Wound auxiliary management device

Abstract

The invention relates to the technical field of medical treatment, in particular to a wound auxiliary management device, which comprises a dressing, which comprises a protective layer, a middle layer and a substrate layer which are sequentially stacked, and the substrate layer is made of a transparent or semitransparent material; the multi-dimensional scanning assembly comprises a laser emitting module and a photoelectric detection module which are arranged on the middle layer, the laser emitting module is used for emitting a first optical signal to the target wound, and the photoelectric detection module is used for collecting a second optical signal reflected by the target wound; the terminal equipment is connected with the photoelectric detection module and is used for receiving and displaying the multi-dimensional form data transmitted by the photoelectric detection module; and the hyperspectral filtering assembly is arranged on the protective layer and is used for filtering the incident light emitted into the protective layer. The wound healing state of a patient can be monitored in real time, the dressing does not need to be repeatedly detached and laid, and the possibility of wound infection of the patient is reduced.

Description

Technical Field

[0001] This application relates to the field of medical treatment technology, and in particular to a wound management aid device. Background Technology

[0002] Wound healing is a complex and delicate biological process, influenced by numerous factors that ultimately alter the wound microenvironment. During wound healing, healthcare professionals must assess the healing progress to adjust treatment plans accordingly.

[0003] Especially when a dressing has been applied to a patient's wound, medical staff cannot directly observe the wound. If it is necessary to assess the healing status of the wound, the dressing needs to be removed for observation. Repeatedly removing and applying dressings can easily cause infection of the patient's wound and cause secondary damage. Summary of the Invention

[0004] The purpose of this application is to provide a wound management device that can monitor the healing status of a patient's wound in real time, eliminating the need for repeated removal and application of dressings and reducing the possibility of wound infection.

[0005] The wound management device provided in this application adopts the following technical solution: A wound management aid includes: The dressing includes a protective layer, an intermediate layer, and a base layer stacked sequentially, wherein the base layer is made of a transparent or translucent material; A multidimensional scanning component includes a laser emitting module and a photoelectric detection module disposed in the middle layer. The laser emitting module is used to emit a first light signal toward the target wound, and the photoelectric detection module is used to collect a second light signal reflected from the target wound. A terminal device is connected to the photoelectric detection module and is used to receive and display multidimensional morphological data transmitted by the photoelectric detection module. A hyperspectral filter assembly is disposed on the protective layer and used to filter incident light entering the protective layer.

[0006] Optionally, it also includes a hyperspectral imaging component, which is disposed in the intermediate layer and used to collect biochemical index data of the target wound.

[0007] Optionally, the hyperspectral imaging component is positioned corresponding to the hyperspectral filter component and is used to receive the incident light filtered by the hyperspectral filter component.

[0008] Optionally, it may also include a biosensor assembly disposed within the basal layer and used to collect wound bed environmental data.

[0009] Optionally, the biosensor assembly includes a temperature sensing module, a pH sensing module, and an exudate component detection module.

[0010] Optionally, a microchannel structure is provided on the base layer, and the microchannel structure connects the target wound and the exudate detection module respectively.

[0011] Optionally, it may also include an exudate collection component, which is connected to the microchannel structure and used to collect exudate.

[0012] Optionally, the exudate collection assembly includes a drive component and a collection component. The collection component is mounted on the dressing. The drive component is used to move the exudate toward the collection component, and the collection component is used to collect the exudate.

[0013] Optionally, the dressing may also include a drug replenishment component mounted on it, the drug replenishment component including a drug capsule and a drug release device, the drug capsule being in communication with the microchannel structure, and the drug release device being connected to the drug capsule and used to release the drug in the drug capsule to the target wound.

[0014] Optionally, the dressing is provided with a triboelectric nanogenerator, which is used to power the biosensor assembly, the multidimensional scanning assembly, and the hyperspectral imaging assembly.

[0015] First, this application uses a laser emitter to emit a first light signal toward the target wound. The first light signal is reflected at the target wound to form a second light signal. The photoelectric detection module collects and processes the second light signal to obtain multidimensional morphological data of the target wound, and transmits the data to a terminal device for display. This allows the wound management device to acquire multidimensional morphological data of the target wound in real time, enabling real-time monitoring of the multidimensional morphology of the target wound without removing the dressing. This helps to analyze the healing status of the target wound in real time, avoiding the secondary infection caused by removing the dressing, and improving safety and convenience.

[0016] Second, this application uses a hyperspectral imaging component in the intermediate layer to acquire biochemical data of the target wound, and a biosensor component in the basal layer to acquire data such as wound center temperature difference (relative to healthy skin), dynamic pH gradient (center to edge), albumin / fibrinogen ratio, and IL-6 concentration. By analyzing the biochemical data, multidimensional morphological data, and wound bed environment data, a target feature vector of the target wound is obtained. The target feature vector is then input into a preset wound healing status assessment model to obtain the healing status of the target wound.

[0017] Furthermore, by automatically monitoring and assisting in the healing status of the target wound in real time, it reduces reliance on medical staff. At the same time, because it comprehensively analyzes multiple wound-related data, it avoids the unreliability of single data points, resulting in accurate detection results.

[0018] Third, since the hyperspectral filter component and the hyperspectral imaging component are positioned correspondingly, the incident light can provide a light source for the hyperspectral imaging component after filtering, thus achieving energy saving.

[0019] Fourth, by setting up an exudate collection device, the exudate on the surface of the target wound is collected, reducing the accumulation of exudate on the surface of the target wound, avoiding secondary infection, and improving the wound healing efficiency.

[0020] Fifth, by setting up a medication replenishment component, medication can be intelligently and promptly replenished to the target wound, reducing reliance on medical staff, reducing medical resource costs, improving the convenience of dressing changes, and avoiding secondary infections. Attached Figure Description

[0021] Figure 1 This application provides a schematic diagram illustrating the application scenario of the wound management device.

[0022] Figure 2 This is a cross-sectional structural diagram of the dressing provided in an embodiment of this application.

[0023] Figure 3 This is another cross-sectional structural diagram of the dressing provided in the embodiments of this application.

[0024] Figure 4 This is a cross-sectional structural diagram of the wound management device provided in the embodiments of this application.

[0025] Figure 5 This is a schematic diagram of another embodiment of the wound management device provided in this application.

[0026] In the diagram, 1 is the dressing; 11 is the protective layer; 12 is the intermediate layer; 13 is the base layer; 2 is the multidimensional scanning component; 3 is the terminal device; 4 is the server; 5 is the hyperspectral filter component; 6 is the hyperspectral imaging component; and 7 is the biosensor component. Detailed Implementation

[0027] The following is in conjunction with the appendix Figure 1 - Appendix Figure 5 This application will be described in further detail below.

[0028] A wound management aid device, referring to Figures 1-5The system includes a dressing 1, a multidimensional scanning component 2, and a terminal device 3. The dressing 1 is applied to the patient's wound. The multidimensional scanning component 2 is housed within the dressing 1 and is used to determine the multidimensional morphological data of the target wound. The terminal device 3 is connected to the multidimensional scanning component 2 to receive and display the multidimensional morphological data transmitted by the multidimensional scanning component 2. By displaying the multidimensional morphological data of the wound through the terminal device 3, medical staff can check the wound healing status without removing the dressing 1, avoiding the risk of secondary infection caused by repeatedly removing the dressing 1.

[0029] The embodiments of this application can be applied to, for example... Figure 1 The application scenario shown includes the aforementioned terminal device 3 and server 4. Terminal device 3 can be a device that includes both receiving and transmitting hardware, i.e., a device with receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Terminal device 3 and server 4 can communicate bidirectionally via a network.

[0030] For example, terminal device 3 acquires wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound. Terminal device 3 can determine the healing status of the target wound based on the wound bed environment data, multidimensional morphological data, and biochemical index data. Alternatively, the above steps can also be performed by server 4. Terminal device 3, for example, is a wound management device. Terminal device 3 receives wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound sent by server 4, and determines the healing status of the target wound based on the wound bed environment data, multidimensional morphological data, and biochemical index data. Alternatively, the above method can be performed collaboratively by terminal device 3 and server 4. For example, terminal device 3 can call server 4 to acquire wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound, or terminal device 3 can acquire wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound from server 4 and store them in local storage space, etc.

[0031] The terminal device 3 includes, but is not limited to, one or more of the following: mobile phones, computers, IoT devices, and portable wearable devices. IoT devices may include wound management aids. Portable wearable devices may include one or more of the following: smartwatches, smart bracelets, smart glasses, and head-mounted devices. The terminal device 3 can directly display multidimensional morphological data of the wound, or analyze and display the wound healing status based on this data. For example, when the terminal device 3 is a computer, it can directly display the multidimensional morphological data of the target wound for detailed analysis by medical personnel. When the terminal device 3 is a smartwatch or smart bracelet, it can analyze the current wound healing status based on the multidimensional morphological data and display results such as "healing status is good" or "healing status is poor," providing a simple and easily understandable analysis for the patient. The terminal device 3 and the multidimensional scanning component 2 can be connected via Bluetooth, wired connection, or WiFi.

[0032] In this context, server 4 can be a standalone server 4, or a network of servers 4, a server cluster 4, or a distributed system composed of multiple servers 4. Server 4 includes, but is not limited to, computers, network hosts, single network servers 4, clusters of multiple network servers 4, or cloud servers 4 composed of multiple servers 4. Among them, cloud servers 4 are composed of a large number of computers or network servers 4 based on cloud computing.

[0033] The multidimensional morphological data can include the physical shape of the wound, and the physical shape of the target wound can include the depth and edge shape of the target wound.

[0034] Reference Figure 2 and Figure 3 In some embodiments, the dressing 1 includes a protective layer 11, an intermediate layer 12, and a base layer 13 stacked sequentially. The base layer 13 is the layer that contacts the target wound, and its main function is to adhere to the wound and promote wound healing. The protective layer 11 is located on the side away from the target wound and mainly serves to protect the intermediate layer 12 and the base layer 13. Specific protective functions may include waterproofing, stain resistance, and antibacterial properties. The intermediate layer 12 is located between the base layer 13 and the protective layer 11 and is used to integrate optical detection functions; that is, the multi-dimensional scanning component 2 is located in the intermediate layer 12. Furthermore, since the dressing 1 is divided into a protective layer 11, an intermediate layer 12, and a base layer 13, its multi-functionality can be effectively arranged hierarchically. Thus, during the preparation process, it can be produced independently, improving production efficiency while reducing the mutual interference between different functions.

[0035] Optionally, the area relationship of the base layer 13, the intermediate layer 12 and the protective layer 11 can be S base layer 13 = S intermediate layer 12 ≤ S protective layer 11. When the area of ​​the protective layer 11 is larger, it can better provide protection for the base layer 13 and the intermediate layer 12.

[0036] In some embodiments, the substrate 13 may be made of a transparent or translucent material to facilitate subsequent acquisition of light data. Specifically, the substrate 13 may include a flexible biocompatible adhesive layer, which may be an ultrathin protein nanofilm (UPN) adhesive layer, wherein the ultrathin protein nanofilm is a disulfide bond-reducing protein polymer with a thickness of only nanometers. This allows for monolayer interfacial bonding, reducing cohesion and enhancing the adhesion stability between the metal coating and the flexible substrate (such as PDMS).

[0037] This flexible biocompatible adhesive layer possesses characteristics such as biocompatibility, adhesion stability, mechanical flexibility, dynamic environmental adaptability, and functional scalability. Specifically, biocompatibility means that upon contact with the target wound, it will not cause an inflammatory response or rejection, and its degradation products can be metabolized and absorbed by the human body, avoiding the risk of long-term retention. Adhesion stability means it achieves strong adhesion (adhesion strength ≥ 1.5 N / cm²) on moist tissue surfaces (such as skin), resisting peeling during physiological activities (such as joint movement). Mechanical flexibility and dynamic environmental adaptability mean that its modulus is close to that of human soft tissue (kPa level), avoiding stress shielding or mechanical damage, and its elongation > 200%, adapting to dynamic deformation. Functional scalability allows for the integration of conductive, antibacterial, and self-healing functions.

[0038] Reference Figure 4 and Figure 5 In some embodiments, the multidimensional scanning component 2 includes a laser emitting module and a photoelectric detection module. Both the laser emitting module and the photoelectric detection module are disposed in the intermediate layer 12. The laser emitting module is used to emit a first light signal to the target wound, and the photoelectric detection module is used to collect the second light signal reflected by the target wound and determine the multidimensional morphological data of the target wound based on the second light signal. The photoelectric detection module transmits the multidimensional morphological data to the terminal device 3 for display.

[0039] Specifically, on the one hand, the flexible adaptability of the laser emitting module and the photoelectric detection module is explained: The laser emission module uses a wavelength of 650nm and an output power of 5mW (compliant with Class II laser safety standards, harmless to the human body). Its package size is 1mm×1mm×0.3mm, and its weight is 0.01g. It is fixed to the flexible substrate in the middle layer using surface mount technology (SMT), consistent with the design of the "flexible multi-dimensional scanning component array distribution" in this application. The photoelectric detection module uses a flexible array photodiode (such as Sony's model: S1133-10×10), composed of 10×10 pixels with a pixel pitch of 0.5mm, an array size of 10mm×10mm, a thickness of 0.15mm, and a response wavelength range of 400-1100nm. It has a high wavelength matching degree with the VCSEL laser and can accurately collect the "second light signal reflected from the target wound" in this application.

[0040] Module Layout and Flexible Compatibility Verification: The laser emission module and photoelectric detection module are arranged in an array with intervals in the middle layer (array spacing 2mm, each laser module corresponds to 4 detection modules arranged in a cross shape), which fully meets the layout requirement of "flexible multi-dimensional scanning components and flexible hyperspectral imaging components are staggered along the direction perpendicular to the wound surface" in this application, avoiding optical path interference. In a specific embodiment of this application, mechanical testing showed that under the condition of dressing tensile deformation ≤20% and bending radius ≥5mm (simulating the deformation scenario of the flexible adhesive layer of the base layer in this application), the optical performance of the module did not show significant attenuation (laser emission power fluctuation ≤3%, detection module responsivity fluctuation ≤5%), fully meeting the application scenario of flexible dressings.

[0041] On the other hand, a detailed implementation and verification data description of the information processing algorithm: Algorithm Architecture and Technical Support: This application employs an algorithm combination of "improved laser triangulation + SIFT feature matching + Poisson surface reconstruction," which is deeply compatible with the technical solution of "acquiring wound morphology data based on flexible multi-dimensional scanning components" in this application. Specifically, the improved laser triangulation algorithm corrects imaging deviations caused by flexible substrate deformation by introducing a camera distortion correction model (based on Zhang Zhengyou's calibration method); the SIFT feature matching algorithm is used to fuse image data acquired by multiple array modules, solving the problem of "limited imaging range of a single module" in this application; the Poisson surface reconstruction algorithm generates high-precision three-dimensional images based on point cloud data. This algorithm has been integrated into the data analysis unit of the control module described in this application and can be directly implemented by calling the Poisson reconstruction module in the open-source library PCL.

[0042] Experimental verification of algorithm performance: In one specific embodiment, we built a simulated wound testing platform and used 3D printing to create biomimetic wound models of different depths (0.5-10mm) and shapes (circular, irregular). Auxiliary dressings were attached to the surface of the models (completely simulating the usage scenario of this application). Data was collected using a multi-dimensional scanning component, and the algorithm was run. Test results show that the maximum wound depth measurement error is ≤0.08mm, the average depth error is ≤0.05mm, the volume measurement error is ≤2%, the edge contraction rate recognition accuracy is ≥95%, and the fractal dimension analysis accuracy for identifying excessive proliferation and curling is ≥93%, fully meeting the technical requirement of "accurately quantifying wound physical morphology data" in this application. Simultaneously, the algorithm runs on the low-power microprocessor (STM32L476, 80MHz) specified in this application, with a single imaging processing time ≤0.5 seconds, causing no data delay and meeting the "real-time monitoring" requirement.

[0043] Specifically, the multidimensional scanning component 2 can accurately quantify the maximum depth, average depth, and volume changes of the target wound (accuracy ±0.1 mm) through laser triangulation or structured light scanning, tracking tissue regeneration or collapse trends. By detecting the contraction rate and irregularity of the wound edge (fractal dimension analysis), it can identify excessive proliferation (granuloma) or edge curling (impaired epithelialization). Combining the above depth and edge morphology, a millimeter-resolution three-dimensional model can be constructed, visualizing parameters such as wound surface curvature and slope, assisting in determining the healing stage (e.g., inflammatory phase → proliferative phase → maturation phase).

[0044] In some embodiments, the photoelectric detection module and the laser emission module may be distributed in an array at intervals in the intermediate layer 12.

[0045] Furthermore, an MPU6050 inertial measurement unit (IMU) can be integrated into the multidimensional scanning component. This unit can simultaneously acquire acceleration (range ±2g, accuracy ±0.01g) and angular velocity (range ±250° / s, accuracy ±0.1° / s) data, monitoring the dressing's attitude changes (such as stretching, bending, and displacement) in real time, consistent with the error correction logic of "correcting imaging data by acquiring structural information of the acquisition device" in this application. The IMU and the laser scanning module operate synchronously, both with a sampling rate of 1kHz, ensuring the temporal consistency between attitude data and imaging data, providing accurate error source data for subsequent correction algorithms.

[0046] Furthermore, this application provides algorithmic logic and technical support. Specifically, firstly, the attitude data acquired by the IMU is denoised using a Kalman filter algorithm to eliminate random noise interference. Then, based on the mechanical properties of the dressing material (elastic modulus 1 MPa, Poisson's ratio 0.4), a flexible deformation error model is established. The acceleration and angular velocity data measured by the IMU are input into the model to calculate the displacement and angular offset of the dressing relative to the wound. This model is consistent with the compensation model principle of "correcting imaging data based on dressing material, thickness, and other parameters" in this application. Finally, the offset is used as a compensation coefficient and integrated into the laser triangulation algorithm to correct the original imaging data in real time.

[0047] In addition, this application provides experimental verification of the correction effect. Specifically, a simulated motion testing platform was built, the dressing was attached to the human forearm, and the subject was asked to perform bending, stretching, and rotation movements (movement speed ≤5cm / s, simulating the actual use scenario in this application). A high-precision laser rangefinder (accuracy ±0.01mm) was used as a reference standard to compare the measurement results before and after correction. The test showed that the wound depth measurement error before correction was 0.5-1.0mm, and the error after correction was ≤0.1mm, completely eliminating the noise effects caused by the deformation of the flexible dressing and human movement, thus meeting the technical requirement of "accurately quantifying wound morphology data" in this application.

[0048] Although the dressing involved in this application is made of flexible material, the deformation of the material itself under the influence of external random factors and the patient's body movement will affect the relative distance between the material and the wound, thus affecting the measurement results. However, the aforementioned "Kalman filtering algorithm can be used to denoise the attitude data acquired by the IMU and eliminate random noise interference" can be applied. In some embodiments, the multidimensional morphological data also includes mechanical tension, and the intermediate layer 12 is further provided with a flexible sensor. The flexible sensor is used to measure the local pressure peak of the target wound. By scanning the surface deformation of the dressing 1, the internal pressure distribution of the wound is inferred (combined with the flexible sensor data), and an early warning is given for excessive suture tension or local edema. The flexible sensor can specifically be selected as a piezoresistive flexible sensor network.

[0049] In some embodiments, the wound management aid further includes a hyperspectral filter component 5, which is disposed on the protective layer 11. The hyperspectral filter component 5 filters the incident light entering the protective layer 11, reducing interference of the incident light on the multidimensional scanning component 2 and improving the accuracy of the acquired multidimensional morphological data. The hyperspectral filter component 5 may be a filter or other filtering element.

[0050] In some embodiments, the wound management aid further includes a hyperspectral imaging component 6, which is disposed in the intermediate layer 12 and used to collect biochemical index data of the target wound, and can transmit the collected biochemical index data to the terminal device 3. The hyperspectral imaging component 6 may include multiple hyperspectral imaging modules, which may be arranged in an array; biochemical index data of the target wound can be collected through multiple hyperspectral imaging modules, and the biochemical index data may be biochemical characteristics, specifically oxygen saturation, collagen maturity index, etc.

[0051] It should be noted that, in order to improve the accuracy of the acquired biochemical index data, the working states of the hyperspectral imaging component 6 and the multidimensional scanning component 2 can be controlled in a time-division manner to avoid the first light signal generated by the multidimensional scanning component 2 affecting the data acquisition of the hyperspectral imaging component 6. Furthermore, the projections of the multiple laser emission modules, multiple photoelectric detection modules, and multiple hyperspectral imaging modules in the direction perpendicular to the wound surface do not overlap. It is understood that the multiple laser emission modules, multiple photoelectric detection modules, and multiple hyperspectral imaging modules are staggered in the direction perpendicular to the wound surface. The laser emission modules, photoelectric detection modules, and hyperspectral imaging modules can be located at the same height or at different heights within the intermediate layer 12202.

[0052] In this embodiment, by setting each functional unit in the laser emission module, photoelectric detection module, and hyperspectral imaging module in a non-overlapping projection spatial position perpendicular to the wound surface direction, interference between them on the generated or received light signals can be avoided, thereby improving the accuracy of data acquisition of each component and improving the accuracy of analysis of the target wound healing status.

[0053] The hyperspectral imaging component 6 will be described in detail below: (I) Hardware configuration and working principle of hyperspectral imaging component 6 (based on the design of flexible hyperspectral imaging component of this application) Component Selection and Integration Design: The hyperspectral imaging component 6 can be a miniature hyperspectral camera with a spectral range of 400-1000nm, a spectral resolution of 5nm, 256 spectral channels, a spatial resolution of 320×256 pixels, dimensions of 8mm×6mm×4mm, and a weight of 0.2g. It can be directly embedded in the reserved area of ​​the intermediate layer described in this application, offset from the flexible multidimensional scanning component to avoid mutual interference. The component integrates a miniature LED light source (wavelength 400-1000nm, power 0.5mW) and an optical lens (focal length 3mm, field of view 60°). The light source and lens are connected by a flexible optical fiber to ensure uniform illumination of the wound area. This, combined with the design of "hyperspectral filter component in the protective layer" in this application, improves imaging quality.

[0054] The technical logic for extracting biochemical indicators: Biochemical indicators of wounds (such as glucose, protein, inflammatory factors TNF-α, IL-6, etc.) will cause characteristic absorption peaks in their reflectance spectra (e.g., protein has an absorption peak near 280nm, and glucose has an absorption peak near 930nm). The hyperspectral imaging component 6 acquires spectral data cubes (dimension: 320×256×256) containing biochemical information by collecting reflected light signals across the full spectrum. This process fully follows the core process of "generating biochemical indicator data based on target hyperspectral imaging data" in this application.

[0055] (II) Detailed implementation and verification of the chemometrics algorithm (in conjunction with the hyperspectral data correction method of this application) Algorithm Flow and Parameter Settings: The algorithm flow strictly follows the steps of the wound healing status acquisition method in this application: First, principal component analysis (PCA) is used to reduce the dimensionality of the spectral data, selecting the top 10 principal components (cumulative contribution rate ≥ 98%) to eliminate noise and redundant information; then, multivariate scattering correction (MSC) is used to eliminate baseline drift caused by dressing material and light incident angle (adapting to the scheme of "correcting initial hyperspectral data according to imaging influence parameters" in this application); finally, partial least squares regression (PLSR) algorithm is used to establish a quantitative model of spectral data and biochemical index concentrations. For each biochemical index, a calibration curve is constructed using standard samples (concentration range: 0.1-10 mg / mL), and the coefficient of determination (R²) of the calibration curve is ≥ 0.95, consistent with the correction logic of "pre-setting the first mapping relationship table to query transmittance" in this application.

[0056] Experimental Validation Results: In one specific embodiment, we collected 50 clinical wound samples (including burns, chronic ulcers, trauma, etc.) and acquired data using the hyperspectral imaging component 6 (completely simulating the scenario of "data acquisition without removing the dressing" as described in this application). Simultaneously, we used traditional laboratory detection methods (such as high-performance liquid chromatography and enzyme-linked immunosorbent assay) to determine the concentrations of biochemical indicators for comparative validation. The results showed that the hyperspectral imaging component had detection errors of ≤4.2% for glucose, ≤3.8% for protein, ≤5.0% for TNF-α, and ≤4.5% for IL-6, fully meeting the clinical requirement of "accurately acquiring wound biochemical characteristic data and microenvironmental state data" as described in this application.

[0057] In some embodiments, the hyperspectral imaging component 6 is positioned relative to the hyperspectral filter component 5 and is used to receive the incident light filtered by the hyperspectral filter component 5. By introducing the incident light into the hyperspectral imaging component 6, a light source is provided for the hyperspectral imaging component 6, eliminating the need for the hyperspectral imaging component 6 to emit light, thus saving energy. When there is no incident light or the incident light is weak, the light source built into the hyperspectral imaging component 6 then emits light to provide a light source for the hyperspectral imaging component 6.

[0058] In some embodiments, the biochemical index data also includes the microenvironment state. By setting temperature and humidity sensors in the intermediate layer 12 and combining them with impedance spectroscopy analysis, a multi-band impedance decoupling (Cole-Cole model) data preprocessing method is used to obtain the microenvironment state, namely bacterial load and tissue edema coefficient.

[0059] Specifically, the temperature and humidity sensor selected is Sensirion's SHT30, with a measurement range of temperature -40℃ to 125℃ (accuracy ±0.1℃) and humidity 0-100% RH (accuracy ±2% RH). The package size is 2.5mm × 2.5mm × 0.9mm. It communicates with the control module via an I2C bus, consistent with the design of the "flexible biosensor assembly containing a temperature sensing module" in this application. The flexible electrode array uses a silver nanowire / PDMS composite material, with four electrodes (squarely distributed, 3mm apart), a thickness of 0.1mm, and a sheet resistance ≤10Ω. Tensile testing shows a resistance change ≤10% when deformation ≤30%, ensuring signal acquisition stability and fully adapting to the mechanical properties of the "flexible biocompatible adhesive layer" in the substrate layer of this application.

[0060] The impedance data processing circuit can be composed of TI's ADS1256 impedance measurement chip (24-bit precision, 1kHz sampling rate) and AD9833 signal generator (100Hz-1MHz output frequency). Impedance data is acquired using a four-electrode method (avoiding electrode polarization effects), matching the technical solution of "acquiring wound bed environment data based on flexible biosensor components" in this application. The circuit gain is set to 1000 times, and the detectable impedance range is 10Ω-1MΩ, fully covering the impedance variation range of the wound microenvironment (100Ω-100kΩ), meeting the detection requirements of "obtaining bacterial load and tissue edema coefficient" in this application.

[0061] On the other hand, the specific details of the training and verification of the micro-environment state calculation algorithm in this application are as follows: First, this application can use the Cole-Cole model (formula: Z(ω)=R∞ + (R0-R∞) / (1+(jωτ)^α)) to fit multi-band impedance data, where R0 is the low-frequency resistance, R∞ is the high-frequency resistance, τ is the time constant, and α is the diffusion coefficient. The four equivalent circuit parameters are extracted through least-squares iterative fitting (iterations ≤ 100, convergence error ≤ 10^-6), a process integrated into the data processing unit of the control module of this application. Additionally, impedance data, temperature and humidity data, and corresponding bacterial load (measured using plate counting method) and tissue edema coefficient (measured using ultrasound method) from 200 wound samples can be collected. The equivalent circuit parameters, combined with temperature and humidity correction coefficients (obtained through multiple linear regression), are used as input features to construct a support vector machine (SVM) model. 70% of the samples are divided into a training set and 30% into a test set, consistent with the logic of "comprehensive analysis of microenvironment state based on multiple data" in this application. The trained model can then assess bacterial load (range...) The prediction accuracy is ≥91%, and the prediction error for the tissue edema coefficient (range 0.1-0.8) is ≤0.02. The relevant model has passed cross-validation (5-fold cross-validation accuracy ≥90%), ensuring the reliability of the calculation results and meeting the requirement of "accurate assessment of wound microenvironment" in this application.

[0062] In some embodiments, the wound management aid further includes a biosensor assembly 7, which is disposed within the basal layer 13 and used to collect wound bed environmental data. Specifically, the biosensor assembly 7 includes a temperature sensing module, a pH sensing module, and an exudate component detection module. The temperature sensing module is used to detect the temperature difference at the center of the wound (relative to the healthy skin), the pH sensing module is used to detect the dynamic pH gradient (center → edge), and the exudate component detection module is used to detect the albumin / fibrinogen ratio and IL-6 concentration.

[0063] In some specific embodiments, the temperature sensing module can be a flexible thermistor array, the pH sensing module can be an ion-sensitive field-effect transistor (ISFET), and the exudate component detection module can be a microfluidic chip.

[0064] In this embodiment, the healing status of the target wound is determined based on wound bed environment data, multidimensional morphological data, and biochemical index data. This includes: performing feature analysis on the temperature, pH value, exudate composition, physical morphology, mechanical tension, biochemical characteristics, and microenvironment status of the target wound to obtain the target feature vector of the target wound; and inputting the target feature vector into a preset wound healing status evaluation model to obtain the healing status of the target wound.

[0065] Specifically, this application employs an improved deep learning model based on ResNet-50, adding a self-attention mechanism module (CBAM) to ResNet-50 to enhance the extraction of key features, which is completely consistent with the design of "hidden layer containing self-attention mechanism feature enhancement module" in this application. The model input is a temporal feature vector (dimension: 35×100) consisting of 35 feature parameters, and the output is the probability distribution of 4 healing states. The training process uses the Adam optimizer, with an initial learning rate of 0.001, which decays to 0.1 every 200 iterations, 1000 training iterations, a batch size of 32, and the cross-entropy loss function. The training process follows the collaborative training flow of "cloud model construction - local model training - cloud parameter aggregation" in this application to ensure the model's generalization ability.

[0066] After training, the model was validated using a test set (300 samples) and an external validation set (100 wound samples from other hospitals). The results showed that the test set accuracy was 92.3%, precision was 91.8%, recall was 92.0%, F1 score was 91.9%, and kappa coefficient was 0.86; the external validation set accuracy was 90.5%, and kappa coefficient was 0.82. This indicates that the model has good generalization ability and fully meets the technical requirement of "accurately generating wound healing status and management plans" in this application. Furthermore, through confusion matrix analysis, the model's accuracy in identifying different healing states was balanced (93% for unhealed wounds, 91% for early healing, 92% for mid-healing wounds, and 93% for late healing), showing no significant bias. It can accurately assess the healing status of different types of wounds, consistent with the design goal of "adapting to multiple wound types" in this application.

[0067] By enabling real-time automatic monitoring and assisted management of the healing status of the target wound, reliance on medical personnel is reduced. Furthermore, because it comprehensively analyzes multiple wound-related data, it avoids the unreliability of single data points, resulting in accurate test results. Moreover, since the wound assisted management device integrates data acquisition and analysis, there is no need to remove the dressing 1 when checking the healing status of the target wound. This not only ensures the accuracy of the test results but also improves the convenience of assisted management of the target wound and reduces the risk of secondary infection caused by frequent removal of the dressing 1.

[0068] The specific usage process is as follows: When a user develops a wound on a certain part of their body due to surgery or injury, the wound management device can be set on the target wound according to preset usage rules, and then the wound management device can be activated. The activation method can be wirelessly controlled. After the wound management device is activated, it can acquire wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound.

[0069] The wound management device can acquire wound bed environment data, multidimensional morphological data, and biochemical index data of the target wound according to a preset acquisition strategy. The acquisition strategy includes real-time acquisition, followed by storing the acquired target data in a preset storage device according to time sequence. Alternatively, the acquisition strategy can also initiate acquisition only upon receiving an acquisition command. In actual operation, the settings can be customized according to specific needs.

[0070] In some embodiments, a microchannel structure is provided on the base layer 13. The microchannel structure connects the target wound and the exudate detection module respectively. By providing the microchannel structure, the exudate generated by the target wound can be conducted to the edge of the exudate detection module and / or the flexible biocompatible adhesive layer. This can help the exudate detection module detect the exudate, while preventing the exudate from accumulating on the surface of the target wound, avoiding secondary infection, and improving the healing efficiency of the target wound.

[0071] In some embodiments, an exudate collection component is also included. The exudate collection component is connected to the microchannel structure and is used to collect exudate. By setting the exudate collection component, the exudate generated by the target wound is collected, reducing the accumulation of exudate on the surface of the target wound, avoiding secondary infection, and improving the healing efficiency of the target wound.

[0072] In a specific embodiment, the exudate collection assembly includes a drive component and a collection component. The collection component is installed on the outside of the dressing 1, and the drive component is used to move the exudate towards the collection component, so that the collection component collects the exudate. In some embodiments, the drive component may be an air pump, and the collection component may be a tank, an absorbent sponge, an absorbent cloth, or other components capable of collecting exudate.

[0073] In some embodiments, the exudate collection component is a detachable structure, which allows for flexible configuration according to the wound condition. For cases where the target wound is large and there is a lot of exudate, by setting up an exudate collection device, the exudate can be effectively treated in a timely manner, ensuring the cleanliness of the wound management device, avoiding secondary infection, and further improving the wound healing efficiency.

[0074] Specifically, when the driving component is an air pump, the air pump is connected to the microfluidic structure through a pipe. A short pipe can be fixed on the microfluidic structure. The air pump is detachably connected to the short pipe through a long pipe, thereby realizing the detachable connection between the air pump and the microfluidic structure. The detachable connection method can be plug-in fixing.

[0075] When the collection device is a tank, it can be detachably connected to the air pump via a pipe to achieve a detachable connection between the collection device and the dressing 1; when the collection device is an absorbent sponge or absorbent cloth, it can be detachably connected to the dressing 1 via Velcro.

[0076] In some embodiments, the exudate composition detection module can also detect the concentration of the exudate. When the exudate concentration is too high, the drive unit can be controlled to actively treat the exudate. Specifically, the decision on whether to control the drive unit to actively treat the exudate can be made by management devices such as computers and controllers.

[0077] In some embodiments, the wound management device further includes a medication replenishment component, which is connected to a microfluidic structure for replenishing medication to the target wound. By incorporating the medication replenishment component, medication can be intelligently and promptly replenished to the target wound, reducing reliance on medical personnel, lowering medical resource costs, improving the convenience of dressing changes, and preventing secondary infections.

[0078] In some specific embodiments, the drug replenishment component includes a drug capsule and a drug release device. The drug capsule is connected to a microfluidic structure, and the drug release device is connected to the drug capsule and used to release the drug in the drug capsule to the target wound. That is, the drug release device allows the drug in the drug capsule to flow into the target wound through the microfluidic structure, facilitating dressing changes. The drug release device can be an air pump; by blowing air through the air pump, the drug is blown into the microfluidic structure and guided by the microfluidic structure to flow into the target wound.

[0079] In some embodiments, the drug release device and the drug capsule can be detachably connected to the dressing 1. The detachable connection method of the drug capsule can be Velcro fastening, clamp fastening, etc., and the detachable connection method of the drug release device can be tube insertion fastening or other methods.

[0080] In some embodiments, a triboelectric nanogenerator is disposed on the dressing 1 to power the biosensor assembly 7, the multidimensional scanning assembly 2, and the hyperspectral imaging assembly 6. Specifically, when the patient needs to move, the drug replenishment assembly and the exudate collection assembly can be detached. The patient's movement causes the triboelectric nanogenerator to generate electricity, powering the biosensor assembly 7, the multidimensional scanning assembly 2, and the hyperspectral imaging assembly 6, thus achieving energy conservation. The triboelectric nanogenerator can be embedded in the intermediate layer 12.

[0081] Specifically, the triboelectric nanogenerator can be a flexible triboelectric nanogenerator based on a double-layer fabric of "PDMS / cotton fabric", which matches the flexible load-bearing requirements of the middle layer 12. Its specific structure is as follows: the upper layer is a PDMS film with a thickness of 0.1 mm (the surface is treated with micro-nano structuring to form a hemispherical array with a diameter of 5 μm to enhance the triboelectric effect), the lower layer is a cotton fabric with a thickness of 0.2 mm (treated with silver nanowires to improve conductivity), and the middle layer 12 is a polytetrafluoroethylene isolation layer with a thickness of 0.05 mm. The overall size is 20 mm × 20 mm × 0.35 mm, and the weight is 0.1 g. It can be embedded in the reserved installation area of ​​the middle layer 12 without affecting the layout of other components.

[0082] Explanation of the power generation performance and energy supply matching test: In one specific embodiment of this application, under simulated human limb activity (5mm repositioning, 1Hz frequency), the generator maintains an open-circuit voltage of 3.2V±0.2V, a short-circuit current of 80μA±5μA, and a power density of 2.5μW / cm². Considering the power requirements of each component in this application (communication module 0.8mW, biosensor component 0.6mW, multi-dimensional scanning component 24mW, hyperspectral imaging component 1.2mW), the total power requirement is 26.6mW, while the actual output power of the generator is 10mW (20mm×20mm area). This, combined with an energy storage capacitor (i.e., a built-in capacitor in the dressing; during patient bed rest, an external power source powers the devices within the dressing and simultaneously powers the capacitor to store energy; during patient activity, the capacitor and generator provide power simultaneously), fully meets the design requirement of "synchronous power supply for multiple components" in this application. Due to the generator, the external usage time of the components in the dressing can be increased to meet the patient's activity needs.

[0083] For continuous power supply solutions during power generation gaps: To address the intermittent nature of triboelectric power generation, this embodiment integrates a 100μF medical-grade tantalum capacitor (3mm×2mm×1mm, 5V withstand voltage) as an energy storage element, echoing the "control module energy storage management" scheme in this application. Experimental testing shows that the capacitor can be fully charged after the generator continuously generates power for 30 seconds, and can stably power all components for ≥50 hours without any power input (test conditions: normal operating power consumption of each component). Simultaneously, a voltage monitoring circuit is implemented; when the capacitor voltage drops below 2.5V, the generator is triggered to prioritize charging the capacitor, ensuring continuous power supply and meeting the "uninterrupted real-time monitoring" requirement of this application.

[0084] In some embodiments, the dressing 1 further includes an anti-reflective layer disposed on the side of the protective layer 11 away from the base layer 13. The anti-reflective layer may include a subwavelength nanostructure, which may be a moth-eye biomimetic anti-reflective layer. Distributing an anti-reflective layer on the outer side of the protective layer 11 effectively improves the light transmittance of the protective layer 11, increases light utilization, provides sufficient light for the hyperspectral imaging component 6, and improves the accuracy of biochemical indicator data.

[0085] In some embodiments, the protective layer 11 is embedded with a carbon nanotube / PDMS conductive film to trigger a wireless alarm when the dressing 1 is torn by external force due to a sudden change in resistance. By embedding the carbon nanotube / PDMS conductive film in the protective layer 11, the working status of the dressing 1 can be monitored, improving the intelligent safety of the wound care device.

[0086] In some embodiments, a visible light-responsive coating is applied to the side of the protective layer 11 away from the base layer 13. The heterojunction generates reactive oxygen species (ROS) under ward lighting, killing drug-resistant bacteria (e.g., MRSA with a kill rate >99.9%). Preferably, the anti-reflective layer is located at... The heterojunction is located on the side away from the protective layer 11; in other embodiments, the antireflective layer and The position of the heterojunction can be interchanged.

[0087] In some embodiments, the protective layer 11 incorporates a temperature-sensitive hydrogel microvalve (such as polyacrylamide-co-acrylic acid). When exudate increases, the micropores expand (pore size 50→200μm), increasing breathability by 5 times and balancing the moist environment with the risk of infection. Specifically, after thoracic drainage, the dressing 1 automatically enhances breathability when humidity > 90%, reducing the incidence of maceration dermatitis by 30%. By incorporating the temperature-sensitive hydrogel microvalve into the protective layer 11, the breathability of the dressing layer 20 is further improved, preventing anaerobic inflammation in the wound.

[0088] In some embodiments, the protective layer 11 uses a temperature-sensitive polymer (such as PNIPAM), which is hydrophobic at body temperature (contact angle > 150°) to prevent blood and exudate from adhering; it switches to a hydrophilic state at low temperatures (< 25°C) for easy cleaning. Specifically, the protective layer 11 of the burn dressing 1 achieves self-cleaning at 37°C, reducing the replacement frequency by 50%. By employing a temperature-sensitive polymer, the protective layer 11 achieves the self-cleaning capability of the dressing 1, improving the safety of the wound management device and increasing the healing efficiency of the target wound.

[0089] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, all equivalent changes made to the structure, shape, and principle of this application should be included within the scope of protection of this application.

Claims

1. A wound auxiliary management device, characterized in that, The wound auxiliary management device comprises a dressing (1), a multi-dimensional scanning assembly (2), a terminal device (3) and a hyperspectral filtering assembly (5). The dressing (1) comprises a protective layer (11), an intermediate layer (12) and a substrate layer (13) arranged in sequence, wherein the substrate layer (13) is made of transparent or semi-transparent material. The multi-dimensional scanning assembly (2) comprises a laser emitting module and a photoelectric detection module arranged in the intermediate layer (12), wherein the laser emitting module is used for emitting a first light signal to a target wound, and the photoelectric detection module is used for collecting a second light signal reflected by the target wound. The terminal device (3) is connected with the photoelectric detection module and used for receiving and displaying multi-dimensional morphological data transmitted by the photoelectric detection module. The hyperspectral filtering assembly (5) is arranged in the protective layer (11) and used for filtering incident light entering the protective layer (11).

2. The wound auxiliary management device according to claim 1, further comprising a hyperspectral imaging assembly (6) arranged in the intermediate layer (12) and used for collecting biochemical index data of the target wound. The hyperspectral imaging assembly (6) corresponds in position to the hyperspectral filtering assembly (5) and is used for receiving filtered incident light of the hyperspectral filtering assembly (5).

3. A wound auxiliary management device according to claim 2, wherein, Further comprising a biological sensor assembly (7) arranged in the substrate layer (13) and used for collecting wound bed environment data.

4. The wound auxiliary management device of claim 1, wherein, The biological sensor assembly (7) comprises a temperature sensing module, a PH sensing module and an exudate component detection module.

5. A wound adjunct management device as defined in claim 4, wherein, The substrate layer (13) is provided with a micro-channel structure, and the micro-channel structure is connected with the target wound and the exudate detection module, respectively.

6. A wound adjunct management device as defined in claim 5, wherein, Further comprising an exudate collection assembly in communication with the micro-channel structure and used for collecting exudate.

7. A wound adjunct management device as defined in claim 6, wherein, The exudate collection assembly comprises a driving member and a collection member, wherein the collection member is installed on the dressing (1), and the driving member is used to drive the exudate to move towards the collection member, and the collection member is used to collect the exudate.

8. A wound auxiliary management device according to claim 7, wherein, Further comprising a drug supplement assembly installed on the dressing (1), wherein the drug supplement assembly comprises a drug capsule and a drug release member, the drug capsule is in communication with the micro-channel structure, and the drug release member is connected with the drug capsule and used to release the drug in the drug capsule to the target wound.

9. A wound auxiliary management device according to claim 7, wherein, The dressing (1) is provided with a friction nanometer generator, and the friction nanometer generator is used to supply power to the biological sensor assembly (7), the multi-dimensional scanning assembly (2) and the hyperspectral imaging assembly (6).

10. A wound auxiliary management device according to claim 9, wherein, ​

Images