulder and trunk, the mapped three-dimensional neural stimulation target candidate region and its priority, and the initial CAD model of the wearable shoulder support structure are imported into the finite element analysis software to simulate the pressure distribution of the support structure and the patient's skin contact surface under different support band tensions, support pad positions and shapes, and calculate the average pressure, pressure gradient and stress-strain relationship with the surrounding tissues of each stimulation target region.

[0094] Based on the priority of the candidate stimulation target regions, a preset effective stimulation pressure range is achieved in the priority target regions, while minimizing the pressure in non-target regions and maximizing patient comfort; the shape, position, and material parameters of the adjustable support strap and support pad are adjusted according to the topology optimization iterative algorithm, and the optimized shape, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are output.

[0095] In this embodiment, the personalized referred pain neural pathway mapping model (especially the three-dimensional neural stimulation target candidate region and its priority) constructed in the previous steps is directly mapped onto the patient's three-dimensional surface model, making the target area where the support structure applies pressure highly biologically and pathologically targeted. Simulation using FEA software can accurately simulate the pressure distribution between the support structure and the patient's skin under different configurations (support band tension, support pad position, and shape), quantitatively assessing and optimizing the pressure applied to specific stimulation targets to ensure the pressure is within a preset effective stimulation range. Minimizing pressure in non-target areas solves the discomfort or pressure sores that may be caused by traditional support devices, significantly improving patient comfort. Topology optimization algorithms adjust the shape, position, and material parameters of the support band and support pad, exploring optimal structures that are difficult to achieve with traditional empirical design, achieving the best balance between efficiency and performance, and reducing the cost and time of manual trial and error.

[0096] Specifically, the initial three-dimensional neural stimulation target candidate regions and their priority rankings are extracted from the constructed personalized referred pain neural pathway mapping model. These three-dimensional stimulation targets and their priority information are then mapped onto the reconstructed three-dimensional surface models of the patient's shoulder and trunk. This can be achieved through coordinate transformation and surface projection techniques, ensuring that each target point corresponds to a target area on the patient's body surface requiring pressure and inherits its priority. Based on the principles of common shoulder support structures (such as shoulder straps, chest straps, and support pads), an initial CAD model of the wearable shoulder support structure is established in CAD software (such as SolidWorks, CATIA, and Fusion 360). Material properties (including Young's modulus, Poisson's ratio, density, and coefficient of friction) are defined for each component of the support structure (such as flexible support straps, elastic support pads, and fasteners). The contact characteristics between the support structure and the patient's skin (three-dimensional surface model), including the coefficient of friction and contact stiffness, are set to accurately simulate the interaction during actual wear. The 3D surface models of the patient's shoulder and torso, the mapped stimulation targets, and the initial CAD model of the wearable shoulder support structure were imported into finite element analysis software (such as ANSYS, Abaqus, COMSOL Multiphysics). All imported models were meshed and discretized into a finite number of elements and nodes. Multiple simulations were performed in the finite element analysis software for different combinations of support band tension, initial position, and morphology of the support pad. For each simulation, the pressure distribution map of the contact surface between the support structure and the patient's skin was calculated and output. The average pressure value, pressure gradient (reflecting the rate of pressure change), and stress-strain relationship of the target area and surrounding tissues were extracted from the simulation results for each stimulation target area. Using the pressure data calculated by the finite element analysis software, target priority, and support structure model as input, a topology optimization iterative algorithm was run. The objective function was to maximize the pressure efficiency of the target area (i.e., achieve the target pressure) and minimize the overall structural flexibility (maintaining support). The algorithm iterated based on preset equilibrium equation constraints (physical and mechanical equilibrium), volume constraints (material usage limits), and manufacturing constraints (minimum structural feature size, smoothness, etc.). In each iteration, the shape and position of the adjustable support straps and support pads, as well as virtual material parameters (such as relative density, which can later be converted into actual material selection or geometric design), are adjusted to gradually converge to the optimal solution. After multiple iterations, the final shape, optimal position (e.g., the precise coordinates and orientation of the support pads on the 3D model), and ideal pressure distribution of the optimized wearable shoulder support structure are output.

[0097] For example, MRI scans of the patient's right shoulder and upper trunk were performed to reconstruct a 3D model of the patient's right scapula, clavicle, ribs, and chest wall muscles. A neural pathway mapping model revealed that the patient's pain was primarily correlated with a projection point of a branch of the right phrenic nerve below the scapula and a segment of the C4 spinal nerve root, with the phrenic nerve branch targeting a higher priority area. Personalized 3D neural stimulation targets (a 2cm diameter circular area approximately 3cm below the scapula and a linear area approximately 1cm beside the C4 spine in the neck) were mapped onto the patient's 3D trunk model. The higher-priority phrenic nerve target was labeled as a "high-priority pressure area," and the C4 area as a "medium-priority pressure area." An initial CAD model of the shoulder support strap was designed, including a strap around the chest, a strap across the right shoulder, and an adjustable silicone support pad connected to the strap. The patient's 3D model and the initial support structure were imported into ANSYS to simulate several different wearing methods, such as wearing method A: a loose support strap, a support pad located in the scapular region, diffuse pressure distribution, and insufficient pressure on the phrenic nerve target. Wearing Method B: Tightening the support band slightly shifts the support pad, increasing pressure at the phrenic nerve target point. However, this also creates excessively high pressure peaks in non-target areas (such as the armpit), reducing patient comfort. Wearing Method C: Adjusting the shape of the support pad to a more ergonomic arc and slightly increasing its area at the phrenic nerve target point while reducing pressure on surrounding tissues. The optimization goal is to achieve an average pressure of 10-12 kPa at the phrenic nerve target point, 5-7 kPa at the C4 target point, and axillary pressure below 3 kPa, while minimizing the overall flexibility of the support structure. This involves iteratively adjusting the geometry of the support pad (to better fit the curved surface at the phrenic nerve target point and create a slender protrusion at the C4 target point), the width and thickness of the support band (thickening at key stress points to distribute stress), and the connection point between the support band and the support pad. An optimized CAD model is output, in which the support pad at the phrenic nerve target point is generated with a customized support pad having a specific curvature and thickness distribution based on the patient's subscapular anatomical shape and target point location. A slightly convex structure is also formed at the C4 target point. The width of the support band is appropriately reduced in the axillary region to avoid compression, while it remains wider above the shoulder to provide support. When patients wear this optimized shoulder support structure, they can feel precise pressure applied to the source of referred pain, and the wearing is comfortable, significantly relieving postoperative referred shoulder pain.

[0098] In one alternative approach, adjusting the shape, position, and material parameters of the adjustable support strip and support pad according to a topology optimization iterative algorithm further includes:

[0099] Construct an objective optimization function that aims to maximize the pressure efficiency of the target area and minimize the overall structural flexibility; wherein the objective optimization function is subject to equilibrium equation constraints, volume constraints, and manufacturing constraints;

[0100] Solving the objective optimization function outputs the optimal material distribution map to determine the optimal shape, position, and material parameter combination of the adjustable support strip and support pad.

[0101] For example, a patient experiencing shoulder pain after laparoscopic surgery has their key stimulation targets calculated using a personalized model. These targets are a point A on the suprascapular nerve pathway and a point B on the trapezius muscle, with point A having higher priority (requiring greater pressure). Traditional methods might involve using a standard shoulder support strap and attempting to compress points A and B by adjusting the strap tightness and adding pads of different shapes. This process is time-consuming and its effectiveness is uncertain; attempting to compress point A may result in other areas becoming too tight, causing pressure discomfort, or the pads may be mismatched, leading to pressure dispersion. In this application, the patient's 3D shoulder model, the precise locations and priorities of target points A and B are input. Iteration begins with a complete rectangular block of material covering the shoulder. After iterative calculations, it was found that to effectively deliver 30 mmHg of pressure to the high-priority point A, a localized, thicker, rigid support arch structure needs to be formed directly behind point A. To deliver 20 mmHg of pressure to point B while maximizing comfort, a larger but thinner flexible support surface is generated below point B to distribute the pressure. To connect the two support points and withstand the tension of the straps, a meandering, variable-section main beam structure was created, avoiding protruding areas such as the scapula to prevent discomfort. The final design resulted in a support structure resembling a biomimetic skeleton, with the optimal shape of the support pads (small and well-defined at point A, large and flat at point B) and the optimal path of the support straps. When wearing the support structure, the patient feels precise pressure applied to the target points requiring treatment, while other areas fit snugly without pressure, achieving a perfect balance between therapeutic efficacy and comfort.

[0102] In one alternative approach, the objective optimization function is:

[0103]

[0104] in, To support the relative density of each finite element in the structure; The number of target regions; Let be the target pressure value for the i-th target area; The average pressure in the region of the i-th target point is calculated through finite element analysis. This is the tolerance parameter; These are the weighting coefficients; This is the global displacement vector; This is the global stiffness matrix;

[0105] The equilibrium equations are constrained as follows:

[0106]

[0107] in, This is a load vector function originating from the tension of the support belt;

[0108] The volume constraint is:

[0109]

[0110] in, Let x be the volume of the x-th finite element; The total number of finite element elements in the design domain; For the maximum permissible material volume, , It is the volume fraction. The total volume of the design domain;

[0111] The manufacturing constraints are:

[0112]

[0113] in, The relative density of the x-th finite element after smoothing; Let be the volume of the j-th finite element; Let be the original relative density of the j-th finite element; For linear weighting functions, , To control the minimum feature size parameter, Center point With the center point The Euclidean distance between them.

[0114] In this embodiment, the actual pressure in each target area Infinitely close to the ideal target pressure calculated based on physiological models Tolerance parameters allow for setting different precision requirements for targets of varying importance (smaller for higher priority targets), ensuring the bioeffectiveness of therapeutic stimulation and maximizing intervention effects from a mechanical standpoint. The high-rigidity support structure deforms less under stress, maintaining the preset shape and pressure distribution more stably and preventing unexpected pressure points and shear forces in non-target areas due to structural deformation. Weighting coefficients balance the goals of precise pressure application and overall comfort, designing more comfortable and longer-tolerable devices for pressure-sensitive patients, significantly improving patient compliance. Volume constraints retain material only in the most mechanically critical areas, naturally creating a lightweight, biomimetic topology (similar to the trabecular structure of bone), directly reducing the weight of the device, lessening the burden on patients, and lowering material costs.

[0115] Step S104: Dynamically explore the optimal combination of stimulation parameters based on the probability correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

[0116] In this embodiment, by dynamically exploring the optimal combination of stimulation parameters, the stimulation parameters can be adjusted according to the patient's real-time feedback (changes in pain intensity, fluctuations in physiological indicators, and neuroelectrophysiological responses), so that the treatment plan can be optimized as the patient's condition progresses, ensuring that the best therapeutic effect is maintained throughout the entire treatment process.

[0117] In this way, the patient's pain intensity can be minimized, the neural plasticity changes in the painful brain region can be maximized, and the patient's discomfort and adverse reactions can be minimized based on the combination of stimulation parameters and reward function.

[0118] In practice, maximizing the neuroplasticity changes in the pain-affecting brain regions not only suppresses pain signals in the short term but also reshapes the brain's perception and processing of pain. Therefore, the intervention has a more lasting effect and helps to fundamentally improve chronic pain. It also minimizes patient discomfort and adverse reactions, avoiding secondary harm or decreased compliance caused by overstimulation or inappropriate parameters.

[0119] According to the solution provided by the present invention, individualized patient data is acquired, including medical imaging data, neurophysiological data, pain scale data, and physiological index data; a personalized referred pain neural pathway mapping model is constructed based on the individualized data, wherein the personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on the imaging data and prioritizes them; and a probabilistic correlation is established between the patient's pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region; based on the three-dimensional neural stimulation target candidate region and its priority ranking, the morphology, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are optimized using a finite element analysis algorithm; and the optimal combination of stimulation parameters is dynamically explored based on the probabilistic correlation, the morphology, position, and pressure distribution of the adjustable support strap and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode. This invention utilizes neural pathway mapping and probabilistic correlation modeling, combined with finite element analysis and multi-objective reward functions for dynamic stimulation parameter adjustment, to achieve a highly individualized, precise, and non-invasive physical intervention method, significantly improving the treatment effect and patient experience for referred pain in the shoulder after laparoscopic surgery. Specifically, a three-dimensional neural stimulation target candidate region is constructed based on the patient's own imaging data and prioritized to accurately locate pain signal pathways. Finite element analysis algorithms are used to optimize the shape, position, and pressure distribution of the adjustable support straps and pads in the wearable shoulder support structure, ensuring that the support structure can achieve the optimal mechanical stimulation effect according to the patient's individual anatomy and target needs, while maximizing patient comfort. Through voxel-level segmentation and fiber tracing of medical imaging data, the main pain signal pathways in peripheral nerves (phrenic nerve, vagal nerve branches, brachial plexus, spinal nerve roots) and related brain regions (primary sensory cortex, secondary sensory cortex, anterior cingulate cortex, insula, thalamus) associated with referred pain are accurately identified.

[0120] Figure 2 A schematic diagram of the frame of the non-invasive physical intervention parameter optimization device for shoulder referred pain after laparoscopic surgery, according to an embodiment of the present invention, is shown. The non-invasive physical intervention parameter optimization device for shoulder referred pain after laparoscopic surgery includes:

[0121] The individualized data acquisition module 210 is used to acquire the patient's individualized data, which includes medical imaging data, neuroelectrophysiological data, pain scale data, and physiological indicator data.

[0122] The personalized referred pain neural pathway mapping module 220 is used to construct a personalized referred pain neural pathway mapping model based on the individualized data. The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes it. It also establishes a probabilistic correlation between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions.

[0123] The wearable shoulder support structure optimization module 230 is used to optimize the shape, position and pressure distribution of the adjustable support band and support pad in the wearable shoulder support structure based on the candidate regions of the three-dimensional nerve stimulation target and their priority ranking, using the finite element analysis algorithm.

[0124] The optimal stimulation parameter optimization module 240 is used to dynamically explore the optimal combination of stimulation parameters based on the probability correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein, the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

[0125] Figure 3 The diagram shows a structural schematic of an embodiment of the computer device of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the computer device.

[0126] like Figure 3 As shown, the computer device may include: a processor 302, a communications interface 304, a memory 306, and a communications bus 308.

[0127] The 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 network elements, such as clients or other servers. The processor 302 executes program 310, specifically performing the relevant steps in the above-described embodiment of the non-invasive physical intervention parameter optimization method for shoulder referred pain after laparoscopic surgery.

[0128] Specifically, program 310 may include program code that includes computer operation instructions.

[0129] 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 computer device includes one or more processors, which 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.

[0130] Memory 306 is used to store program 310. Memory 306 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0131] According to the solution provided by the present invention, individualized patient data is acquired, including medical imaging data, neurophysiological data, pain scale data, and physiological index data; a personalized referred pain neural pathway mapping model is constructed based on the individualized data, wherein the personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on the imaging data and prioritizes them; and a probabilistic correlation is established between the patient's pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region; based on the three-dimensional neural stimulation target candidate region and its priority ranking, the morphology, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are optimized using a finite element analysis algorithm; and the optimal combination of stimulation parameters is dynamically explored based on the probabilistic correlation, the morphology, position, and pressure distribution of the adjustable support strap and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode. This invention utilizes neural pathway mapping and probabilistic correlation modeling, combined with finite element analysis and multi-objective reward functions for dynamic stimulation parameter adjustment, to achieve a highly individualized, precise, and non-invasive physical intervention method, significantly improving the treatment effect and patient experience for referred pain in the shoulder after laparoscopic surgery. Specifically, a three-dimensional neural stimulation target candidate region is constructed based on the patient's own imaging data and prioritized to accurately locate pain signal pathways. Finite element analysis algorithms are used to optimize the shape, position, and pressure distribution of the adjustable support straps and pads in the wearable shoulder support structure, ensuring that the support structure can achieve the optimal mechanical stimulation effect according to the patient's individual anatomy and target needs, while maximizing patient comfort. Through voxel-level segmentation and fiber tracing of medical imaging data, the main pain signal pathways in peripheral nerves (phrenic nerve, vagal nerve branches, brachial plexus, spinal nerve roots) and related brain regions (primary sensory cortex, secondary sensory cortex, anterior cingulate cortex, insula, thalamus) associated with referred pain are accurately identified.

[0132] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination of all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed can be employed. 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 serves the same, equivalent, or similar purpose. Furthermore, those skilled in the art will understand 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 invention and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices may be embodied by the same hardware item. Unless otherwise specified, the steps in the above embodiments should not be construed as limiting the order of execution.

Claims

1. A method for optimizing parameters of non-invasive physical intervention for referred pain in the shoulder after laparoscopic surgery, characterized in that, include: Acquire individualized patient data, including medical imaging data, neurophysiological data, pain scale data, and physiological indicator data; A personalized referred pain neural pathway mapping model is constructed based on the individualized data. The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them. Furthermore, a probabilistic correlation is established between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions. Based on the candidate regions of the three-dimensional neural stimulation targets and their priority ranking, the morphology, position and pressure distribution of the adjustable support straps and support pads in the wearable shoulder support structure are optimized using the finite element analysis algorithm. The optimal combination of stimulation parameters is dynamically explored based on the probabilistic correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

2. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 1, characterized in that, The medical imaging data includes MRI, CT, or ultrasound images of the shoulder region after laparoscopic surgery; The neurophysiological data includes surface electromyography, event-related potentials, or evoked potential data. The pain scale data includes the Visual Analogue Scale (VAS) and the Numerical Rating Scale (NRS); The physiological data include the patient's heart rate variability (HRV), skin conductance (EDA), and respiratory rate.

3. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 1, characterized in that, The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them, further including: The medical imaging data is segmented at the voxel level to extract key anatomical structures, including potential sources of referred pain, related peripheral nerves, and related brain regions. The related peripheral nerves include the phrenic nerve, branches of the vagus nerve, the brachial plexus, and related spinal nerve roots. The related brain regions include the primary sensory cortex, the secondary sensory cortex, the anterior cingulate cortex, the insula, and the thalamus. Each potential pain signal pathway is obtained by simulating the triangular mesh model or voxel label map of the key anatomical structure based on the fiber tracing algorithm of diffusion tensor imaging. The pathway is a pain signal transmission path that starts from the potential referred pain source, transmits along the relevant peripheral nerve to the spinal cord gray matter, and then reaches the relevant brain region through the ascending pathway of the spinothalamic tract. Calculate the efficiency weight map of each potential pain signaling pathway in pain signal transmission, wherein the efficiency weight map includes the length, diameter, degree of myelination, number of synapses, and excitability threshold of neurons along the pathway; Identify the primary pain signaling pathway that is most efficient and best matches the patient’s current pain perception pattern from the efficiency weighting graph; Key ganglia, nerve fiber tract segments, and projection regions in the cerebral cortex of the main pain signaling pathway are mapped as initial three-dimensional neural stimulation target candidate regions; the initial three-dimensional neural stimulation target candidate regions are prioritized and stimulation weights are assigned to each target.

4. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 3, characterized in that, Further, based on establishing probabilistic correlations between patients' pain perception patterns, physiological responses, neuroanatomical structures, and cerebral cortex regions, this includes: The pain scale data is used to quantify the patient's current pain intensity, pain type, and pain distribution as the initial input for the pain perception pattern; the physiological index data is used to analyze the patient's autonomic nervous system activity, stress level, and breathing pattern as the initial input for the physiological response; and the neuroelectrophysiological data is used to analyze the latency, amplitude, and spatial distribution of the response of different regions of the cerebral cortex to pain stimuli as the initial input for the activity of the cerebral cortex regions. The initial input of the pain perception pattern, the initial input of the physiological response, the initial input of the activity of the cerebral cortex, the key ganglia, nerve fiber tract segments and projection areas in the cerebral cortex of the main pain signal pathway are correlated and analyzed to obtain a multimodal correlation feature set. A nonlinear probabilistic association model is established based on the multimodal association feature set to connect the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region.

5. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 3, characterized in that, Optimizing the shape, position, and pressure distribution of the adjustable support straps and support pads in a wearable shoulder support structure using finite element analysis algorithms further includes: A three-dimensional surface model of the patient's shoulder and torso is generated based on the medical imaging data, serving as the geometric basis for the design of the wearable shoulder support structure. The initial three-dimensional neural stimulation target candidate regions and their priorities are sorted and mapped onto the three-dimensional surface model of the patient's shoulder and torso, serving as the target regions and priorities for which the support structure needs to apply pressure. An initial CAD model of the wearable shoulder support structure is established, and its material properties and contact properties are defined. The three-dimensional surface model of the patient's shoulder and trunk, the mapped three-dimensional neural stimulation target candidate region and its priority, and the initial CAD model of the wearable shoulder support structure are imported into the finite element analysis software to simulate the pressure distribution of the support structure and the patient's skin contact surface under different support band tensions, support pad positions and shapes, and calculate the average pressure, pressure gradient and stress-strain relationship with the surrounding tissues of each stimulation target region. Based on the priority of the candidate stimulation target regions, a preset effective stimulation pressure range is achieved in the priority target regions, while minimizing the pressure in non-target regions and maximizing patient comfort; the shape, position, and material parameters of the adjustable support strap and support pad are adjusted according to the topology optimization iterative algorithm, and the optimized shape, position, and pressure distribution of the adjustable support strap and support pad in the wearable shoulder support structure are output.

6. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 5, characterized in that, Adjusting the shape, position, and material parameters of the adjustable support strip and support pad according to the topology optimization iterative algorithm further includes: Construct an objective optimization function that aims to maximize the pressure efficiency of the target area and minimize the overall structural flexibility; wherein the objective optimization function is subject to equilibrium equation constraints, volume constraints, and manufacturing constraints; Solving the objective optimization function outputs the optimal material distribution map to determine the optimal shape, position, and material parameter combination of the adjustable support strip and support pad.

7. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 6, characterized in that, The objective optimization function is: ； in, To support the relative density of each finite element in the structure; The number of target regions; Let be the target pressure value for the i-th target area; The average pressure in the region of the i-th target point is calculated through finite element analysis. This is the tolerance parameter; These are the weighting coefficients; This is the global displacement vector; This is the global stiffness matrix; The equilibrium equations are constrained as follows: ； in, This is a load vector function originating from the tension of the support belt; The volume constraint is: ； in, Let x be the volume of the x-th finite element; The total number of finite element elements in the design domain; For the maximum permissible material volume, , It is the volume fraction. The total volume of the design domain; The manufacturing constraints are: ； in, The relative density of the x-th finite element after smoothing; Let be the volume of the j-th finite element; Let be the original relative density of the j-th finite element; For linear weighting functions, , To control the minimum feature size parameter, Center point With the center point The Euclidean distance between them.

8. The method for optimizing non-invasive physical intervention parameters for shoulder referred pain after laparoscopic surgery according to claim 4, characterized in that, The nonlinear probabilistic association model established based on the multimodal association feature set, relating the pain perception pattern, physiological response, neuroanatomical structure, and cerebral cortex region, further includes: The multimodal association feature set is defined as a high-dimensional feature vector F, wherein the high-dimensional feature vector contains d features extracted from pain perception patterns, physiological responses, cerebral cortex activity, and neuroanatomical structures. Construct a probabilistic graphical model based on conditional random fields, wherein the Gibbs distribution of the probabilistic graphical model is: ； in, These are implicit state variables, representing the activation state of neuroanatomical structures and cerebral cortical regions; It is the partition function; Let be the set of all cliques in the graph; Let be the potential function defined on the clique c; ; This is a univariate feature function used to model the relationship between the activation probability of a single neural node and the multimodal feature F; It is a binary feature function used to model the coactivation relationship between peripheral nerves and cortical regions; , represents the model parameters to be learned; i and j represent the i-th and j-th neural nodes, respectively; The optimal parameters are obtained by maximizing the log-likelihood function of the model on the training data, and then iteratively updated using the stochastic gradient ascent method to obtain a nonlinear probabilistic correlation model.

9. A non-invasive physical intervention parameter optimization device for referred pain in the shoulder after laparoscopic surgery, characterized in that, include: The individualized data acquisition module is used to acquire the patient's individualized data, which includes medical imaging data, neuroelectrophysiological data, pain scale data, and physiological indicator data. The personalized referred pain neural pathway mapping module is used to construct a personalized referred pain neural pathway mapping model based on the individualized data. The personalized referred pain neural pathway mapping model generates a three-dimensional neural stimulation target candidate region for referred pain in the patient's shoulder based on imaging data and prioritizes them; and establishes a probabilistic correlation between the patient's pain perception pattern, physiological response and neuroanatomical structure and cerebral cortex regions. The wearable shoulder support structure optimization module is used to optimize the shape, position and pressure distribution of the adjustable support band and support pad in the wearable shoulder support structure based on the candidate regions of the three-dimensional neural stimulation target and their priority ranking, using the finite element analysis algorithm. The optimal stimulation parameter optimization module is used to dynamically explore the optimal combination of stimulation parameters based on the probability correlation, the shape, position, and pressure distribution of the adjustable support band and support pad; wherein, the combination of stimulation parameters includes frequency, intensity, pulse width, and stimulation mode.

10. A computer device, comprising: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the non-invasive physical intervention parameter optimization method for shoulder referred pain after laparoscopic surgery as described in any one of claims 1-8.

Images