An artificial intelligence-based wound pH detection method and a matching enzymatic in-situ forming hydrogel dressing and preparation method
By combining enzymatic in situ molding of hydrogel dressings with convolutional neural networks, the problems of insufficient adaptability of hydrogel dressings in emergency treatment scenarios and insufficient monitoring of healing status were solved. Real-time detection of wound pH and intelligent analysis of healing status were achieved, improving treatment efficiency and antibacterial performance.
Patent Information
- Application Number
- CN202411039944.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2026-08-25
- Estimated Expiration
- 2044-07-31
AI Technical Summary
Existing hydrogel dressings are difficult to adapt flexibly to various irregular wounds in emergency treatment scenarios, and lack real-time monitoring of wound healing status, resulting in inaccurate treatment plans and complicated nursing procedures.
A hydrogel dressing prepared using enzymatic in situ molding technology is combined with a convolutional neural network for pH detection. Gel polymerization is initiated by the reaction of glucose oxidase at the wound site to generate H2O2 and ·OH free radicals, and pH changes are indicated by phenol red. Combined with the antibacterial properties of quaternized chitosan, precise monitoring and analysis of wounds can be achieved.
It enables real-time detection of wound pH and intelligent analysis of healing status, improving treatment efficiency, simplifying nursing procedures, reducing costs, and possessing good antibacterial properties.
Smart Images
Figure CN119027379B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of biomedical materials technology, specifically relating to an artificial intelligence-based wound pH detection method and a matching enzymatic in-situ forming hydrogel dressing and its preparation method. Background Technology
[0002] Currently, the hydrogel materials used in wound detection and care are mostly limited to standardized specifications, making it difficult to fully adapt to wounds of varying shapes and sizes in real-world situations, especially in emergency treatment scenarios where the adaptability of pre-formed hydrogel dressings is limited. To address this issue, in-situ molding hydrogel technology has emerged, which can flexibly conform to various irregular wounds, significantly improving the accuracy and effectiveness of wound care.
[0003] Over the past decade, researchers have made significant progress in exploring the mechanisms of in-situ hydrogel formation, encompassing various strategies such as chemical crosslinking, enzymatic crosslinking, hydrogen bonding, and host-guest interactions. Among these, methods based on bio-enzymatic reactions have become a research hotspot in the field of medical materials due to their mildness and good biocompatibility.
[0004] It is worth noting that pH levels undergo dynamic changes during the wound healing process. This physiological indicator can reflect the wound healing status in real time, providing patients with intuitive health information. Leveraging this factor, a novel wound monitoring platform is being developed to monitor the wound healing status in real time during treatment. This not only helps to accurately adjust treatment plans and achieve personalized treatment, but also effectively simplifies nursing procedures and reduces the psychological and economic burden on patients and their families. This has profound significance for promoting the modernization of chronic wound management.
[0005] Furthermore, artificial intelligence, big data, and image processing technologies offer promising and feasible solutions for developing more reliable and simpler intelligent wound management in the future era of precision and personalized medicine.
[0006] Therefore, it is of great significance to explore and develop a novel, safe, and efficient enzyme-driven hydrogel technology to overcome the limitations of existing technologies. This technology can detect wound pH and provide intelligent analysis through machine learning, which is of great importance for improving wound care and promoting rapid patient recovery, and shows broad application prospects. Summary of the Invention
[0007] The purpose of this invention is to provide an artificial intelligence-based wound pH detection method and a corresponding enzymatic in situ forming hydrogel dressing and preparation method.
[0008] To achieve the above objectives, the present invention provides an artificial intelligence-based wound pH detection method, the steps of which are as follows:
[0009] (1) Using an image acquisition tool, the original images of the simulated blood-induced enzymatic in situ forming hydrogel dressing prepared in this invention under different pH conditions are acquired, and the acquired original images are cropped, labeled and data augmented.
[0010] Image acquisition tools include, but are not limited to, mobile phones and cameras; image cropping is to minimize invalid content in the image and prevent invalid images from interfering with recognition; data augmentation includes, but is not limited to, operations such as flipping and shifting to increase the number of photos.
[0011] (2) The images obtained from the preprocessing operation in step (1) are grouped into training set and validation set by random sampling. The training set and validation set are used for training and validation of the model, respectively.
[0012] (3) The generated training set is used to train the convolutional neural network to construct a wound healing period classification training model related to pH value; the training model is then validated through the validation set, and finally the optimal model trained in vitro is obtained.
[0013] The convolutional neural network includes convolutional layers, pooling layers, and fully connected layers. The number of convolutional layers is 2 to 100, the number of pooling layers is 2 to 100, the pooling method of the pooling layers is max pooling or average pooling, and the number of fully connected layers is at least 1.
[0014] (4) Using an image acquisition tool, acquire images of a wound covered with the enzyme-catalyzed in situ forming hydrogel dressing for actual wounds as described in this invention, and acquire the pH value of the wound using a pH meter; perform image cropping, labeling, and data augmentation preprocessing on the acquired original images; manually label the images according to the pH meter reading, and round the pH value to the nearest integer; group the images obtained after preprocessing and pH labeling by random sampling into training set and validation set, use the generated training set to train the optimal in vitro training model obtained in step (3), and then validate the optimal model through the validation set to obtain the optimal model again;
[0015] (5) Using an image acquisition tool, acquire an image of a wound covered with an enzyme-catalyzed in-situ formed hydrogel dressing for actual wounds as described in this invention. Based on the wound image and the optimal model obtained in step (4), automatically identify and determine the pH value at the wound site, and then determine the wound healing status.
[0016] The present invention also provides an enzymatically induced in-situ formed hydrogel dressing for use in the above-mentioned artificial intelligence-based wound pH detection method and its preparation method.
[0017] The preparation method of the enzymatically catalytically formed hydrogel dressing of this invention first utilizes nanogel technology to prepare nanogel glucose oxidase to improve the stability of the enzyme during long-term storage. Then, methacrylated polyvinyl alcohol (PVAMa) is synthesized as a macromolecular crosslinking agent, and phenol red is added, followed by copolymerization with quaternary ammonium chitosan (QCS), hydroxyethyl acrylate (HEA), and simulated blood to prepare the enzymatically catalytically formed hydrogel dressing. Phenol red has the ability to indicate wound pH and can be used to detect wound healing. The enzymatically catalytically formed hydrogel technology utilizes the reaction of glucose oxidase with glucose in the wound surface, generating hydrogen peroxide (H2O2) and hydroxyl radicals (·OH) to initiate gel polymerization, thereby forming the dressing in situ and achieving hemostasis. Quaternary ammonium chitosan has good antibacterial properties and, as a natural material, possesses good biocompatibility. Simultaneously, the macromolecular crosslinking agent, due to its large molecular weight, more easily forms a hydrogel network structure, enabling rapid in-situ formation.
[0018] All the raw materials involved in this invention are commercially available.
[0019] The preparation method of the enzyme-catalyzed in-situ formed hydrogel dressing according to the present invention comprises the following steps:
[0020] (1) Preparation of nanogel glucose oxidase
[0021] Weigh 1-ethyl-(3-dimethylaminopropyl)carbodiimide, N-hydroxysuccinimide, and glucose oxidase and add them to deionized water, then add methacrylic acid and react at room temperature for 20–30 hours. Next, weigh acrylamide, N,N-methylenebisacrylamide, and ammonium persulfate and add them to the mixture. After thorough mixing, inject N2 for deoxygenation for 1–2 hours, then add tetramethylethylenediamine and react at room temperature for 20–30 hours. Finally, use M... W Dialysis was performed using dialysis bags with a concentration of 3000-3500, and the retained solution was lyophilized to obtain nano-gel glucose oxidase powder.
[0022] The components include: 1-ethyl-(3-dimethylaminopropyl)carbodiimide with a mass concentration of 10%–30%; N-hydroxysuccinimide to 1-ethyl-(3-dimethylaminopropyl)carbodiimide with a molar ratio of 1:1–2; glucose oxidase with a mass concentration of 0.1%–1%; methacrylic acid with a volume concentration of 1%–10%; acrylamide with a mass concentration of 1%–3%; N,N-methylenebisacrylamide with a mass concentration of 0.1%–1%; ammonium persulfate with a mass concentration of 0.1%–1%; and tetramethylethylenediamine with a mass concentration of 0.1%–1%.
[0023] (2) Preparation of macromolecular crosslinking agents
[0024] Weigh out polyvinyl alcohol and dissolve it in dimethyl sulfoxide, add glycidyl methacrylate and tetramethylethylenediamine, and stir in an oil bath at 55°C for 18 hours; finally, use M W Dialysis was performed using a dialysis bag with a concentration of 3000-3500, and the retained solution was lyophilized to obtain methacrylated polyvinyl alcohol powder.
[0025] The polyvinyl alcohol has a mass concentration of 5% to 10%, glycidyl methacrylate has a mass concentration of 1% to 3%, and tetramethylethylenediamine has a mass concentration of 0.1% to 1%.
[0026] (3) Preparation of quaternized chitosan
[0027] Chitosan was weighed and added to deionized water, followed by glacial acetic acid. The mixture was stirred at room temperature for 1–3 hours. Then, 2,3-epoxypropyltrimethylammonium chloride was added, and the mixture was reacted at 50–60°C for 15–20 hours. After filtration, the crude product was dissolved in deionized water, precipitated with acetone, and then filtered. The precipitate was then dissolved in deionized water, and finally dialyzed using a dialysis bag with a MW of 3000–3500. The retained solution was freeze-dried to obtain quaternized chitosan powder.
[0028] The chitosan has a mass concentration of 1% to 5%, the glacial acetic acid has a volume concentration of 1% to 2%, and the molar ratio of chitosan to 2,3-epoxypropyltrimethylammonium chloride is 1:2.
[0029] (4) Preparation of simulated blood
[0030] Glucose, heme chloride, and ascorbic acid were weighed and added to deionized water. The mixture was then magnetically stirred at room temperature for 1–2 hours to obtain simulated blood.
[0031] The concentrations of heme chloride, glucose, and ascorbic acid range from 0.01% to 0.1%.
[0032] (5) Preparation of enzymatically induced in-situ formed hydrogel dressings:
[0033] Weigh 10-20 mg of nanogel glucose oxidase, 100-150 mg of methacrylated polyvinyl alcohol, 1-2 mg of quaternized chitosan, and phenol red, add them to 1-2 mL of deionized water, dissolve, and then add 200-300 μL of hydroxyethyl acrylate. Mix well to obtain a gel prepolymer solution. Pour the gel prepolymer solution into a container and let it come into contact with blood. It will quickly solidify within 40 seconds to obtain an enzymatically induced in-situ hydrogel dressing that can detect the pH of the actual wound. Alternatively, mix simulated blood with the gel prepolymer solution at a volume ratio of 1:10-20, pour it into a container, and it will quickly solidify within 40 seconds to obtain an enzymatically induced in-situ hydrogel dressing that can detect the pH of the simulated blood.
[0034] The multifunctional enzymatic in situ molding hydrogel dressing described in this invention is prepared by the above method.
[0035] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:
[0036] The multifunctional hydrogel dressing for enzymatic in-situ molding and simultaneous wound pH detection provided in this invention utilizes glucose oxidase to consume glucose, generating free radicals that initiate in-situ polymerization of the hydrogel. On one hand, enzymatic in-situ polymerization is more economical in terms of raw material consumption, and the glucose oxidase reaction at the human wound site is safer than other in-situ polymerization methods such as UV-induced polymerization. On the other hand, quaternized chitosan has good antibacterial properties, and pH detection can reflect the wound healing status. Finally, by combining a convolutional neural network algorithm, the optimal recognition model generated by the phenol red reagent introduced into the hydrogel network is identified and analyzed to assess the wound healing status, providing a new strategy for monitoring and immediate diagnosis. This invention can improve treatment efficiency and save treatment costs. It is safe, efficient, and multifunctional. Attached Figure Description
[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.
[0038] Figure 1 Before gelation of the antibacterial hydrogel dressing prepared in Example 1 of this invention ( Figure 1 (left) and after gelation ( Figure 1 The photo on the right.
[0039] Figure 2 This is a graph showing the changes in the UV-Vis spectrum of the antibacterial hydrogel dressing prepared in Example 2 of the present invention at different pH values when monitoring the pH value of the wound.
[0040] Figure 3 This image shows a comparison of the inhibitory effects of the hydrogel dressing prepared in Example 1 of this invention on Escherichia coli (E. coli). The left image shows the effect before inhibition, and the right image shows the effect after inhibition.
[0041] Figure 4 This image shows a comparison of the inhibitory effects of the hydrogel dressing prepared in Example 1 of this invention on Staphylococcus aureus. The left image shows the effect before inhibition, and the right image shows the effect after inhibition.
[0042] Figure 5 Before gelation of the antibacterial hydrogel dressing prepared in Example 2 of this invention ( Figure 5 (left) and after gelation ( Figure 5 The photo on the right.
[0043] Figure 6 This is a comparison of the inhibitory effect of the hydrogel dressing prepared in Example 2 of the present invention on Escherichia coli (E. coli). The left image shows the effect before inhibition, and the right image shows the effect after inhibition.
[0044] Figure 7 This image shows a comparison of the inhibitory effects of the hydrogel dressing prepared in Example 2 of this invention on Staphylococcus aureus. The left image shows the effect before inhibition, and the right image shows the effect after inhibition.
[0045] Figure 8 The experiment on the healing of chronic wounds in mice was conducted using the hydrogel dressings prepared in Examples 1 and 2 of this invention and a blank control group.
[0046] Figure 9 This is a confusion matrix diagram showing the accuracy of the training model tested in Embodiment 2 of the present invention. Detailed Implementation
[0047] Those skilled in the art will understand that the numerical ranges in the embodiments of this application should be understood to also include every intermediate value between the upper and lower limits of the range. Every smaller range between any stated value or intermediate value within a stated range, and every other stated value or intermediate value within said range, is also included within this invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.
[0048] Unless otherwise stated, the technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. While only preferred methods and materials are described herein, any methods and materials similar or equivalent to those described herein may be used in the implementation or testing of this application. All references to this specification are incorporated by way of citation to disclose and describe the methods and / or materials associated with those references. In the event of any conflict with any introduced reference, the content of this specification shall prevail.
[0049] Unless otherwise specified, the methods used in the following examples are all conventional methods. The reagents and materials used in the following examples are all conventional biochemical reagents.
[0050] Example 1
[0051] (1) Weigh 678 mg of 1-ethyl-(3-dimethylaminopropyl)carbodiimide, 400 mg of N-hydroxysuccinimide, and 30 mg of glucose oxidase and add them to 5 mL of water. Then add 200 μL of methacrylic acid and react at room temperature for 24 h. Next, weigh 100 mg of acrylamide, 10 mg of N,N-methylenebisacrylamide, and 10 mg of ammonium persulfate and add them to the mixture. After mixing well, inject N2 for deoxygenation for 1 hour, then add 5 mg of tetramethylethylenediamine and react at room temperature for 24 hours. Finally, use M W Dialysis was performed using a dialysis bag with a capacity of 3000. After lyophilizing the retained solution, 35 mg of nanogel glucose oxidase powder was obtained.
[0052] (2) Weigh 640 mg of polyvinyl alcohol and dissolve it in 10 mL of dimethyl sulfoxide. Then, add 100 mg of glycidyl methacrylate and 15 μL of tetramethylethylenediamine to the reaction solution and stir in an oil bath at 60 °C for 24 h. Finally, use M W Dialysis was performed using a dialysis bag with a capacity of 3000, and the retained solution was lyophilized to obtain 650 mg of polyvinyl alcohol macromolecular crosslinking agent.
[0053] (3) Weigh 3g of chitosan and add it to 100mL of distilled water, then add 1mL of glacial acetic acid and stir at room temperature for 1h; then add 5.6g of 2,3-epoxypropyltrimethylammonium chloride and react at 55℃ for 18 hours. After filtration, dissolve the crude product in deionized water, precipitate with acetone and filter; then dissolve the precipitate in deionized water, and finally dialyze it with a dialysis bag with MW=3000. After lyophilizing the retained solution, 3.1g of quaternized chitosan powder is obtained.
[0054] (4) Weigh 10 mg of glucose, 0.1 mg of heme chloride and 1 mg of ascorbic acid, add them to 100 μL of deionized water, shake to dissolve, and obtain simulated blood;
[0055] (5) Weigh 10 mg of nanogel glucose oxidase, 100 mg of methacrylated polyvinyl alcohol, 1 mg of quaternized chitosan, and 0.1 mg of phenol red, add them to 1 mL of deionized water, dissolve them, add 200 μL of hydroxyethyl acrylate, mix them evenly to obtain a gel prepolymer solution; pour the gel prepolymer solution into a container and let it come into contact with blood. It can be rapidly formed within 40 seconds to obtain an enzymatic in situ formed hydrogel dressing that can detect the pH of the actual wound; in this embodiment, simulated blood is used instead of actual blood. Mix 100 μL of simulated blood with the gel prepolymer solution, pour it into a container, and it can be rapidly formed within 40 seconds to obtain an enzymatic in situ formed hydrogel dressing containing simulated blood that can detect the pH of the wound, denoted as H1.
[0056] Example 2
[0057] The reaction steps and conditions in Example 2 were the same as in Example 1, resulting in an enzymatically catalytically formed hydrogel dressing containing simulated blood that enhances bactericidal efficacy and can detect wound pH. This dressing was denoted as H2. The difference was that in step (5) of preparing the H2 gel prepolymer solution, the amount of quaternized chitosan used was 50 mg.
[0058] Test Example (1): Performance testing of dressings prepared in Examples 1 and 2.
[0059] a. Measuring gel setting time using the inverted vial method
[0060] At regular time intervals (e.g., every 30 seconds or 1 minute), gently tilt the vial to a 45-degree angle and hold for several seconds, observing whether the solution flows. If the solution flows freely, the gel has not yet formed. Repeat this process until the solution stops flowing, indicating that a gel has formed.
[0061] The results are as follows Figure 1 and Figure 5 As shown, the gel solution initially flows as the vial tilts; after 45 seconds, the gel solution in both vials stops flowing and forms a gel.
[0062] b. Determination of the antibacterial properties of antibacterial gel
[0063] Cut H1 and H2 into pieces, take 10 mg of each, and place them separately into 1 mL of a solution containing 10 mg of H2. 7 Incubate CFU / mL Escherichia coli and Staphylococcus aureus in physiological saline for 12 h, then dilute with PBS to 10. 4 The gel was then multiplied by 50 μL and spread onto a nutrient agar plate to quantitatively evaluate its antibacterial effect.
[0064] The results are as follows Figure 3 , 4 As shown in Figures 6 and 7, enzyme-initiated hydrogel dressings, due to their reaction with glucose in the blood during polymerization to produce H2O2 and reactive oxygen species (·OH) for initiation, can both initiate free radical polymerization and act as an effective antibacterial component, effectively eliminating bacteria. Furthermore, increasing the amount of quaternized chitosan can significantly eliminate even more bacteria.
[0065] (II) pH curves were prepared for the dressings prepared in Example 2, and a model was established:
[0066] H2 was immersed in 10 μL of PBS solution at different pH values (pH = 5–9). The samples were then scanned at 565 nm using a microplate reader. Figure 2 As shown, the hydrogel color varies significantly at different pH levels, corresponding to the color difference caused by the n-π* transition of phenol red from benzene to quinone excitons when the ambient solution becomes more alkaline.
[0067] Original images of the enzyme-catalyzed in-situ molding hydrogel dressing H2 containing simulated blood, prepared according to this invention, were collected using smart digital devices (such as mobile phones) under different pH conditions (100 images were collected for each pH group). The collected images were cropped so that the image of the enzyme-catalyzed in-situ molding hydrogel dressing prepared according to this invention occupied the main part of the image. The images were manually labeled according to pH, and then data augmentation operations such as inversion and rotation were performed on the images to double the number of photos. Then, 80% of the images were randomly selected to generate a training set, and 20% were used to generate a validation set. The convolutional neural network was trained on the training set to construct a wound healing period classification training model related to pH value; then the training model was validated using the validation set, and finally the optimal model trained in vitro was obtained.
[0068] In this embodiment, the convolutional neural network is based on the PyTorch framework, the algorithm adopts ResNet-18, and the hyperparameters are set as follows: InputSize, (224, 224); BatchSize, (32); Epochs, (100); BaseLearningRate, (0.0001); there are 2 convolutional layers and 2 pooling layers, the pooling method of the pooling layers is max pooling, and the fully connected layer is 1 layer.
[0069] (III) Wound healing experiments were conducted on the gel prepolymer solutions prepared in Examples 1 and 2:
[0070] Mice were anesthetized with ether, and then their backs were shaved with a depilatory agent. Two circular incisions with a diameter of 8 mm were made in the skin, and 20 μL of 1×10⁻⁶ solution was injected into the wounds. 8 A chronic wound infection model was established by infecting patients with a CFU / mL suspension of S. aureus bacteria for 2 days.
[0071] The enzymatic in situ forming hydrogel dressings obtained from the actual wounds in steps (5) of Examples 1 and 2 were applied to the wounds.
[0072] like Figure 8 As shown, compared with the blank control group that only used gauze to wrap the wound, both H1 and H2 can promote wound healing. Among them, H2 has a significantly shorter healing period than the other two groups due to its stronger antibacterial properties.
[0073] (iv) An optimal model was constructed by conducting an artificial intelligence-based wound healing period classification experiment on the dressings prepared in Example 2 for actual wounds:
[0074] Images of mouse wounds covered with the enzyme-catalyzed in-situ hydrogel dressing described in this invention, obtained in step (III), were collected every other day using a smart digital device (such as a mobile phone) for a total of 12 days. The pH value at the wound site was also measured using a pH meter. The collected images were cropped and enlarged to ensure the image of the enzyme-catalyzed in-situ hydrogel dressing prepared according to this invention dominated the frame. Data augmentation operations such as inversion and rotation were then performed on the images to double the number of photos. The images were manually labeled according to the pH meter readings, with the pH value rounded to the nearest integer. Then, 80% of the images were randomly selected to generate a training set, and 20% to generate a validation set. The optimal model obtained in step (II) was further optimized using the training set, and the optimized model was validated using the validation set to obtain the optimal model again.
[0075] In this embodiment, the convolutional neural network is based on the PyTorch framework, the algorithm uses ResNet-18, and the hyperparameters are set as follows: InputSize, (224, 224); BatchSize, (32); Epochs, (100); BaseLearningRate, (0.0001); there are 2 convolutional layers and 2 pooling layers, the pooling method of the pooling layers is max pooling, and the fully connected layer is 1 layer.
[0076] (v) Verification of wound healing in mice using Example 2:
[0077] Images of mouse wounds covered with the enzyme-catalyzed in-situ hydrogel dressing described in this invention, which were not captured in step (iv) using a smart digital device (such as a mobile phone), were collected every other day for a total of 12 days. The pH value at the wound site was also measured using a pH meter. The collected images were cropped, and the portion covered with the hydrogel dressing was enlarged to ensure that the image of the enzyme-catalyzed in-situ hydrogel dressing prepared in this invention occupied the majority of the image. The images were manually labeled according to the pH meter readings, and the pH values were rounded to the nearest integer. A portion of the images was randomly selected as a test set, which was used to evaluate the final model after training and was input into the optimal model obtained in step (iv) for testing.
[0078] The optimal model accuracy confusion matrix obtained in step (5) is shown in the figure below. Figure 9 As shown, the model has an accuracy of over 95%, can automatically identify and determine the pH value at the wound site, and thus determine the wound healing status, meeting the classification requirements for the wound healing period.
[0079] The above examples are merely illustrative descriptions of the materials used in this invention and are not intended to limit the scope of this application. It should not be construed that the specific implementation of this invention is limited to these descriptions. For those skilled in the art, any substitutions of materials, adjustments to the order of addition, or changes in the proportions of additions to the technical solutions and inventive concepts of this invention should be considered within the scope of protection of this invention.
Claims
1. A method for preparing an enzyme-catalyzed in-situ formed hydrogel dressing, comprising the following steps: (1) Preparation of nanogel glucose oxidase Weigh out 1-ethyl-(3-dimethylaminopropyl)carbodiimide, N-hydroxysuccinimide, and glucose oxidase and add them to deionized water, then add methacrylic acid and react at room temperature for 20-30 hours; then weigh out acrylamide, N,N-methylenebisacrylamide, and ammonium persulfate and add them to the mixture. After thorough mixing, inject N2 for deoxygenation for 1-2 hours, then add tetramethylethylenediamine and react at room temperature for 20-30 hours; finally, use M... W Dialysis was performed using dialysis bags with a concentration of 3000~3500, and the retained solution was freeze-dried to obtain nano-gel glucose oxidase powder. (2) Preparation of macromolecular crosslinking agents Weigh out polyvinyl alcohol and dissolve it in dimethyl sulfoxide, add glycidyl methacrylate and tetramethylethylenediamine, and stir in an oil bath at 55°C for 18 hours; finally, use M W Dialysis was performed using a dialysis bag with a concentration of 3000~3500, and the retained solution was lyophilized to obtain methacrylated polyvinyl alcohol powder. (3) Preparation of quaternized chitosan Chitosan was weighed and added to deionized water, followed by glacial acetic acid. The mixture was stirred at room temperature for 1-3 hours. Then, 2,3-epoxypropyltrimethylammonium chloride was added, and the mixture was reacted at 50-60°C for 15-20 hours. After filtration, the crude product was dissolved in deionized water, precipitated with acetone, and then filtered. The precipitate was then dissolved in deionized water, and finally dialyzed using a dialysis bag with a MW of 3000-3500. The retained solution was lyophilized to obtain quaternized chitosan powder. (4) Preparation of simulated blood Weigh out glucose, heme chloride, and ascorbic acid and add them to deionized water. Stir magnetically at room temperature for 1-2 hours to obtain simulated blood. (5) Preparation of enzymatically induced in-situ hydrogel dressings: Weigh 10-20 mg of nanogel glucose oxidase, 100-150 mg of methacrylated polyvinyl alcohol, 1-2 mg of quaternized chitosan, and phenol red, add them to 1-2 mL of deionized water, dissolve them, add 200-300 μL of hydroxyethyl acrylate, mix well, and obtain the gel prepolymer solution. Pour the gel prepolymer into a container and let it come into contact with blood. It will quickly solidify within 40 seconds to obtain an enzymatically induced in-situ hydrogel dressing that can detect the pH of the actual wound. Alternatively, mix simulated blood with the gel prepolymer at a volume ratio of 1:10~20, pour it into a container, and it will quickly solidify within 40 seconds to obtain an enzymatically induced in-situ hydrogel dressing that can detect the pH of the simulated blood.
2. The method for preparing an enzymatically catalytically formed hydrogel dressing as described in claim 1, characterized in that: In step (1), the mass concentration of 1-ethyl-(3-dimethylaminopropyl)carbodiimide is 10%~30%, the molar ratio of N-hydroxysuccinimide to 1-ethyl-(3-dimethylaminopropyl)carbodiimide is 1:1~2, the mass concentration of glucose oxidase is 0.1%~1%, the volume concentration of methacrylic acid is 1%~10%, the mass concentration of acrylamide is 1%~3%, the mass concentration of N,N-methylenebisacrylamide is 0.1%~1%, the mass concentration of ammonium persulfate is 0.1%~1%, and the mass concentration of tetramethylethylenediamine is 0.1%~1%.
3. The method for preparing an enzymatically catalyzed in-situ hydrogel dressing as described in claim 1, characterized in that: In step (2), the mass concentration of polyvinyl alcohol is 5%~10%, the mass concentration of glycidyl methacrylate is 1%~3%, and the mass concentration of tetramethylethylenediamine is 0.1%~1%.
4. The method for preparing an enzymatically catalytically formed hydrogel dressing as described in claim 1, characterized in that: In step (3), the mass concentration of chitosan is 1%~5%, the volume concentration of glacial acetic acid is 1%~2%, and the molar ratio of chitosan to 2,3-epoxypropyltrimethylammonium chloride is 1:
2.
5. The method for preparing an enzymatically catalytically formed hydrogel dressing as described in claim 1, characterized in that: In step (4), the mass concentration of heme chloride is 0.01~0.1%, the mass concentration of glucose is 1~20%, and the concentration range of ascorbic acid is 0.01~0.1%.
6. An artificial intelligence-based method for detecting wound pH, comprising the following steps: (1) Using an image acquisition tool, original images of the enzymatically induced in-situ hydrogel dressing induced by simulated blood were acquired under different pH conditions. The acquired original images were then cropped, labeled, and preprocessed with data augmentation. The enzyme-catalyzed in-situ formed hydrogel dressing is prepared by the method according to any one of claims 1 to 5; (2) The images obtained from the preprocessing operation in step (1) are grouped into a training set and a validation set by random sampling. The training set and the validation set are used for training and validation of the model, respectively. (3) The generated training set is used to train the convolutional neural network to construct a wound healing period classification training model related to pH value; the training model is then validated through the validation set, and finally the optimal model trained in vitro is obtained. (4) Use an image acquisition tool to acquire images of wounds covered with enzymatic in situ formed hydrogel dressings caused by actual wounds, and use a pH meter to acquire the pH value at the wound site; perform image cropping, labeling and data augmentation preprocessing on the acquired original images; manually label the images according to the pH meter reading, and round the pH to the nearest integer; group the images obtained after preprocessing and pH labeling by random sampling into training set and validation set, use the generated training set to train the optimal model obtained in step (3), and then use the validation set to validate the optimal model to obtain the optimal model again; (5) Using an image acquisition tool, acquire images of wounds covered with enzymatic in situ formed hydrogel dressings caused by actual wounds. Based on the wound image and the optimal model obtained in step (4), automatically identify and determine the pH value at the wound site, and then determine the wound healing status.
7. The wound pH detection method based on artificial intelligence as described in claim 6, characterized in that: Image acquisition tools include mobile phones or cameras; image cropping reduces invalid content in images to prevent invalid images from interfering with recognition; data augmentation increases the number of photos by flipping, shifting, and deforming.
8. The wound pH detection method based on artificial intelligence as described in claim 6, characterized in that: A convolutional neural network includes convolutional layers, pooling layers, and fully connected layers. The number of convolutional layers is 2 to 100, the number of pooling layers is 2 to 100, the pooling method of the pooling layers is max pooling or average pooling, and the number of fully connected layers is at least 1.