A wearable postoperative pain assessment and analgesia pump linkage system

Abstract

This invention provides a wearable postoperative pain assessment and analgesia pump linkage system, belonging to the field of medical rehabilitation equipment technology. The system includes a perception layer, an edge computing layer, a collaborative decision engine, and an execution feedback layer. The perception layer synchronously collects multimodal signals of patient movement, local wound mechanics, and pain physiological responses through a wearable motion sensing unit, wound biomechanical sensing patch, and physiological signal acquisition module. The edge computing layer completes spatiotemporal alignment of multimodal data, action recognition based on temporal convolutional networks, and quantitative calculation of wound stress risk index and action-pain response index. The collaborative decision engine incorporates a four-level risk assessment model to classify the risk of postoperative activities and performs dual-modal quantitative verification of occult wound risks using drug-masked and sleep-masked methods, outputting differentiated intervention decisions. The execution feedback layer achieves graded drug administration through an intelligent analgesia pump and pushes real-time physical intervention guidance via patient-side interactive devices.

Description

Technical Field

[0001] This application relates to the field of medical rehabilitation equipment technology, specifically the design of a wearable postoperative pain assessment and analgesia pump linkage system. Background Technology

[0002] Postoperative pain management is a core aspect of surgical clinical rehabilitation. Its intervention effect directly affects patients' willingness to move in the early stages, the incidence of postoperative complications, and the length of hospitalization for rehabilitation. It is also a key and challenging aspect of clinical postoperative nursing.

[0003] Existing wearable pain assessment and analgesia linkage systems mainly rely on systemic physiological signals such as heart rate variability and skin conductance, combined with limb movement signals collected by inertial sensors, attempting to achieve precise "action-triggered" analgesia. However, such systems have significant technical shortcomings. First, they lack the ability to perceive the local mechanical environment of the wound and cannot monitor key mechanical parameters such as tension and shear force at the surgical incision site. Postoperative patients in areas such as the abdomen, chest, and groin often experience a sudden surge in wound tension during routine rehabilitation activities such as coughing, turning over, and sitting up. If analgesics completely mask pain perception, patients are prone to making dangerous movements that excessively stretch the wound in a painless state, leading to poor wound healing or even wound dehiscence. Second, they ignore the potential conflict between analgesia and wound protection, treating pain as the sole intervention target and passively administering medication based on pain signals without considering the mechanical damage risks to the wound from different movements. They also cannot provide patients with immediate physical behavioral guidance while providing analgesia. Verbal instructions from clinical medical staff cannot provide continuous 24-hour supervision, and patients are also unlikely to accurately perform protective actions when pain suddenly occurs.

[0004] Furthermore, the existing system lacks a quantitative risk grading and assessment mechanism, making it impossible to distinguish between different scenarios such as simple pain, pain accompanied by high wound risk, and hidden wound risk. This results in a lack of targeted analgesia intervention strategies, either leading to over-administration that masks wound danger signals or under-administration that affects the patient's recovery experience. At the same time, it is impossible to quantify and record intervention events and effects, making it difficult to provide objective data support for dynamically adjusting analgesia and rehabilitation plans in clinical practice.

[0005] In summary, there is an urgent clinical need for a closed-loop postoperative pain assessment and analgesia pump linkage system that can simultaneously analyze the dynamic relationship between "motor intention, wound stress, and pain response," ensuring analgesia while achieving intelligent wound protection, and enabling quantitative risk assessment and intervention. This system would address many shortcomings of existing technologies and improve the clinical effectiveness of postoperative pain management and wound care. Summary of the Invention

[0006] The purpose of this application is to address the technical problems in existing pain management systems that cannot perceive the local mechanical environment of the wound, where there is a contradiction between analgesia and wound protection, a lack of real-time physical intervention guidance, and no quantitative risk grading assessment mechanism, making it difficult to ensure wound safety while guaranteeing analgesia.

[0007] To address the aforementioned technical problems, this application provides the following technical solution:

[0008] A wearable postoperative pain assessment and analgesia pump linkage system, characterized by comprising a sensing layer, an edge computing layer, a collaborative decision engine, and an execution feedback layer.

[0009] The sensing layer is used to collect patient motion signals, local wound biomechanical signals, and pain physiological response signals.

[0010] The sensing layer includes a wound mechanics sensor patch, which, from bottom to top, consists of a medical-grade silicone substrate, an array-type micro-strain sensing layer, a flexible pressure sensing layer, a signal convergence and preprocessing circuit layer, and a waterproof and breathable outer protective layer. The array-type micro-strain sensing layer is used to collect the local tension distribution and shear force of the wound. The signal convergence and preprocessing circuit layer is used to collect the sensor signals from the array-type micro-strain sensing layer and the flexible pressure sensing layer and transmit them to the edge computing layer.

[0011] The edge computing layer originates from the local real-time processing of multimodal data;

[0012] The collaborative decision engine is the core of the system's decision-making process. It is used to assess the risk level based on the data output from the edge computing layer and to output differentiated intervention decision instructions.

[0013] The execution feedback layer is used to receive and execute intervention decision instructions from the collaborative decision engine.

[0014] Preferably, the array-type micro-strain sensing layer is distributed in a grid pattern with a node spacing of 5 mm and a detection range of 0-50 N. The local tension of the wound is calculated by the resistance change in the array-type micro-strain sensing layer, as shown in the following formula: ;

[0015] in Tension value (unit: N). For calibration coefficients, This represents the change in resistance (in Ω). This is the initial resistance value (in Ω).

[0016] Preferably, the sensing layer further includes a wearable motion sensing unit and a physiological signal acquisition module, wherein the wearable motion sensing unit is worn on the patient's waist or upper arm, and the wearable motion sensing unit includes a six-axis inertial measurement unit and a 4-channel surface electromyography sensor.

[0017] The physiological signal acquisition module is used to acquire signals of heart rate variability, skin conductance, and local electromyography around the wound. It is used to simultaneously acquire the patient's pain physiological response signals to quantify the intensity of pain stress and provide core parameters for subsequent action-pain response index calculation.

[0018] Preferably, the core of the edge computing layer is an edge computing processing unit, which can be placed at the patient's bedside, on a nursing cart, or integrated into the nurse station terminal. The edge computing processing unit includes a multimodal data spatiotemporal alignment module, an action recognition module, and a dual-exponential calculation module.

[0019] Preferably, the multimodal data spatiotemporal alignment module includes a time alignment module and a spatial comparison module. The time alignment module adds hardware timestamps to the data packets collected by all sensing units. When an action event is detected, it automatically extracts a time window from 200ms before the start of the action to 500ms after the end of the action, and matches the motion, mechanical, and physiological data within the window one by one according to the time axis to eliminate data timing deviation.

[0020] The spatial alignment module allows medical staff to mark the actual location, direction, and length of the incision on an electronic schematic diagram of the wound biomechanical sensing patch using a companion mobile application. The system automatically establishes a mapping relationship between the spatial coordinates of each sensing unit of the patch and the anatomical location of the patient's incision based on the markings. When a specific action is identified, the system extracts the direction vector of the action and, in conjunction with the anatomical mapping relationship, calculates the directional distribution of tension in each area of ​​the incision under that action.

[0021] Preferably, the action recognition module is a built-in action recognition model based on a temporal convolutional network. The model inputs are six-axis data from an IMU and time-frequency features from sEMG, and outputs the category probability of common postoperative actions. The dual-index calculation module calculates the wound stress risk index (SI) and the action-pain response index (PI) simultaneously based on aligned multimodal data. Both indices are normalized to the 0-1 range. The larger the value, the higher the corresponding risk / pain level, providing a quantitative basis for the four-level risk assessment.

[0022] The action-pain response index (PI) is used to quantify the intensity of pain stress induced by the patient's actions.

[0023] The wound stress risk index (SI) is used to quantify the risk of mechanical damage to the surgical incision caused by movement.

[0024] Preferably, the wound stress risk index (SI) and action-pain response index (PI) are calculated using the following methods:

[0025] The first step is to extract the wound stress feature vector corresponding to the action: peak tension. Peak occurrence time , rate of change of tension Stress concentration factor (SCF), (the ratio of peak tension to average tension in the region), and tension duration (the total duration during which the tension exceeds 20% of the base value).

[0026] The second step is to calculate the mechanical damage potential:

[0027]

[0028] in For individualized safety thresholds, , , To map the original parameters to the nonlinear mapping function in the 0-1 interval, =0.5、 =0.3、 =0.2 is the weighting coefficient determined by finite element simulation and clinical data;

[0029] The third step is to introduce the time decay cumulative effect and calculate the cumulative stress hazard index:

[0030]

[0031] in The cumulative factor is 0.8. With an attenuation coefficient of 0.1, As an integral variable (representing a certain moment in the past), this step can quantify the cumulative mechanical damage effect of repeated actions within a short period of time, avoiding wound damage caused by a single low-stress but high-frequency action.

[0032] Preferably, the individualized security threshold The core parameters for dynamically generating the stress hazard index calculation These are not fixed values, but are dynamically generated based on the patient's surgical type, postoperative days, incision length, and suture method, to reflect the incision healing status of different patients. Specific values ​​are as follows:

[0033] By surgery type + postoperative days:

[0034] 1-3 days after abdominal surgery 4-7 days after surgery ;

[0035] 1-3 days after breast surgery 4-7 days after surgery ;

[0036] Groin surgery It improved by 20% compared to abdominal surgery at the same postoperative day.

[0037] Fine-tune according to the incision length: for every 5cm increase in incision length Reduced by 10%;

[0038] Fine-tuning according to the suture method: When using tension-reducing sutures It improves the efficiency by 15% compared to conventional suturing.

[0039] Preferably, the collaborative decision-making engine incorporates a risk assessment model, using the Stress Risk Index (SI) and Action-Pain Response Index (PI) output from the edge computing layer as core inputs to determine the risk level. It then performs dual-mode quantitative verification on the highest-risk Level IV occult wound risk, ultimately outputting differentiated intervention decision instructions. The quantitative standards for the four-level risk assessment are as follows:

[0040] Level I (Safe Activity): SI < 0.3 and PI < 0.3, the patient experiences no significant pain during the activity and there is no risk of mechanical damage to the wound;

[0041] Grade II (routine pain activity): SI < 0.3 and PI ≥ 0.3, the patient experiences significant pain during the activity, but there is no risk of mechanical damage to the wound, requiring only analgesic intervention;

[0042] Grade III (High-risk pain activity): SI≥0.3 and PI≥0.3, the patient's actions are not only painful, but also pose a mechanical risk to the wound, requiring physical intervention first + reward analgesia;

[0043] Grade IV (Hidden Wound Risk): SI ≥ 0.3 and PI < 0.3. The patient's movements pose a mechanical risk to the wound, but there is no obvious pain perception. There is a risk of hidden wound damage, and analgesia should be temporarily suspended and a graded warning should be issued.

[0044] Dual-modal quantitative validation of Grade IV occult wound risk: For Grade IV risk, the system initiates dual-modal quantitative validation of drug-masked and sleep-masked wounds to clarify the causes of pain perception loss and provide precise evidence for clinical intervention. The validation criteria are as follows:

[0045] Drug-masked type: The patient's pain perception is masked by the analgesic drug when the smart analgesia pump administers the drug ≥2 times in the past 30 minutes, or the cumulative dose administered in the past 30 minutes is ≥2mL.

[0046] Sleep-masked type: The patient's heart rate variability (LF / HF) is ≥1.5 in the past hour and the number of body movements is <5 times / minute, which is considered to be a deep sleep state, and the pain perception is masked by sleep.

[0047] Preferably, the execution feedback layer is the core of the system's intervention and recording, including an intelligent analgesia pump, a patient-end interactive device, a nurse station terminal, and a cloud / electronic medical record interface. Based on the intervention instructions output by the collaborative decision engine, it executes differentiated analgesia administration, physical intervention guidance, and graded early warning, and records and synchronizes all monitoring data, intervention operations, and effects in real time, realizing a closed loop of "intervention-feedback-recording".

[0048] Compared with the prior art, this application has at least the following beneficial effects:

[0049] This application introduces local wound biomechanical monitoring into a wearable pain assessment system. Through a flexible and stretchable wound biomechanical sensing patch, it achieves real-time, high-precision monitoring of tension, shear force, and vertical pressure at the surgical incision site for the first time. Combined with motion signals and pain physiological signals, it constructs a three-in-one collaborative perception network of "action-stress-pain", which simultaneously quantifies the patient's action type, pain level, and wound biomechanical damage risk. This solves the core technical blind spot of existing technologies that cannot perceive the wound safety boundary, and realizes a full-dimensional and multi-dimensional quantitative assessment of postoperative activity risk.

[0050] The system provided in this application establishes a four-level quantitative risk assessment to accurately balance the needs of analgesia and wound protection. Based on the quantitative thresholds of the stress hazard index and the action-pain response index, a four-level risk assessment model is established to accurately classify postoperative patients' routine safe activities, activities involving only pain, activities with pain and high wound risk, and activities with hidden wound risk. For level IV hidden wound risk, a dual-mode quantitative verification of drug-masked and sleep-masked methods is designed to clarify the specific reasons for the lack of pain perception. Through differentiated analgesia intervention strategies, precise drug administration is administered during simple pain, physical guidance is prioritized when pain is accompanied by high wound risk, and drug administration is temporarily suspended and an early warning is given when there is hidden wound risk. This effectively balances the needs of postoperative analgesia and wound protection, avoids the masking of wound danger signals by drugs, and reduces the clinical risks of wound traction and poor healing.

[0051] The system provided in this application integrates real-time physical intervention guidance, upgrading passive analgesia to active rehabilitation education. In high-risk pain activities (Level III), a closed-loop intervention logic of "real-time physical intervention guidance + reward-based analgesia" is introduced. Through patient-end interactive devices such as bone conduction headphones and vibrators, personalized protective movement prompts are pushed for different actions. After the patient performs the movement and the wound stress decreases to a safe range, a small dose of reward-based analgesic medication is administered. This design not only immediately reduces the risk of mechanical damage to the wound but also helps patients learn and solidify correct postoperative rehabilitation movements during pain management. This upgrades the system from a simple passive analgesia device to an active postoperative rehabilitation education device, significantly improving patients' self-rehabilitation ability and accelerating the postoperative recovery process.

[0052] The system provided in this application reduces the clinical nursing burden and provides quantitative rehabilitation basis through intelligent closed-loop control of the entire process. The system realizes full-process automation and intelligence of multi-source data collection, spatiotemporal alignment analysis, risk classification assessment, differentiated intervention execution, and effect feedback recording, without the need for real-time manual intervention by medical staff, which greatly reduces the workload of clinical postoperative pain management and wound care. At the same time, all monitoring data, action recognition results, risk assessment levels, intervention operations and effects are synchronized to the hospital's electronic medical record system in real time, automatically generating a joint report on functional rehabilitation and wound healing. This provides objective and quantitative data analysis basis for clinicians to dynamically adjust analgesia plans and formulate personalized rehabilitation plans, thereby improving the scientific and precise level of postoperative care.

[0053] The system provided in this application is highly adaptable and practical, and easy to implement in clinical practice. The system hardware adopts a medical-grade lightweight and flexible design. The wearable motion sensing unit is adapted to different wearing positions. The wound biomechanical sensing patch has multiple specifications and uses a medical-grade low-sensitivity adhesive layer, which is suitable for surgical incisions of different lengths and patients with different skin types, ensuring high wearing comfort and clinical safety. The sensing layer supports multi-protocol wireless / wired communication with the computing and execution layers, adapting to the communication environment of different wards. It can also interface with analgesia pumps from mainstream brands such as Smith, B. Braun, and Mindray. The electronic medical record interface is compatible with mainstream medical data protocols such as HL7FHIR and DICOM. No large-scale modification of existing clinical medical equipment is required. It has good clinical compatibility and practicality and can be widely used in postoperative rehabilitation and nursing scenarios for various surgical procedures such as abdominal, thoracic, and groin areas.

[0054] The system provided in this application enhances the postoperative recovery experience for patients and reduces postoperative complications. While ensuring wound safety, the system achieves precise and on-demand analgesia, avoiding adverse reactions from over-administration and effectively alleviating postoperative activity pain, thus improving the patient's postoperative recovery experience and willingness to move early. Simultaneously, through real-time monitoring and intelligent protection of wound biomechanical risks, it significantly reduces the incidence of postoperative complications such as wound dehiscence and poor healing, shortens the patient's hospitalization and recovery period, and reduces clinical medical costs, demonstrating high clinical and social value. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the system architecture of the present invention. Detailed Implementation

[0056] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. These embodiments are only used to explain the present invention and are not intended to limit the present invention. Those skilled in the art can make various modifications or variations based on the core technical ideas of the present invention, all of which fall within the protection scope of the present invention.

[0057] Please see Figure 1 The wearable postoperative pain assessment and analgesia pump linkage system described in this invention achieves simultaneous monitoring of movement, stress, and pain, four-level risk quantification assessment, and etiologically targeted differentiated intervention. The system consists of a four-layer architecture: a perception layer, an edge computing layer, a collaborative decision engine, and an execution feedback layer. These layers work together to form a closed-loop control system of "multi-source data acquisition - spatiotemporal alignment analysis - risk grading assessment - differentiated intervention - effect feedback recording." It is adaptable to various postoperative rehabilitation scenarios after abdominal, chest, and groin surgeries. All hardware is made of medical-grade materials, meeting clinical requirements for safety, disinfection, and compatibility. The system workflow is as follows: Figure 1 As shown.

[0058] Specifically, the hardware and functional configurations of each layer of the wearable postoperative pain assessment and analgesia pump linkage system are as follows:

[0059] (a) Perception layer:

[0060] The sensing layer is the core of the system's data acquisition, enabling synchronous and high-precision acquisition of patient motion signals, local wound mechanical signals, and pain physiological response signals. All sensing units have a uniform sampling frequency of ≥100 Hz and clock synchronization is achieved via Bluetooth protocol, with synchronization accuracy controlled within 10 ms to ensure consistency of multimodal data timing. Each unit adopts a medical-grade lightweight design, taking into account both wearing comfort and detection accuracy.

[0061] In one embodiment, the sensing layer includes a wearable motion sensing unit, a wound biomechanical sensing patch, and a physiological signal acquisition module.

[0062] The wearable motion sensing unit weighs less than 30g and is worn on the patient's torso or upper arm, adaptable to patients of different body types. Its core includes a six-axis inertial measurement unit (IMU, triaxial acceleration + triaxial angular velocity, acceleration range ±16g, angular velocity range ±2000° / s) and a four-channel surface electromyography (sEMG, sampling rate 1000Hz) sensor. This allows for real-time capture of the patient's torso / limb movement trajectory, posture changes, and muscle contraction activity signals, providing raw data for subsequent motion recognition. The unit is encapsulated in a medical-grade silicone shell with an IP65 protection rating, can withstand medical alcohol wiping, and has a battery life of ≥24 hours.

[0063] The wound mechanics sensing patch is a flexible, stretchable, multi-layered structure that is applied around the surgical incision. It includes an array of strain sensors and a capacitive pressure sensor to collect local tension distribution, shear force, and vertical pressure signals at a sampling frequency of no less than 100 Hz.

[0064] Specifically, in one embodiment, the wound mechanical sensing patch consists of, from bottom to top, a medical-grade silicone substrate, an array-type micro-strain sensing layer, a flexible pressure sensing layer, a signal convergence and preprocessing circuit layer, and a waterproof and breathable outer protective layer; the overall thickness is <1 mm, the elongation is ≥50%, and it is adapted to the deformation characteristics of the skin around the surgical incision.

[0065] Specifications and compatibility: Available in three sizes: 5×8 cm, 8×12 cm, and 12×15 cm, to fit surgical incisions of different lengths: 2-8 cm, 8-15 cm, and 15-20 cm, respectively. Made with medical-grade hypoallergenic acrylic adhesive, it is suitable for different skin types such as dry, oily, and sensitive skin. A single patch can be worn continuously for 12-24 hours without the risk of skin irritation.

[0066] Sensing parameters: The array-type micro-strain sensing layer is distributed in a grid pattern (node ​​spacing 5 mm), with a detection range of 0-50N and a detection accuracy of ±0.2N, used to collect local tension distribution and shear force in wounds; the capacitive pressure sensing layer has a detection range of 0-30kPa and a detection accuracy of ±0.5 kPa, used to collect vertical pressure in wounds; the signal convergence and preprocessing circuit layer integrates a multi-channel analog front-end and a 16-bit analog-to-digital converter, polling and collecting signals from all sensing units at a sampling frequency of 100 Hz. After amplification, filtering, and digitization, the signals are transmitted to the edge computing layer via Bluetooth 5.0 Low Power, with a signal transmission distance ≤15 m and transmission stability ≥99.5%.

[0067] Tension calculation: The local tension of the wound is calculated by the resistance change of the micro-strain sensing unit, using the following formula: ;

[0068] in Tension value (unit: N). For calibration coefficients, This represents the change in resistance (in Ω). The initial resistance value (in Ω) is given. This formula has been calibrated using large clinical sample data, and the calculation error is ≤5%.

[0069] The wound biomechanical sensing patch is available in three sizes: 5×8 cm, 8×12 cm, and 12×15 cm, suitable for surgical incisions of different lengths from 2 to 20 cm. It uses a medical-grade low-sensitivity adhesive layer and can be directly applied to the dressing around the incision or to the skin surface. The patch can be used for 12-24 hours.

[0070] The physiological signal acquisition module integrates heart rate variability, skin conductance, and local electromyography (EMG) acquisition units around the wound for synchronously acquiring the patient's pain physiological response signals. The clock synchronization accuracy of all sensors in the sensing layer is controlled within 10 ms. In one embodiment, the biosignal acquisition module can be integrated into a wearable motion sensing unit (integrated design) or worn independently (weight <20 g). The core includes a heart rate variability (HRV) acquisition unit (photoelectric, sampling rate 50 Hz), a skin conductance (EDA) acquisition unit (electrode, range 0-20 μS), and a local EMG acquisition unit around the wound (2 channels) for synchronously acquiring the patient's pain physiological response signals induced by movement, quantifying the intensity of pain stress, and providing core parameters for subsequent action-pain response index calculation.

[0071] (ii) Edge computing layer:

[0072] The core of the edge computing layer is an embedded edge computing processing unit (main control chip ARM Cortex-A72, equipped with NPU neural network processing unit, size 5×7×2 cm), which can be placed at the patient's bedside, nursing cart, or integrated into the nurse station terminal to realize local real-time processing of multimodal data, avoid cloud transmission delays, and ensure rapid response to intervention commands. The edge computing processing unit includes a multimodal data spatiotemporal alignment module, an action recognition module, and a dual-exponential calculation module, which are implemented as follows:

[0073] The multimodal data spatiotemporal alignment module is used to achieve precise temporal and spatial matching of motion, mechanical, and physiological signals, which is the basis for subsequent index calculation and risk assessment. The multimodal data spatiotemporal alignment module includes a time alignment module and a spatial comparison module.

[0074] Specifically, the time alignment module adds hardware timestamps to the data packets collected by all sensing units. When an action event is detected, it automatically extracts a time window from 200 ms before the action starts to 500 ms after the action ends, and maps the motion, mechanical and physiological data within the window to the time axis to eliminate data timing deviations.

[0075] The spatial alignment module allows medical staff to mark the actual location, direction, and length of the incision on an electronic diagram of the wound biomechanical sensing patch using a companion mobile application. The system automatically establishes a mapping relationship between the spatial coordinates of each sensing unit of the patch and the anatomical location (upper end, lower end, left side, right side) of the patient's incision based on the markings. When a specific action is identified, the system extracts the direction vector of the action (such as the trunk flexion angle and rotation angle), and calculates the directional distribution of tension in each area of ​​the incision under that action by combining the anatomical mapping relationship.

[0076] The action recognition module is a built-in action recognition model based on a temporal convolutional network (TCN). The model takes six-axis data from an IMU and time-frequency features from sEMG as input, and outputs the category probabilities of 15 common postoperative actions, such as coughing, turning over, sitting up, bending knees, walking, lying on one's side, and standing. The model has been pre-trained with a large sample of action data from over 1000 postoperative patients, achieving a basic recognition accuracy of ≥95%. After deployment, it supports incremental learning and can be fine-tuned in real time according to the individual patient's action patterns and limb activity characteristics to further improve recognition accuracy. The action recognition response latency is ≤0.3s.

[0077] The dual-index calculation module simultaneously calculates the wound stress risk index (SI) and the action-pain response index (PI) based on aligned multimodal data. Both indices are normalized to the 0-1 range, with higher values ​​representing higher risk / pain levels, providing a quantitative basis for the four-level risk assessment.

[0078] The action-pain response index (PI) is used to quantify the intensity of pain stress induced by the patient's action. The calculation formula is the decrease value, skin conductance slope, and local electromyography amplification. The HRV decrease value, skin conductance slope, and local electromyography amplification are all normalized, and the weighting coefficients are calibrated based on clinical pain assessment data, which can truly reflect the patient's subjective pain experience.

[0079] The wound stress risk index (SI) is used to quantify the risk of mechanical damage to the surgical incision caused by the action. It is the core quantitative indicator of this invention. The calculation is divided into three steps, and the cumulative stress risk index is used as the evaluation basis in the end.

[0080] The first step is to extract the wound stress feature vector corresponding to the action: peak tension. Peak occurrence time , rate of change of tension Stress concentration factor (SCF), (the ratio of peak tension to average tension in the region), and tension duration (the total duration during which the tension exceeds 20% of the base value).

[0081] The second step is to calculate the mechanical damage potential:

[0082]

[0083] in For individualized safety thresholds, , , To map the original parameters to the nonlinear mapping function in the 0-1 interval, =0.5、 =0.3、 =0.2 is the weighting coefficient determined by finite element simulation and clinical data;

[0084] The individualized safety threshold The core parameters for dynamically generating the stress hazard index calculation These are not fixed values, but are dynamically generated based on the patient's surgical type, postoperative days, incision length, and suture method, to reflect the incision healing status of different patients. Specific values ​​are as follows:

[0085] (I) By surgical type + postoperative days:

[0086] 1-3 days after abdominal surgery 4-7 days after surgery ;

[0087] 1-3 days after breast surgery 4-7 days after surgery ;

[0088] Groin surgery It improved by 20% compared to abdominal surgery at the same postoperative day.

[0089] (II) Fine-tune according to the incision length: for every 5cm increase in incision length, Reduced by 10%;

[0090] (III) Fine-tuning according to the suture method: When using tension-reducing sutures, It improves the efficiency by 15% compared to conventional suturing.

[0091] The third step is to introduce the time decay cumulative effect and calculate the cumulative stress hazard index:

[0092]

[0093] in The cumulative factor is 0.8. With an attenuation coefficient of 0.1, As an integral variable (representing a certain moment in the past), this step can quantify the cumulative mechanical damage effect of repeated actions within a short period of time, avoiding wound damage caused by a single low-stress but high-frequency action.

[0094] (III) Collaborative Decision Engine:

[0095] As the core of the system's decision-making, it incorporates a four-level risk assessment model. Using the stress hazard index (SI) and action-pain response index (PI) output from the edge computing layer as core inputs, it completes the risk level determination and performs dual-mode quantitative verification on the highest risk level IV occult wound risk. Finally, it outputs differentiated intervention decision instructions. All judgment logic is solidified into algorithms with a response latency of ≤0.2s.

[0096] The Level 4 risk assessment quantitative standard is based on the quantitative thresholds of SI and PI to classify risks. The thresholds are calibrated using clinical postoperative rehabilitation data, taking into account both analgesia needs and wound protection. The specific standards are as follows:

[0097] Level I (Safe Activity): SI < 0.3 and PI < 0.3, the patient experiences no significant pain during the activity and there is no risk of mechanical damage to the wound;

[0098] Grade II (routine pain activity): SI < 0.3 and PI ≥ 0.3, the patient experiences significant pain during the activity, but there is no risk of mechanical damage to the wound, requiring only analgesic intervention;

[0099] Grade III (High-risk pain activity): SI≥0.3 and PI≥0.3, the patient's actions are not only painful, but also pose a mechanical risk to the wound, requiring physical intervention first + reward analgesia;

[0100] Grade IV (Hidden Wound Risk): SI ≥ 0.3 and PI < 0.3. The patient's movements pose a mechanical risk to the wound, but there is no obvious pain perception. There is a risk of hidden wound damage, and analgesia should be temporarily suspended and a graded warning should be issued.

[0101] Dual-modal quantitative validation of Grade IV occult wound risk: For Grade IV risk, the system initiates dual-modal quantitative validation of drug-masked and sleep-masked wounds to clarify the causes of pain perception loss and provide precise evidence for clinical intervention. The validation criteria are as follows:

[0102] Drug-masked type: The patient's pain perception is masked by the analgesic drug when the smart analgesia pump administers the drug ≥2 times in the past 30 minutes, or the cumulative dose administered in the past 30 minutes is ≥2mL.

[0103] Sleep-masked type: The patient's heart rate variability (LF / HF) is ≥1.5 in the past hour and the number of body movements is <5 times / minute, which is considered to be a deep sleep state, and the pain perception is masked by sleep.

[0104] (iv) Execution Feedback Layer:

[0105] The execution feedback layer is the core of the system's intervention and recording, including an intelligent analgesia pump, patient-side interactive devices, nurse station terminals, and cloud / electronic medical record interfaces. Based on the intervention instructions output by the collaborative decision engine, it executes differentiated analgesia administration, physical intervention guidance, and graded early warning, and records and synchronizes all monitoring data, intervention operations, and effects in real time, realizing a closed loop of "intervention-feedback-recording." The specific configuration and functions are as follows:

[0106] The intelligent analgesia pump is compatible with mainstream medical analgesia pumps from brands such as Smith & Nephew, B. Braun, and Mindray, with a dosing accuracy of ±0.1 mL. It supports wireless / wired communication with the edge computing layer, executes differentiated dosing strategies based on risk levels, and has a dosing command response delay of ≤0.5s. Specific dosing parameters are as follows:

[0107] Level I risk: No medication administration is performed; only patient activity quality is recorded.

[0108] Level II risk: Triggers micro-pre-filled bolus administration, dose 0.5-1 mL, administration rate 0.2 mL / s, rapid relief of common pain;

[0109] Level III Risk: Direct administration should be postponed until the sensor detects that the patient is performing a protective action and the wound stress risk index SI < 0.3, then reward bolus administration should be triggered at a dose of 0.3-0.5 mL and an administration rate of 0.1 mL / s to achieve analgesia while protecting the wound.

[0110] Level IV risk: Immediately suspend medication until the wound stress risk index (SI) is <0.3 or healthcare personnel manually unlock medication access via the nurse station terminal to prevent the medication from further masking wound risk signals.

[0111] The patient-side interactive device includes at least one of a vibrator, bone conduction headphones, and a miniature speaker, adaptable to different postoperative scenarios (e.g., bone conduction headphones are suitable for bed rest, and vibrators are suitable for quiet environments). Its core functions are real-time physical intervention guidance and risk alerts, specifically:

[0112] At Level III risk, personalized voice / vibration prompts are pushed with a response delay of ≤0.5 s. Targeted protective action guidance is given for different actions, such as "Please press the wound gently with your hand before coughing" when coughing, "Please sit up slowly and support your abdomen with a pillow" when sitting up, and "Please turn your torso gently to avoid pulling on the wound" when turning over.

[0113] At Level IV risk, a continuous vibration / voice warning will be issued to remind the patient to stop the dangerous action.

[0114] The nurse station terminal serves as the operation and monitoring terminal for medical staff, supporting real-time viewing of patient motion signals, wound biomechanics data, risk levels, intervention records, etc. Its core functions include:

[0115] The tiered early warning system displays: Level IV risks are marked with an orange color code and prominently displayed on the terminal interface. It distinguishes between drug-masked and sleep-masked risks and pushes corresponding clinical recommendations (e.g., for drug-masked risks, the system suggests "the patient has adequate analgesia but the wound stress is excessive, and it is recommended to assess the wound condition and consider reducing the background infusion rate"; for sleep-masked risks, the system suggests "the wound stress is excessive during the patient's sleep, and it is recommended not to wake the patient temporarily, but to increase nighttime patrols").

[0116] Manual intervention function: Medical staff can manually unlock medication permissions for Level IV risk, adjust analgesia pump medication parameters, and send personalized nursing instructions to the patient via the terminal;

[0117] Data visualization: Displays the changing trends of the patient's stress risk index and action-pain response index in the form of curves and reports, intuitively reflecting the postoperative recovery status.

[0118] The cloud / electronic medical record interface is compatible with mainstream medical data protocols such as HL7FHIR and DICOM, and can be seamlessly connected with hospital electronic medical record systems and postoperative rehabilitation management cloud platforms. All monitoring data, motion recognition results, risk assessment levels, intervention operations and effects are synchronized to electronic medical records and the cloud in real time, automatically generating a joint report on functional rehabilitation and wound healing. This provides clinicians with objective and quantitative evidence for dynamically adjusting postoperative analgesia and rehabilitation plans, while also enabling long-term data storage and traceability.

[0119] Based on the above description, the system of the present invention has good clinical compatibility, adaptability, and practicality, and can be directly applied to existing hospital postoperative rehabilitation nursing scenarios without the need for large-scale modification of existing equipment. Specific adaptability is as follows:

[0120] The wearable motion sensing unit and wound biomechanical sensing patch of this application system are adapted to patients of different body types and skin types. The three specifications of the wound biomechanical sensing patch cover more than 95% of the surgical incision lengths in clinical practice. The intelligent analgesia pump is compatible with mainstream brands such as Smith, B. Braun, and Mindray, and does not require replacement of existing analgesia pump equipment, thus having good hardware compatibility.

[0121] In this application, the perception layer and the edge computing layer use Bluetooth 5.0 low-power communication, and the edge computing layer and the execution feedback layer support three communication methods: Bluetooth 5.0, Wi-Fi 6, and medical wired Ethernet, which can be adapted to the communication environment of different wards (such as wired Ethernet in wireless shielded wards).

[0122] The cloud / electronic medical record interface in the system provided in this application is compatible with mainstream medical data protocols such as HL7FHIR and DICOM, and can be seamlessly connected with hospital electronic medical record systems and postoperative rehabilitation management platforms of different brands to achieve data interconnection;

[0123] The system supports various postoperative rehabilitation scenarios such as bedside, ward, and rehabilitation room. The wearable device is lightweight and wireless, so it does not affect the patient's routine postoperative activities (such as turning over, sitting up, getting out of bed and walking).

[0124] To further illustrate the technical solution of the present invention, three specific embodiments are given in combination with three common postoperative scenarios in clinical practice: abdominal surgery, thoracic surgery, and groin surgery. In each embodiment, the system parameters, risk assessment, and intervention strategies are all adjusted individually according to the patient's surgical type and postoperative days. Those skilled in the art can reproduce the technical effects of the present invention based on these embodiments.

[0125] Example 1: Patient 2 days after abdominal surgery

[0126] Patient basic information: Male, 55 years old, 2 days after laparoscopic cholecystectomy, incision length 6 cm, routine suturing, no underlying diseases, wound biomechanical sensor patch of 5×8 cm size was selected and worn around the abdominal incision, wearable motion sensing unit was worn on the trunk and waist.

[0127] Individualized safety threshold: 1-3 days post-abdominal surgery The incision length is 6 cm (<8 cm), no fine-tuning is required, and the final result is... ;

[0128] Monitoring and Recognition: When the patient coughs, the system recognizes the action with 100% accuracy and collects the peak tension of the wound. The stress hazard index SI = 0.42 and the action-pain response index PI = 0.55.

[0129] Risk assessment: SI≥0.3 and PI≥0.3, classified as Level III (high-risk pain activity);

[0130] Intervention strategies:

[0131] ① The patient's bone conduction headphones push out a voice prompt: "Please gently press the wound with your hand before coughing."

[0132] ②After the sensor detected that the patient performed a protective action, the peak tension of the wound dropped to 12N, SI=0.25;

[0133] ③ The intelligent analgesia pump triggers reward bolus administration at a dose of 0.4 mL and an administration rate of 0.1 mL / s;

[0134] Intervention effect: The patient's pain was significantly relieved and the wound was not mechanically damaged. After the intervention, PI=0.18 and SI=0.25, returning to a safe state. All data were synchronized to the electronic medical record.

[0135] Example 2: Patient 3 days after thoracic surgery

[0136] Patient basic information: Female, 60 years old, 3 days after radical resection of lung cancer, incision length 12 cm, tension-reducing suture, history of diabetes, wound biomechanical sensor patch of 8×12 cm size is selected and worn around the chest incision, wearable motion sensing unit is worn on the upper arm of the limb.

[0137] Individualized safety threshold: 1-3 days post-chest surgery The incision length is 12 cm (an increase of 8 cm). Reduced by 16%, tension-reducing sutures increased by 15%, ultimately ;

[0138] Monitoring and identification: The system accurately identified patients turning over during sleep with a 98% accuracy rate in recognizing these movements and collected data on peak wound tension. The stress hazard index SI = 0.38 and the action-pain response index PI = 0.22.

[0139] Risk assessment: SI ≥ 0.3 and PI < 0.3, classified as Level IV (hidden wound risk), dual-mode validation initiated;

[0140] Dual-mode verification: The patient's LF / HF ratio was ≥1.8 within the past hour, and the number of body movements was <5 (3 times / minute), indicating sleep masking.

[0141] Intervention strategies:

[0142] ① The intelligent analgesia pump temporarily delays drug administration;

[0143] ② The nurse station terminal displays an orange alert and pushes a suggestion: "The patient's wound stress exceeds the standard during sleep. It is recommended not to wake the patient and to strengthen night patrols."

[0144] ③ Record risk events in electronic medical records;

[0145] Intervention effect: Medical staff strengthened night patrols, the patient was not awakened, the wound was not subjected to further mechanical damage, and the system pushed vibration prompts in advance when turning over, and the wound tension was controlled within a safe range.

[0146] Example 3: Patient 5 days post-groin surgery

[0147] Patient basic information: Male, 45 years old, 5 days after inguinal hernia repair surgery, incision length 5 cm, conventional suturing, wound biomechanical sensing patch of 5×8 cm size was selected and worn around the incision in the groin area, wearable motion sensing unit was worn on the trunk and waist.

[0148] Individualized safety threshold: 4-7 days post-groin surgery The incision length is 5 cm, no fine-tuning is required, and the final result is... ;

[0149] Monitoring and Recognition: When the patient started walking, the system recognized the movement with an accuracy rate of 99%, and collected the peak tension of the wound. The stress hazard index SI = 0.28 and the action-pain response index PI = 0.45.

[0150] Risk assessment: SI < 0.3 and PI ≥ 0.3, classified as Grade II (routine pain activity);

[0151] Intervention strategy: Intelligent analgesia pump triggers micro-pre-filled bolus administration, dose 0.8 mL, administration rate 0.2 mL / s;

[0152] Intervention effect: The patient's pain during walking was quickly relieved, the PI after the intervention was 0.20, and the wound tension was always controlled within a safe range with no risk of mechanical damage. The system recorded the quality of the patient's walking activities to the cloud.

[0153] In summary, this invention, through the collaborative work of a four-layer architecture, breaks through the bottleneck of existing postoperative pain management technologies, achieving a technological leap from "simple action-triggered analgesia" to "precise intervention and wound protection integrating action, stress, and pain." Compared with existing technologies, it has the following significant technical effects:

[0154] Filling the blind spot in wound biomechanical monitoring and achieving full-dimensional risk quantification, local wound biomechanical monitoring is introduced into a wearable pain assessment system, constructing a three-in-one collaborative perception network of "action-stress-pain". For the first time, real-time, high-precision monitoring of local tension, shear force, and vertical pressure of surgical incision is achieved, simultaneously quantifying the patient's actions, pain, and wound biomechanical risks, solving the core problem that existing technologies cannot perceive the wound safety boundary.

[0155] A four-level risk quantification assessment, balancing analgesia and wound protection, was established based on the quantification thresholds of the stress hazard index and the action-pain response index. This four-level risk assessment model enabled precise classification of postoperative activity risks, and a dual-mode quantification validation was performed for level IV occult wound risks to clarify the risk type. Through differentiated analgesia dosing strategies, precise medication was administered during routine pain, physical guidance was prioritized during high-risk pain, and medication was postponed during occult risks, effectively balancing postoperative analgesia and wound protection needs and avoiding the masking of wound danger signals by drugs.

[0156] Real-time physical intervention guidance upgrades passive analgesia to active rehabilitation. In high-risk pain activities of level III, a closed-loop strategy of "real-time physical intervention guidance + reward analgesia" is introduced. Through personalized voice / vibration prompts, patients are guided to perform protective actions. This not only effectively reduces the mechanical risk of wounds, but also helps patients learn the correct postoperative activity methods in pain management. This upgrades the system from a simple passive analgesia device to an active postoperative rehabilitation education device, significantly promoting the speed of postoperative recovery for patients.

[0157] The fully closed-loop intelligent control reduces the burden on clinical nursing by achieving fully closed-loop intelligent control of "multi-source data acquisition - spatiotemporal alignment analysis - risk classification assessment - differentiated intervention - effect feedback recording". All operations are completed automatically without the need for real-time manual intervention by medical staff, which effectively reduces the workload of clinical nursing staff in postoperative pain management. At the same time, all monitoring data, intervention operations and effects are synchronized to electronic medical records and the cloud in real time, generating a quantitative functional rehabilitation-wound healing joint report, which provides objective basis for clinicians to dynamically adjust rehabilitation plans and improves the intelligence and precision of postoperative care.

[0158] With high compatibility and practicality, the system is easy to promote and implement in clinical practice. All hardware is made of medical-grade materials, meeting clinical safety and disinfection requirements. The wearable device features a lightweight and wireless design, balancing wearing comfort and detection accuracy. The system is compatible with existing mainstream brands of analgesic pumps and hospital electronic medical record systems, requiring no large-scale modification of existing medical equipment. It has excellent conditions for clinical promotion and implementation, and can be widely used in postoperative rehabilitation care for various surgical procedures, including abdominal, thoracic, and groin surgeries.

Claims

1. A wearable postoperative pain assessment and analgesia pump linkage system, characterized in that: It includes a perception layer, an edge computing layer, a collaborative decision engine, and an execution feedback layer. The sensing layer is used to collect patient motion signals, local wound biomechanical signals, and pain physiological response signals. The sensing layer includes a wound mechanics sensor patch, which, from bottom to top, consists of a medical-grade silicone substrate, an array-type micro-strain sensing layer, a flexible pressure sensing layer, a signal convergence and preprocessing circuit layer, and a waterproof and breathable outer protective layer. The array-type micro-strain sensing layer is used to collect the local tension distribution and shear force of the wound. The signal convergence and preprocessing circuit layer is used to collect the sensor signals from the array-type micro-strain sensing layer and the flexible pressure sensing layer and transmit them to the edge computing layer. The edge computing layer originates from the local real-time processing of multimodal data; The collaborative decision engine is the core of the system's decision-making process. It is used to assess the risk level based on the data output from the edge computing layer and to output differentiated intervention decision instructions. The execution feedback layer is used to receive and execute intervention decision instructions from the collaborative decision engine.

2. The wearable postoperative pain assessment and analgesia pump linkage system according to claim 1, characterized in that: The array-type micro-strain sensing layer is distributed in a grid pattern with a node spacing of 5 mm and a detection range of 0-50 N. The local tension of the wound is calculated by the resistance change in the array-type micro-strain sensing layer, as shown in the following formula: ; in Tension value (unit: N). For calibration coefficients, This represents the change in resistance (in Ω). This is the initial resistance value (in Ω).

3. The wearable postoperative pain assessment and analgesia pump linkage system according to claim 2, characterized in that: The sensing layer also includes a wearable motion sensing unit and a physiological signal acquisition module. The wearable motion sensing unit is worn on the patient's waist or upper arm. The wearable motion sensing unit includes a six-axis inertial measurement unit and a 4-channel surface electromyography sensor. The physiological signal acquisition module is used to acquire signals of heart rate variability, skin conductance, and local electromyography around the wound. It is used to simultaneously acquire the patient's pain physiological response signals to quantify the intensity of pain stress and provide core parameters for subsequent action-pain response index calculation.

4. The wearable postoperative pain assessment and analgesia pump linkage system according to claim 1, characterized in that: The core of the edge computing layer is the edge computing processing unit, which can be placed at the patient's bedside, on a nursing cart, or integrated into the nurse station terminal. The edge computing processing unit includes a multimodal data spatiotemporal alignment module, an action recognition module, and a dual-exponential calculation module.

5. A wearable postoperative pain assessment and analgesia pump linkage system according to claim 4, characterized in that: The multimodal data spatiotemporal alignment module includes a time alignment module and a spatial comparison module. The time alignment module adds hardware timestamps to the data packets collected by all sensing units. When an action event is detected, it automatically extracts a time window from 200ms before the start of the action to 500ms after the end of the action, and matches the motion, mechanical, and physiological data within the window one by one according to the time axis to eliminate data timing deviation. The spatial alignment module allows medical staff to mark the actual location, direction, and length of the incision on an electronic schematic diagram of the wound biomechanical sensing patch using a companion mobile application. The system automatically establishes a mapping relationship between the spatial coordinates of each sensing unit of the patch and the anatomical location of the patient's incision based on the markings. When a specific action is identified, the system extracts the direction vector of the action and, in conjunction with the anatomical mapping relationship, calculates the directional distribution of tension in each area of ​​the incision under that action.

6. The wearable postoperative pain assessment and analgesia pump linkage system according to claim 1, characterized in that: The action recognition module is a built-in action recognition model based on a temporal convolutional network. The model inputs are six-axis data from an IMU and time-frequency features from sEMG, and outputs the probability of common postoperative actions. The dual-index calculation module calculates the wound stress risk index (SI) and the action-pain response index (PI) simultaneously based on aligned multimodal data. Both indices are normalized to the 0-1 range, with higher values ​​representing higher risk / pain levels, providing a quantitative basis for the four-level risk assessment. The action-pain response index (PI) is used to quantify the intensity of pain stress induced by the patient's actions. The wound stress risk index (SI) is used to quantify the risk of mechanical damage to the surgical incision caused by movement.

7. A wearable postoperative pain assessment and analgesia pump linkage system according to claim 6, characterized in that: The methods for calculating the wound stress risk index (SI) and the action-pain response index (PI) are as follows: The first step is to extract the wound stress feature vector corresponding to the action: peak tension. Peak occurrence time , rate of change of tension Stress concentration factor (SCF), (the ratio of peak tension to average tension in the region), and tension duration (the total duration during which the tension exceeds 20% of the base value). The second step is to calculate the mechanical damage potential: in For individualized safety thresholds, , , To map the original parameters to the nonlinear mapping function in the 0-1 interval, =0.5、 =0.3、 =0.2 is the weighting coefficient determined by finite element simulation and clinical data; The third step is to introduce the time decay cumulative effect and calculate the cumulative stress hazard index: in The cumulative factor is 0.

8. With an attenuation coefficient of 0.1, As an integral variable (representing a certain moment in the past), this step can quantify the cumulative mechanical damage effect of repeated actions within a short period of time, avoiding wound damage caused by a single low-stress but high-frequency action.

8. A wearable postoperative pain assessment and analgesia pump linkage system according to claim 7, characterized in that: The individualized safety threshold The core parameters for dynamically generating the stress hazard index calculation These are not fixed values, but are dynamically generated based on the patient's surgical type, postoperative days, incision length, and suture method, to reflect the incision healing status of different patients. Specific values ​​are as follows: By surgery type + postoperative days: 1-3 days after abdominal surgery 4-7 days after surgery ; 1-3 days after breast surgery 4-7 days after surgery ; Groin surgery The success rate is 20% higher than that of abdominal surgeries performed within the same postoperative period; Fine-tune according to the incision length: for every 5cm increase in incision length Reduced by 10%; Fine-tuning according to the suture method: When using tension-reducing sutures It improves the efficiency by 15% compared to conventional suturing.

9. A wearable postoperative pain assessment and analgesia pump linkage system according to claim 8, characterized in that: The collaborative decision-making engine incorporates a risk assessment model. Using the Stress Risk Index (SI) and Action-Pain Response Index (PI) output from the edge computing layer as core inputs, it determines the risk level and performs dual-mode quantitative verification on the highest-risk Level IV occult wound risk. Finally, it outputs differentiated intervention decision instructions. The quantitative standards for the four-level risk assessment are as follows: Level I (Safe Activity): SI < 0.3 and PI < 0.3, the patient experiences no significant pain during the activity and there is no risk of mechanical damage to the wound; Grade II (routine pain activity): SI < 0.3 and PI ≥ 0.3, the patient experiences significant pain during the activity, but there is no risk of mechanical damage to the wound, requiring only analgesic intervention; Grade III (High-risk pain activity): SI≥0.3 and PI≥0.3, the patient's actions are not only painful, but also pose a mechanical risk to the wound, requiring physical intervention first + reward analgesia; Grade IV (Hidden Wound Risk): SI ≥ 0.3 and PI < 0.

3. The patient's movements pose a mechanical risk to the wound, but there is no obvious pain perception. There is a risk of hidden wound damage, and analgesia should be temporarily suspended and a graded warning should be issued. Dual-modal quantitative validation of Grade IV occult wound risk: For Grade IV risk, the system initiates dual-modal quantitative validation of drug-masked and sleep-masked wounds to clarify the causes of pain perception loss and provide precise evidence for clinical intervention. The validation criteria are as follows: Drug-masked type: The patient's pain perception is masked by the analgesic drug when the smart analgesia pump administers the drug ≥2 times in the past 30 minutes, or the cumulative dose administered in the past 30 minutes is ≥2mL. Sleep-masked type: The patient's heart rate variability (LF / HF) is ≥1.5 in the past hour and the number of body movements is <5 times / minute, which is considered to be a deep sleep state, and the pain perception is masked by sleep.

10. A wearable postoperative pain assessment and analgesia pump linkage system according to claim 1, characterized in that: The execution feedback layer is the core of the system's intervention and recording, including an intelligent analgesia pump, patient-end interactive devices, nurse station terminals, and cloud / electronic medical record interfaces. Based on the intervention instructions output by the collaborative decision engine, it executes differentiated analgesia administration, physical intervention guidance, and graded early warning, and records and synchronizes all monitoring data, intervention operations, and effects in real time, realizing a closed loop of "intervention-feedback-recording".

Images