Remaining needle falling and permeation early warning device and method

Abstract

The invention relates to an early warning device and method for falling and permeation of an indwelling needle. The device comprises a sensing patch and a local processor, and the sensing patch is connected with the local processor and used for monitoring stress data of a first target area, monitoring electrical characteristic data of a second target area and monitoring liquid leakage data of a third target area. The stress data, the electrical characteristic data and the liquid leakage data are sent to the local processor; and the local processor is used for predicting the abnormal state of the remaining needle according to the received stress data, the electrical characteristic data and the liquid leakage data, and determining the falling and permeation early warning information of the remaining needle according to the abnormal state of the remaining needle. By adopting the method, early warning can be carried out at the early stage of liquid medicine seepage, and the early warning accuracy of falling and seepage risks is remarkably improved.

Description

Technical Field

[0001] This application relates to the field of medical monitoring technology, and in particular to an indwelling needle dislodgement and leakage early warning device and method. Background Technology

[0002] Indwelling intravenous catheters, also known as cannulas, are the main tool for clinical infusion therapy. Their application greatly reduces the pain of repeated punctures for patients and also reduces the workload of nursing staff. The safety of indwelling catheters during indwelling has always been a key and difficult point in clinical nursing work, mainly including: (1) Unplanned extubation, i.e., accidental dislodgement: due to the patient's unconscious pulling (especially when sleeping or agitated), excessive limb movement, improper fixation, etc., the indwelling catheter may be partially or completely dislodged from the blood vessel. (2) Extravasation and leakage: due to improper puncture, increased vascular permeability, or displacement of the catheter tip, the drug solution may seep into the subcutaneous tissue. Especially when infusing chemotherapy drugs, hypertonic solutions and other irritating drugs, extravasation can lead to serious tissue damage and necrosis.

[0003] To address the aforementioned issues, existing technologies include several alarm devices for monitoring infusion tubing. These devices primarily monitor the physical parameters of the tubing, such as detecting blockages, empty drips, or air ingress by monitoring droplets in the drip chamber, changes in tubing pressure, or air bubbles.

[0004] However, existing technologies generally struggle to monitor the "invisible" state of indwelling catheters within subcutaneous blood vessels. They also cannot provide early warnings of fluid leakage at its very early stages, leading to delays in intervention. Summary of the Invention

[0005] Therefore, it is necessary to provide an indwelling needle dislodgement and permeation early warning device and method that can provide early warning of drug leakage in the early stage and significantly improve the accuracy of early warning of dislodgement and permeation risks, in order to address the above-mentioned technical problems.

[0006] In a first aspect, this application provides an indwelling needle dislodgement and permeation early warning device, including a sensor patch and a local processor, wherein:

[0007] The sensor patch is connected to the local processor and is used to monitor the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area. The force data, electrical characteristic data and liquid leakage data are sent to the local processor.

[0008] The local processor is used to predict abnormal states of the indwelling needle based on the received force data, electrical characteristic data, and liquid leakage data, and to determine early warning information for indwelling needle dislodgement and leakage based on the abnormal states of the indwelling needle.

[0009] In one embodiment, the sensing patch includes an adhesive layer, a circuit layer, and a sensor layer, wherein:

[0010] An adhesive layer, located at the bottom of the sensor patch, is used to adhere the sensor patch to the patient's skin surface;

[0011] The circuit layer, located above the adhesive layer, is used to carry and connect the various sensors in the sensor layer;

[0012] The sensor layer, embedded in the circuit layer, is used to collect force data, electrical characteristic data, and liquid leakage data.

[0013] In one embodiment, the sensor layer includes a stress sensor, a bioimpedance sensor, and a humidity sensor, wherein:

[0014] A stress sensor is used to obtain force data by monitoring the pressure exerted on the patient's skin by the fixation wings of the indwelling needle in the first target area, and the magnitude and direction of the tension generated by the patient's movements;

[0015] A bioimpedance sensor is used to obtain electrical characteristic data by monitoring the conductivity of liquid permeation in a second target region;

[0016] A humidity sensor is used to obtain liquid leakage data by monitoring the humidity at the skin puncture point in a third target area.

[0017] In one embodiment, the stress sensor includes a first sensor unit and a second sensor unit, wherein:

[0018] The first sensor unit, located in the projection area where the fixation wing of the indwelling needle contacts the patient's skin, is used to monitor the pressure exerted by the fixation wing on the patient's skin.

[0019] The second sensor unit, located in the stress concentration area of ​​the indwelling needle catheter on the patient's skin, is used to monitor the magnitude and direction of tension generated by the traction of the indwelling needle catheter and the patient's movements.

[0020] In one embodiment, the local processor is specifically used for:

[0021] The received force data, electrical characteristic data, and liquid leakage data are input into a state prediction model trained based on machine learning algorithms to obtain the abnormal state of the indwelling needle.

[0022] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0023] In one embodiment, the local processor is also used for:

[0024] Based on the early warning information of indwelling needle dislodgement and penetration, the early warning level of the early warning information of indwelling needle dislodgement and penetration is determined. The early warning level is used to characterize the urgency of the early warning information of indwelling needle dislodgement and penetration.

[0025] In one embodiment, it further includes a communication module and a central monitoring platform, wherein:

[0026] The communication module is used to send early warning information about indwelling needle dislodgement and penetration, as well as the early warning level, to the central monitoring platform;

[0027] The central monitoring platform is used to display the received early warning information on indwelling needle dislodgement and infiltration, and to issue an alarm when the early warning level reaches the warning threshold.

[0028] Secondly, this application also provides a method for early warning of indwelling needle dislodgement and leakage, including:

[0029] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0030] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0031] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0032] In one embodiment, predicting abnormal states of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data includes:

[0033] By inputting the stress data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle can be obtained.

[0034] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage, including:

[0035] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0036] In one embodiment, it also includes:

[0037] Based on a pre-set database, the parameters of the state prediction model and the early warning model are periodically updated.

[0038] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0039] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0040] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0041] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0042] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0043] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0044] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0045] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0046] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0047] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0048] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0049] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0050] The aforementioned indwelling needle dislodgement and leakage early warning device, method, computer equipment, computer-readable storage medium, and computer program product collect force data, electrical characteristic data, and fluid leakage data through sensor patches, which are then analyzed by a local processor. This predicts abnormal states of the indwelling needle and, based on these abnormal states, determines early warning information for dislodgement and leakage. By monitoring force data in the first target area, it effectively warns of the risk of accidental dislodgement due to patient activity or traction. By monitoring electrical characteristic data in the second target area, it indirectly determines the relative position of the catheter tip and the blood vessel, solving the problem of indwelling needles being "invisible" under the skin and thus unable to detect displacement. By monitoring fluid leakage data in the third target area, it can determine leakage information in the early stages of fluid leakage, achieving rapid early warning at the early stage of drug leakage and overcoming the lag of traditional methods. Simultaneously, multi-parameter comprehensive judgment effectively avoids false alarms or missed alarms from single monitoring, significantly improving the accuracy of early warning for dislodgement and leakage risks. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a schematic diagram of an indwelling needle dislodgement and leakage early warning device in one embodiment;

[0053] Figure 2 This is a schematic diagram of a sensing patch in one embodiment;

[0054] Figure 3 This is a schematic diagram illustrating the application of the sensor patch in one embodiment;

[0055] Figure 4 This is another schematic diagram of the indwelling needle dislodgement and leakage early warning device in one embodiment;

[0056] Figure 5 This is a schematic diagram of a central monitoring platform in one embodiment;

[0057] Figure 6 This is a flowchart illustrating the indwelling needle dislodgement and penetration early warning method in another embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] In one exemplary embodiment, such as Figure 1 As shown, an indwelling needle dislodgement and leakage early warning device is provided, including a sensor patch and a local processor, wherein:

[0060] The sensor patch is connected to the local processor and is used to monitor the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area. The force data, electrical characteristic data and liquid leakage data are sent to the local processor.

[0061] The local processor is used to predict abnormal states of the indwelling needle based on the received force data, electrical characteristic data, and liquid leakage data, and to determine early warning information for indwelling needle dislodgement and leakage based on the abnormal states of the indwelling needle.

[0062] Traditional indwelling needle monitoring devices can only monitor the status of the indwelling needle by monitoring the infusion tubing. They cannot detect in time when the patient's indwelling needle leaks or is at risk of falling out, nor can they issue an early warning, which leads to prominent safety issues during the indwelling period.

[0063] Based on the above problems, this application proposes an indwelling needle dislodgement and leakage early warning device. It collects force data, electrical characteristic data, and fluid leakage data through a sensor patch, which are then analyzed by a local processor. The device predicts abnormal states of the indwelling needle and determines dislodgement and leakage early warning information based on these abnormal states. By monitoring force data in the first target area, it can effectively warn of the risk of accidental dislodgement caused by patient activity or traction. By monitoring electrical characteristic data in the second target area, it can indirectly determine the relative position of the catheter tip and the blood vessel, solving the problem of the indwelling needle being "invisible" under the skin and thus unable to detect displacement. By monitoring fluid leakage data in the third target area, it can determine leakage information in the early stages of fluid leakage, achieving rapid early warning in the early stages of drug leakage and overcoming the lag of traditional methods. Simultaneously, the comprehensive judgment of multiple parameters effectively avoids false alarms or missed alarms from single monitoring, significantly improving the accuracy of early warning of dislodgement and leakage risks.

[0064] For example, by Figure 1 The sensor patch monitors multimodal data related to abnormal states of the indwelling needle during its indwelling period. This includes monitoring force data in a first target area, electrical characteristic data in a second target area, and fluid leakage data in a third target area. The force data reflects the pressure experienced by the indwelling needle; the electrical characteristic data reflects the conductivity around the needle; and the fluid leakage data reflects the humidity around the skin puncture site. The first target area corresponds to the external fixation structure of the indwelling needle; the second target area corresponds to the subcutaneous tissue area; and the third target area corresponds to the skin puncture site.

[0065] Furthermore, after the sensing patch sends the force data, electrical characteristic data, and liquid leakage data to the local processor, the local processor can calculate the pressure changes on the indwelling needle, whether the indwelling needle is loose, the conductivity changes around the indwelling needle caused by liquid leakage, and the humidity changes caused by liquid passing through the skin puncture point of the indwelling needle. Then, based on the force data, electrical characteristic data, and liquid leakage data, it predicts abnormal states of the indwelling needle; and further determines indwelling needle dislodgement and leakage warning information based on the abnormal states. The abnormal states of the indwelling needle can include retention abnormalities and liquid leakage abnormalities. In specific implementation, the local processor can determine the retention abnormality of the indwelling needle based on the force data, determine the liquid leakage abnormality based on the electrical characteristic data and liquid leakage data, and then determine indwelling needle dislodgement and leakage warning information based on the retention abnormality and liquid leakage abnormality. In practice, based on stress data, abnormal conditions of the indwelling needle can be comprehensively predicted by analyzing whether the pressure, shear force, or tension is abnormal. For example, determining only pressure decay may indicate that the indwelling needle is fixed in place, but the fixation wings have loosened, potentially indicating a tendency for catheter displacement. Decreasing pressure coupled with increasing tension may indicate loosening of the fixation wings, while the catheter remains under stress, potentially indicating a tendency for catheter displacement. Furthermore, abnormal liquid leakage conditions can be determined based on electrical characteristic data and liquid leakage data. For example, a continuous decrease in impedance and a continuous increase in humidity may indicate that liquid leakage has already occurred.

[0066] Compared to current common technologies that monitor indwelling needle status by monitoring the infusion tubing, the aforementioned indwelling needle dislodgement and extravasation early warning device effectively identifies risks by analyzing changes in multimodal data before visible slippage and swelling occur, buying valuable time for intervention. Through multimodal data fusion, it can effectively distinguish between transient stress caused by normal activity and actual dislodgement, reducing false alarms. In particular, its early warning of extravasation of irritating drugs can greatly prevent serious tissue damage and ensure patient safety.

[0067] In one exemplary embodiment, the sensing patch includes an adhesive layer, a circuit layer, and a sensor layer, wherein:

[0068] An adhesive layer, located at the bottom of the sensor patch, is used to adhere the sensor patch to the patient's skin surface;

[0069] The circuit layer, located above the adhesive layer, is used to carry and connect the various sensors in the sensor layer;

[0070] The sensor layer, embedded in the circuit layer, is used to collect force data, electrical characteristic data, and liquid leakage data.

[0071] Figure 2 This is a schematic diagram of the sensor layer in one embodiment, such as... Figure 2As shown, the sensor patch can adopt a disposable medical flexible composite membrane structure. The shape of the sensor patch can be compatible with existing transparent fixation dressings. It can be directly covered and pasted on the puncture point and surrounding skin of a standard indwelling needle to sense multimodal data related to abnormal status of the indwelling needle in real time.

[0072] For example, the sensing patch may include an adhesive layer 1, a circuit layer 2, a circuit layer 3, and a sensor layer 4, wherein:

[0073] The adhesion layer 1, located at the bottom of the sensing patch, can be a skin-friendly adhesive layer made of medical-grade pressure-sensitive adhesive. It is used to firmly and comfortably adhere the entire sensing patch to the patient's skin surface, ensuring good adhesion between the sensors within the patch and the skin. The adhesion layer can have micro-windows corresponding to the areas where each sensor is located within the sensor layer, or it can be made of wave-transparent or conductive materials to ensure effective signal transmission from each sensor within the sensor layer.

[0074] Circuit layer 2 and circuit layer 3 are located above the adhesive layer and can be flexible circuit layers. They are made of flexible polymer materials such as polyimide or polyethylene terephthalate and have microcircuits printed or etched on them to connect and carry various sensor elements.

[0075] Sensor layer 4 is embedded in or integrally formed with the flexible circuit layer. It is the core functional layer of the sensor patch, integrating multiple micro sensors and arranging them in an optimized spatial position.

[0076] The sensor patch may also include an outer protective layer 5, wherein:

[0077] The outer protective layer 5 is located on the top layer of the sensor patch and can be made of waterproof and breathable medical non-woven fabric or polyurethane film. It is used to protect the internal circuitry of the sensor patch while allowing the skin to breathe normally through the sensor patch.

[0078] In addition, the sensor patch can also be equipped with flexible leads and micro connectors. The flexible leads are flat, flexible ribbon cables extending from the edge of the flexible circuit layer. The micro connectors are located at the ends of the flexible leads and are used for pluggable electrical and physical connections with the local processor.

[0079] In this embodiment, by integrating the adhesion layer, circuit layer, and sensor layer into the same sensing patch, and embedding multiple sensor layers into the circuit layer, it is possible to simultaneously acquire force data, electrical characteristic data, and liquid leakage data, while ensuring good adhesion between the patch and the skin and patient comfort, thus providing a reliable data foundation for subsequent multimodal data fusion analysis.

[0080] In one exemplary embodiment, the sensor layer includes a stress sensor, a bioimpedance sensor, and a humidity sensor, wherein:

[0081] A stress sensor is used to obtain force data by monitoring the pressure exerted on the patient's skin by the fixation wings of the indwelling needle in the first target area, and the magnitude and direction of the tension generated by the patient's movements;

[0082] A bioimpedance sensor is used to obtain electrical characteristic data by monitoring the conductivity of liquid permeation in a second target region;

[0083] A humidity sensor is used to obtain liquid leakage data by monitoring the humidity at the skin puncture point in a third target area.

[0084] For example, each sensor is located in the sensor layer. When collecting multimodal data, its target area is distributed in different areas of the sensing patch. That is, the effective area of ​​each sensor is a preset area corresponding to the indwelling position of the indwelling needle, wherein:

[0085] The first target area includes the projection area corresponding to the skin contact between the catheter fixation wings of the indwelling needle and the skin, and the stress concentration area corresponding to the point where the catheter of the indwelling needle and the transparent dressing protrude from the skin. The force data includes the pressure of the fixation wings of the indwelling needle on the patient's skin, and the shear force and tension generated by tubing traction, patient movement, etc. The second target area includes the subcutaneous tissue projection area corresponding to the area directly below the puncture point of the indwelling needle and within a preset range around the puncture point. The electrical characteristic data includes the electrical conductivity of the subcutaneous tissue. The third target area includes the body surface area directly below the puncture point of the indwelling needle corresponding to the sensing patch, and the absorbent pad of the dressing that is in close contact with the skin or the indwelling needle. The liquid leakage data includes the humidity around the skin puncture point of the indwelling needle.

[0086] Stress sensors may include micro-stress sensing arrays and tension sensing arrays; bioimpedance sensors may include a pair of microneedle electrodes or planar interdigital electrodes; humidity sensors may include a pair of conductive electrodes coated with a moisture-sensitive material at a predetermined spacing.

[0087] For example, Figure 3 This is a schematic diagram illustrating the application of the sensor patch in one embodiment. Figure 3 The application method of the sensor patch in the embodiments of this application will be described using an example.

[0088] Applying the sensor patch: Remove from the aseptic packaging Figure 2 The sensor patch shown is directly and smoothly applied to the puncture site of the indwelling needle and the surrounding skin. During application, it is crucial to ensure that the central area of ​​the patch, containing the built-in bioimpedance sensor and humidity sensor, is precisely aligned with the skin puncture point. Simultaneously, the stress sensor array at the patch's edge naturally covers the catheter fixation wings of the indwelling needle and the stress concentration area where the catheter exits the skin. The patch's adhesive layer ensures a tight fit with the skin, guaranteeing reliable signal acquisition.

[0089] Connecting the local processor: Remove the miniature connector from the end of the flexible lead at the edge of the patch and connect it to a magnetic, button-sized local processor. Then, clip the local processor to the edge of the patient's gown collar or pocket and secure the lead with the accompanying flexible cable clip to prevent pulling.

[0090] In this embodiment, a stress sensor monitors the pressure of the fixed wing in the first target area and the magnitude and direction of tension generated by the patient's movements, a bioimpedance sensor monitors the conductivity of fluid permeation in the second target area, and a humidity sensor monitors the humidity at the puncture point in the third target area. This enables multi-dimensional synchronous monitoring of abnormal states of indwelling needle retention, subcutaneous tissue fluid, and surface leakage. The collaborative data collection by the three sensors provides a comprehensive and complementary data foundation for subsequent multi-parameter fusion analysis.

[0091] In an exemplary embodiment, the stress sensor includes a first sensor unit and a second sensor unit, wherein:

[0092] The first sensor unit, located in the projection area where the fixation wing of the indwelling needle contacts the patient's skin, is used to monitor the pressure exerted by the fixation wing on the patient's skin.

[0093] The second sensor unit, located in the stress concentration area of ​​the indwelling needle catheter on the patient's skin, is used to monitor the magnitude and direction of tension generated by the traction of the indwelling needle catheter and the patient's movements.

[0094] For example, the stress sensor may be composed of a micro-stress array or a tension sensing array; for example, an array composed of multiple miniature capacitive or piezoresistive stress sensor units.

[0095] In practice, at least two sensor units can be set up. For example, a first sensor unit and a second sensor unit with the same physical structure can be set up. The first sensor unit is used to monitor the pressure exerted on the patient's skin by the fixation wing; the second sensor unit is used to monitor the magnitude and direction of the tension generated by the tubing pull of the indwelling needle and the patient's movements.

[0096] Its spatial distribution features are as follows: the first sensor unit is configured in the first target area, corresponding to the projection area of ​​the indwelling needle catheter fixation wing in contact with the skin; the second sensor unit is configured in the first target area, corresponding to the stress concentration area of ​​the indwelling needle catheter and the point where the transparent dressing is applied to the skin.

[0097] The bioimpedance sensor can consist of at least one pair of microneedle electrodes or planar interdigital electrodes. Its spatial distribution is characterized by being positioned in the central region of the patch, specifically within the subcutaneous tissue projection area directly below and around the indwelling needle puncture point within a 1-3 cm radius. By applying a weak, safe alternating current to the tissue, the impedance value between the electrodes is measured to monitor the subcutaneous tissue conductivity. The fluid used includes the patient's blood and the medication in the indwelling needle.

[0098] A humidity sensor can consist of a pair of conductive electrodes spaced at a preset distance, coated with a moisture-sensitive material. Its spatial distribution is characterized by: an inner layer precisely positioned directly below the patch puncture point, closely adhering to the skin or indwelling needle, used to detect whether blood or medication leaks through the skin puncture point onto the body surface.

[0099] For example, after entering continuous monitoring mode, each sensor synchronously collects three core signals from the sensor patch at a preset frequency and sends them to the local sensor:

[0100] Force data acquisition:

[0101] First sensor unit: continuously monitors the static pressure value and its minute fluctuations exerted on the skin by the fixation wing. When the fixation wing is slightly lifted due to loosening of the adhesive tape, the static pressure value will show a slow downward trend.

[0102] The second sensor unit monitors the external tension and shear force on the catheter in real time in a three-dimensional vector form. When the patient's arm moves unconsciously, periodic tension fluctuations with small amplitudes are generated; while when the catheter is continuously pulled by an external force, a tension signal with a clear direction and continuously increasing amplitude is detected.

[0103] Electrical characteristic data acquisition:

[0104] Bioimpedance sensor: A high-frequency, weak constant current alternating current is applied to subcutaneous tissue, and the voltage between electrodes is measured simultaneously to calculate the tissue's bioimpedance value. Under normal circumstances, tissue fluid is relatively stable, and the impedance value fluctuates around a baseline. Once medication or blood seeps out, the local tissue fluid increases, conductivity rises, and the impedance value decreases rapidly and continuously.

[0105] Liquid leakage data collection:

[0106] Humidity sensor: Detects minute amounts of humidity around the puncture site. Under normal circumstances, this area remains dry. If even a tiny amount of blood or medication seeps along the catheter wall onto the skin surface, the humidity sensor will instantly detect a sudden increase in humidity.

[0107] In this embodiment, by dividing the stress sensor into a first sensor unit and a second sensor unit, and respectively configuring them in the projection area of ​​the fixation wing and the stress concentration area where the catheter exits the skin, the fixation status and stress status of the indwelling needle can be monitored. The first sensor unit specifically monitors the static pressure between the fixation wing and the skin, facilitating subsequent judgment of whether the fixation is loose. The second sensor unit specifically monitors the magnitude and direction of the dynamic tension and shear force on the catheter, facilitating subsequent judgment of whether it is subjected to external traction. This provides direct differentiated data support for distinguishing the risk of dislodgement caused by two types of forces: fixation loosening and external traction, thereby improving the accuracy of subsequent status judgment.

[0108] In one exemplary embodiment, the local processor is specifically used for:

[0109] The received force data, electrical characteristic data, and liquid leakage data are input into a state prediction model trained based on machine learning algorithms to obtain the abnormal state of the indwelling needle.

[0110] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0111] For example, the local processor can be a miniaturized, low-power embedded system, which can be packaged as a standalone small clip-on device, connected to the sensor patch via flexible leads, or integrated into the edge of the patch.

[0112] After receiving the aforementioned data, the local processor first preprocesses the multimodal data, including filtering and noise reduction, which can be achieved using Kalman filtering to remove noise caused by the patient's slight tremors or external electromagnetic interference. Then, the data is timestamped and normalized to form standardized data frames, preparing for subsequent model inference. Simple calculations are then performed on the preprocessed data. For example, based on the magnitude and direction of tension collected by the stress sensor, the trend and rate of change of tension are determined; based on the conductivity collected by the bioimpedance sensor, the conductivity change around the indwelling needle caused by fluid leakage is determined; and based on the humidity around the skin puncture point of the indwelling needle collected by the humidity sensor, the humidity change around the skin puncture point of the indwelling needle is determined.

[0113] Here's an example to illustrate the process of determining early warning information using a state prediction model and an early warning model:

[0114] The system enters continuous monitoring mode, and the sensor patch collects three types of data in real time:

[0115] Force data: The first sensor unit monitors the pressure of the fixed wing on the skin; the second sensor unit monitors the magnitude and direction of the tension and shear force generated by the duct under traction.

[0116] Electrical characteristic data: The bioimpedance sensor monitors the electrical impedance value of the subcutaneous tissue below the puncture site, reflecting the conductivity around the indwelling needle.

[0117] Liquid leakage data: A humidity sensor monitors the humidity at the puncture site.

[0118] This raw data is transmitted to the local processor via a flexible circuit layer.

[0119] After receiving the three types of data, the local processor first performs filtering and noise reduction and feature extraction, then calculates the rate of change for each type of data to construct a multi-dimensional feature vector. This vector contains the following core features: static pressure attenuation index, dynamic tension integral value, tension direction entropy, impedance decrease rate, impedance fluctuation amplitude, and humidity step indicator. Specifically, the static pressure attenuation index is determined based on the pressure of the indwelling needle's fixation wings on the patient's skin, collected by the first sensor unit; the dynamic tension integral value and tension direction entropy are determined based on the shear force and tension generated by tubing traction and patient movement, collected by the second sensor unit; the impedance decrease rate and impedance fluctuation amplitude are determined based on the conductivity collected by the bioimpedance sensor; and the humidity step indicator is determined based on the humidity collected by the humidity sensor.

[0120] Among them, the static pressure decay index is used to reflect the long-term changing trend of the pressure of the fixed wing on the skin monitored by the first sensor unit; the tension integral value is used to reflect the magnitude of the cumulative tension on the catheter monitored by the second sensor unit in a recent period; the tension direction entropy is used to reflect the degree of disorder of the force direction change monitored by the second sensor unit; the impedance decrease rate is used to reflect how fast the subcutaneous tissue impedance value monitored by the bioimpedance sensor decreases over time; the impedance fluctuation amplitude is used to reflect the stability of the impedance value monitored by the bioimpedance sensor in a short period of time; and the humidity step indicator reflects whether the humidity at the puncture point monitored by the humidity sensor has suddenly increased, which can be set to a binary value.

[0121] The aforementioned feature vectors are input into a state prediction model deployed on a local processor. The state prediction model is constructed using machine learning algorithms suitable for small-sample, multi-classification tasks, such as gradient boosting trees, eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM); or support vector machines, such as Support Vector Machine (SVM).

[0122] In this embodiment, a lightweight model trained using the gradient boosting tree algorithm can be employed. This model consists of an ensemble of hundreds of decision trees, and the processing flow is as follows:

[0123] First, the feature vectors are simultaneously sent to all decision trees. Each tree, based on its own judgment logic, traverses different branches and finally outputs an initial score for a preset category at the leaf node. The preset categories include abnormal retention states of the indwelling needle and abnormal liquid leakage states. For example, a tree specifically for detecting leakage will sequentially check conductivity features (e.g., impedance decrease rate and impedance fluctuation amplitude) and humidity features (e.g., humidity step indicator) to determine the abnormal liquid leakage state. Similarly, a tree specifically for detecting displacement will sequentially check features corresponding to force data such as pressure, shear force, and tension (e.g., tension integral value, tension direction entropy, and static pressure decay index) to determine the abnormal retention state of the indwelling needle.

[0124] In practice, each decision tree, based on the input feature vector, judges the abnormal retention state of the indwelling needle or the abnormal liquid leakage state according to the aforementioned method, and determines the features present in each feature vector and the initial score of the preset category. For example, the score of the feature vector in the category of abnormal retention state of the indwelling needle is determined based on the tension integral value, tension direction entropy and static pressure decay index.

[0125] Secondly, the model sums the scores of all trees using weighted averages to obtain the raw cumulative scores for each category. The weighting coefficients are determined during the model training phase and reflect the importance of each tree in the final decision.

[0126] Finally, the raw scores can be grouped using the Softmax function to transform them into probability values ​​between 0 and 1, with a total sum of 100%. The model selects the category with the highest probability as the abnormal state of the indwelling needle at the current moment. For example, the output at a certain moment might be: normal catheter retention 5%, catheter displacement trend 85%, emergency dislodgement 10%; extravasation 18%, no leakage 92%. Then the model outputs: the predicted parameters for catheter displacement trend and the predicted parameters for leakage. Among them, normal catheter retention, catheter displacement trend, and emergency dislodgement are output items reflecting the abnormal retention state; extravasation and no leakage are output items reflecting the abnormal leakage state; the output items for abnormal retention state and abnormal leakage state together reflect the abnormal state of the indwelling needle.

[0127] In this embodiment, by inputting force data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle is determined. Then, the abnormal state of the indwelling needle is input into an early warning model to generate corresponding early warning information. This enables deep fusion and two-level intelligent processing of multimodal data. This two-level model architecture ensures both the accuracy of state identification and the refined classification of early warning information, effectively solving the technical problems of traditional monitoring methods' inability to provide early warnings and high false alarm rates. The prediction parameters are presented in data form, and the state prediction model outputs corresponding, intuitively suggestive early warning information based on the input prediction parameters.

[0128] In one exemplary embodiment, the local processor is further configured to:

[0129] Based on the early warning information of indwelling needle dislodgement and penetration, the early warning level of the early warning information of indwelling needle dislodgement and penetration is determined. The early warning level is used to characterize the urgency of the early warning information of indwelling needle dislodgement and penetration.

[0130] For example, the local processor determines the warning level based on the indwelling needle dislodgement and leakage warning information. This facilitates the subsequent comparison of the warning level and warning threshold by the central monitoring platform, thereby determining whether to issue a warning, and the corresponding warning form and scope. Specifically, the indwelling needle dislodgement and leakage warning information is used to determine the catheter retention status, catheter displacement trend, fluid extravasation data, and catheter dislodgement data. These data are then input into a preset processing model to obtain the final warning information and warning level. The catheter retention status, catheter displacement trend, and catheter dislodgement data can correspond to the aforementioned abnormal retention status, and the fluid extravasation data can correspond to the aforementioned abnormal fluid leakage status.

[0131] In this embodiment, the local processor further determines the corresponding warning level based on the indwelling needle dislodgement and extravasation warning information. The catheter retention status, catheter displacement trend, extravasation data, and catheter dislodgement data are mapped to warning levels of different urgency levels, which enables hierarchical management of abnormal indwelling needle status. This avoids frequent false alarms that interfere with medical staff and ensures that emergency events can be prioritized and responded to in a timely manner, thereby improving the efficiency of clinical monitoring and patient safety.

[0132] In one exemplary embodiment, a communication module and a central monitoring platform are also included, wherein:

[0133] The communication module is used to send early warning information about indwelling needle dislodgement and penetration, as well as the early warning level, to the central monitoring platform;

[0134] The central monitoring platform is used to display the received early warning information on indwelling needle dislodgement and infiltration, and to issue an alarm when the early warning level reaches the warning threshold.

[0135] Exemplary, exemplary Figure 4 This is another schematic diagram of the indwelling needle dislodgement and leakage early warning device in one embodiment. Figure 5 This is a schematic diagram of a central monitoring platform in one embodiment, such as... Figure 4 , Figure 5 As shown in the figure, the indwelling needle dislodgement and penetration early warning device proposed in this embodiment of the invention also includes a communication module and a central monitoring platform.

[0136] In practice, the warning information and warning level are sent to the communication module. Figure 5 The central monitoring platform shown can also transmit hospital Wi-Fi, alarm information, and real-time trend charts to the nurse station's large screen and mobile nursing system, enabling monitoring of the entire hospital area.

[0137] Abnormal status of the indwelling needle is transmitted to the early warning model, which determines the early warning level based on the status category and probability value, for example:

[0138] Blue alert: Multiple consecutive predictions of "catheter displacement trend" with a probability exceeding 50%; "no fluid leakage" with a probability exceeding 60%. The local processor emits a brief, gentle vibration to prompt the patient to adjust their position and avoid further traction. Simultaneously, this information is sent to the central monitoring platform, displayed as a blue "attention" icon in the corresponding bed's monitoring area.

[0139] Yellow alert: Predicted as "catheter displacement trend" with a probability exceeding 50%; "fluid extravasation" with a probability exceeding 60%, and the rate of impedance decrease exceeds the threshold. The local processor will emit intermittent, noticeable vibrations and beeps. The central monitoring platform will immediately display a yellow exclamation mark on the bed icon and pop up a prompt box, such as: Patient [Bed Number] - Risk of indwelling needle leakage, please check promptly.

[0140] Red Alert: Predicted as "emergency dislodgement," regardless of output probability, or "fluid extravasation," with a probability exceeding 70%. The local processor issues a continuous, rapid audible and visual alarm. The central monitoring platform issues a high-priority, continuously ringing alarm, and simultaneously forces the display of detailed information on the nurse station screen and the on-duty nurse's mobile terminal, such as: Patient [Bed Number] - Indwelling needle dislodged / serious leakage, please handle immediately!

[0141] In this embodiment, the communication module sends early warning information and warning levels regarding indwelling needle dislodgement and penetration to the central monitoring platform, which then displays the information visually. When the warning level reaches a threshold, an alarm is automatically triggered, enabling seamless integration from bedside monitoring to centralized monitoring at the nurses' station. In the event of a high-risk incident, the system proactively pushes an alarm, allowing nurses to promptly identify abnormalities without frequent rounds. This reduces the workload of nursing staff while ensuring that emergencies are detected and addressed immediately, significantly improving clinical monitoring efficiency and patient safety.

[0142] In one exemplary embodiment, the local processor can also convert the trained model into a lightweight format, deploy it in the microcontroller of the local processor, collect new clinical data periodically, retrain and optimize the model periodically, and update the local module via a wireless network to continuously improve the accuracy of the warning information.

[0143] At the same time, the system can identify a preset number of features that contribute more than a preset threshold to the current decision and generate a brief explanation, such as: main judgment criteria: excessively high tension integral value + excessively low directional entropy + pressure decay, to help nurses understand the source of risk.

[0144] A complete embodiment is provided to illustrate the indwelling needle dislodgement and leakage early warning device provided in the embodiments of this application.

[0145] After the nurse performs intravenous puncture and secures the indwelling needle according to standard procedures, she removes the sensing patch from this embodiment, replaces the ordinary dressing, and applies it flat to the puncture site. The patch's miniature interface is connected to a button-sized wireless transmitter module, which can be clipped onto the patient's gown.

[0146] After the system completes its self-test, the indicator light turns green. The nurse then uses a handheld PDA to bind the device to the bed.

[0147] When the patient frequently flexes and extends their wrist, the local processor analyzes the tension and finds that it changes periodically but does not have a continuous unidirectional trend, which is judged as low risk and no alarm is triggered.

[0148] If a patient unconsciously pulls on the catheter at night, the processor detects a unidirectional tension exceeding a threshold that lasts for several seconds, immediately triggering a local vibration to alert the patient to stop the action, and simultaneously sending a yellow warning message to the nurses' station.

[0149] If leakage occurs, the subcutaneous tissue resistance will continue to decrease within minutes. Even if there is no swelling, the system will issue a medium-level warning of "leakage risk," allowing nurses to intervene in a timely manner.

[0150] In this embodiment, force data, electrical characteristic data, and liquid leakage data are collected by a sensor patch and then fused and analyzed by a local processor. By monitoring the force and electrical characteristic data, the traction status and tip position of the catheter are perceived in real time, solving the problem that the indwelling needle is "invisible" under the skin and cannot be judged for displacement. By monitoring the liquid leakage data, rapid early warning is achieved in the early stage of drug leakage, overcoming the lag of traditional methods. At the same time, the comprehensive judgment of multiple parameters effectively avoids false alarms or missed alarms from single monitoring, significantly improving the accuracy of early warning of dislodgement and leakage risks.

[0151] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0152] Based on the same inventive concept, this application also provides a method for providing early warning of indwelling needle dislodgement and penetration in the aforementioned indwelling needle dislodgement and penetration early warning device. The solution provided by this method is similar to the solution described in the aforementioned device. Therefore, the specific limitations in one or more embodiments of the indwelling needle dislodgement and penetration early warning method provided below can be found in the limitations of the indwelling needle dislodgement and penetration early warning device described above, and will not be repeated here.

[0153] In one exemplary embodiment, such as Figure 6 As shown, a method for early warning of indwelling needle dislodgement and permeation is provided, including:

[0154] Step 602: Obtain the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0155] Step 604: Based on the stress data, electrical characteristic data, and liquid leakage data, predict the abnormal state of the indwelling needle;

[0156] Step 606: Based on the abnormal status of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0157] The indwelling needle dislodgement and penetration early warning method provided in this application embodiment can be implemented by a computer device. This computer device may include a terminal and a server. The terminal communicates with the server via a network. A data storage system can store the data that the server needs to process. The data storage system can be integrated on the server or placed in the cloud or on other network servers. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart TVs, projection devices, etc. Portable wearable devices can be smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0158] In this embodiment, by acquiring the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area, and predicting the abnormal state of the indwelling needle based on these multimodal data, and determining the corresponding dislodgement and leakage early warning information, it is possible to achieve comprehensive real-time monitoring of the indwelling needle status and provide timely and accurate decision-making basis for clinical intervention.

[0159] In one exemplary embodiment, predicting abnormal states of an indwelling needle based on force data, electrical characteristic data, and fluid leakage data includes:

[0160] By inputting the stress data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle can be obtained.

[0161] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage, including:

[0162] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0163] In this embodiment, by inputting force data, electrical characteristic data, and liquid leakage data into the state prediction model to obtain the abnormal state of the indwelling needle, and then inputting the abnormal state into the early warning model to obtain the final early warning information, the collaborative processing of the two-level models can be realized. This ensures the accuracy of state identification and achieves refined classification of early warning information, effectively improving the intelligence level of the system.

[0164] In one exemplary embodiment, it further includes:

[0165] Based on a pre-set database, the parameters of the state prediction model and the early warning model are periodically updated.

[0166] In this embodiment, the parameters of the state prediction model and the early warning model are updated periodically according to the preset database, which can continuously optimize the model parameters and thereby continuously improve the accuracy of identifying abnormal states of indwelling needles, effectively adapting to changes in different patient groups and application scenarios.

[0167] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0168] In an exemplary embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for early warning of needle dislodgement and penetration. The display unit of the computer device is used to form a visually visible image and may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0169] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0170] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0171] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0172] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0173] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0174] By inputting the stress data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle can be obtained.

[0175] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0176] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0177] Based on a pre-set database, the parameters of the state prediction model and the early warning model are periodically updated.

[0178] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0179] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0180] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0181] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0182] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0183] By inputting the stress data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle can be obtained.

[0184] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0185] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0186] Based on a pre-set database, the parameters of the state prediction model and the early warning model are periodically updated.

[0187] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0188] Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area;

[0189] Predict abnormal conditions of indwelling needles based on stress data, electrical characteristic data, and fluid leakage data;

[0190] Based on the abnormal condition of the indwelling needle, determine the early warning information for needle dislodgement and leakage.

[0191] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0192] By inputting the stress data, electrical characteristic data, and liquid leakage data into a state prediction model trained based on machine learning algorithms, the abnormal state of the indwelling needle can be obtained.

[0193] The abnormal status of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

[0194] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0195] Based on a pre-set database, the parameters of the state prediction model and the early warning model are periodically updated.

[0196] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0197] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0198] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0199] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A device for early warning of indwelling needle dislodgement and leakage, characterized in that, Includes sensor patches and a local processor, wherein: The sensing patch is connected to the local processor and is used to monitor the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area, and to send the force data, the electrical characteristic data and the liquid leakage data to the local processor. The local processor is used to predict abnormal states of the indwelling needle based on the received force data, electrical characteristic data, and liquid leakage data, and to determine early warning information for indwelling needle dislodgement and leakage based on the abnormal states of the indwelling needle.

2. The apparatus as claimed in claim 1, characterized in that, The sensing patch includes an adhesive layer, a circuit layer, and a sensor layer, wherein: The adhesive layer, located at the bottom of the sensing patch, is used to adhere the sensing patch to the patient's skin surface; The circuit layer, located above the adhesive layer, is used to carry and connect the sensors of the sensor layer. The sensor layer, embedded in the circuit layer, is used to collect the force data, the electrical characteristic data, and the liquid leakage data.

3. The apparatus as described in claim 2, characterized in that, The sensor layer includes a stress sensor, a bioimpedance sensor, and a humidity sensor, wherein: The stress sensor is used to obtain force data by monitoring the pressure exerted on the patient's skin by the fixation wings of the indwelling needle in the first target area, and the magnitude and direction of the tension generated by the patient's movements; The bioimpedance sensor is used to obtain electrical characteristic data by monitoring the conductivity of liquid permeation in the second target region; The humidity sensor is used to obtain liquid leakage data by monitoring the humidity at the skin puncture point in the third target area.

4. The apparatus as described in claim 3, characterized in that, The stress sensor includes a first sensor unit and a second sensor unit, wherein: The first sensor unit is located in the projection area where the fixing wing of the indwelling needle contacts the patient's skin, and is used to monitor the pressure exerted by the fixing wing on the patient's skin; The second sensor unit is located in the stress concentration area of ​​the indwelling needle catheter on the patient's skin and is used to monitor the magnitude and direction of tension generated by the traction of the indwelling needle catheter and the patient's movements.

5. The apparatus as claimed in claim 1, characterized in that, The local processor is specifically used for: The received force data, electrical characteristic data, and liquid leakage data are input into a state prediction model trained based on a machine learning algorithm to obtain the abnormal state of the indwelling needle. The abnormal state of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

6. The apparatus as claimed in claim 1, characterized in that, The local processor is also used for: Based on the indwelling needle dislodgement and penetration warning information, the warning level of the indwelling needle dislodgement and penetration warning information is determined, and the warning level is used to characterize the urgency of the indwelling needle dislodgement and penetration warning information.

7. The apparatus as claimed in claim 6, characterized in that, It also includes a communication module and a central monitoring platform, among which: The communication module is used to send the early warning information of the indwelling needle dislodgement and penetration, and the early warning level, to the central monitoring platform; The central monitoring platform is used to display the received early warning information on needle dislodgement and penetration, and to issue an alarm when the early warning level reaches the early warning threshold.

8. A method for early warning of indwelling needle dislodgement and permeation, characterized in that, include: Acquire the force data of the first target area, the electrical characteristic data of the second target area, and the liquid leakage data of the third target area; Based on the stress data, electrical characteristic data, and liquid leakage data, predict abnormal states of indwelling needles; Based on the abnormal state of the indwelling needle, early warning information for needle dislodgement and leakage is determined.

9. The method as described in claim 8, characterized in that, Based on the stress data, electrical characteristic data, and fluid leakage data, predict abnormal states of the indwelling needle, including: The force data, electrical characteristic data, and liquid leakage data are input into a state prediction model trained based on a machine learning algorithm to obtain the abnormal state of the indwelling needle. The step of determining early warning information for indwelling needle dislodgement and leakage based on the abnormal state of the indwelling needle includes: The abnormal state of the indwelling needle is input into an early warning model trained based on a machine learning algorithm to obtain early warning information on needle dislodgement and penetration.

10. The method as described in claim 9, characterized in that, Also includes: Based on a preset database, the parameters of the state prediction model and the early warning model are periodically updated.

Images