Anti-pollution medical consumable multi-code fusion tracing method and system

Abstract

The invention relates to an anti-pollution medical consumable multi-code fusion tracing method and system, and relates to the technical field of medical information process.The method comprises the steps that multi-modal analysis data of consumables is collected, the multi-modal analysis data comprises original image data and current environment data, and the current environment data comprises the original image data and the current environment data; the original image data comprises at least one visual tagging label; extracting a label image at the tagging label in the original image data, decoding according to the pollution condition of the label image to obtain an identification code, judging the similarity between the current environment data and the environment check code, determining the life cycle data of the corresponding consumable based on the mapping relationship between the identification codes when the similarity is higher than a preset value, and sending the life cycle data to the corresponding consumable. And traceability of the medical consumables when the labels are polluted is realized.

Description

Technical Field

[0001] This application relates to the field of medical information processing technology, and in particular to a method and system for fusion and traceability of multi-code medical consumables based on a pollution-resistant approach. Background Technology

[0002] Currently, hospitals and medical consumables supply chains generally use UDI codes (Unique Device Identifiers) or batch codes for traceability management of medical consumables. However, in actual application scenarios, due to the special nature of the medical environment and the limitations of existing technology, label contamination and damage can lead to traceability interruptions, making it difficult to meet the needs of clinical safety and regulatory compliance.

[0003] In medical settings, high-humidity environments such as operating rooms and storage rooms, as well as splashes of liquids like blood, disinfectants, and chemical reagents, can easily contaminate, blur, or even damage barcode or QR code labels. Traditional scanning equipment relies solely on visible light for recognition, achieving a success rate of less than 60% for identifying contaminated labels. Once a label becomes unreadable, the traceability information for consumables is directly interrupted, making it impossible to link production and usage records, severely impacting adverse event tracing and quality traceability. Summary of the Invention

[0004] To enable traceability of medical consumables in the event of label contamination, this application provides a method and system for fusion traceability of multi-code medical consumables that is resistant to contamination.

[0005] Firstly, this application provides a method for multi-code fusion traceability of contamination-resistant medical consumables, employing the following technical solution: A method for multi-code fusion traceability of contamination-resistant medical consumables, the method comprising: Collect multimodal analysis data of consumables, the multimodal analysis data including raw image data and current environmental data, wherein the raw image data includes at least one visually visible coded label; The label image at the coded label location is extracted from the original image data, and the label image is decoded according to the pollution status to obtain the identification code. The identification code is at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identification codes. The basic code is a globally unique item identifier for consumables, and the dynamic batch code is used to identify the attributes of consumables that change over time. Determine the similarity between the current environmental data and the environmental check code. The environmental check code is generated based on the environmental data when the code assignment label is generated, and the environmental check code has a unique mapping relationship with the dynamic batch code. When the similarity is higher than a preset value, the lifecycle data of the corresponding consumable is confirmed based on the mapping relationship between the identification codes.

[0006] In one possible implementation, decoding is performed based on the contamination status of the label image to obtain an identification code, including: Pixel-level segmentation is performed based on the labeled image, and the mask region is marked according to the segmented pixel-level mask. The mask region includes background, code region, contaminated region, and occlusion region. The contaminated area ratio is calculated based on the number of pixels in the contaminated area and the total number of pixels in the code area; The pollution levels are classified according to the percentage of contaminated area. Based on the classification, the corresponding decoding strategy is selected first to obtain the identification code.

[0007] In one possible implementation, the step of preferentially selecting the corresponding decoding strategy based on the level classification to obtain the identifier code includes: The pollution level classification is set according to a preset range of pollution area ratios, including no pollution level, light pollution level, moderate pollution level, and heavy pollution level. When there is no pollution level, the label image is scanned using a standard decoder to obtain the identification code; At a light pollution level, the label image is enhanced, a near-infrared image of the enhanced label image is acquired and decoded to obtain the identification code; At the moderate pollution level, the pollution area of ​​the label image is located by semantic segmentation, the missing code elements in the pollution area are filled, and the missing code elements are completed based on the preset encoding rules of the identifier code to obtain the identifier code; The severe pollution level is determined by using depth prediction to complete the label image and obtain the identification code.

[0008] In one possible implementation, the identification code is obtained by completing the label image through depth prediction, including: Based on the fixed grid characteristics of QR codes, a standard grid is constructed by detecting positioning points, and the state of each symbol is marked to obtain the sequence of damaged symbols; Input the current environment data and the damaged code sequence into the trained deep prediction model to complete the label image and obtain the identification code.

[0009] In one possible implementation, obtaining the identifier code includes: The type of the identifier is obtained based on the explicit fixed features and multi-dimensional implicit features of the identifier. The explicit fixed features include encoding length, prefix format, character set, and check bit rules. The multi-dimensional implicit features include segmentation structure, character distribution, check bit association, and semantic association.

[0010] In one possible implementation, the base code is generated using a hash algorithm of product model, specifications, internal material ID, and factory ID, and the base code exists in the fields of the first coding tag, RFID chip, consumable internal packaging imprint, and UDI carrier.

[0011] In one possible implementation, the dynamic batch code is generated using an encryption algorithm that combines the production batch number, sterilization batch number, timestamp, and batch counter. The dynamic batch code is present on the second coding label, variable QR code layer, and production identification field of the UDI on the outer packaging of the consumables.

[0012] Secondly, this application provides a multi-code fusion traceability system for contamination-resistant medical consumables, employing the following technical solution: A pollution-resistant multi-code fusion traceability system for medical consumables, the system comprising: The data acquisition module is used to acquire multimodal analysis data of consumables. The multimodal analysis data includes raw image data and current environmental data. The raw image data includes at least one visually visible coded label. The decoding processing module is used to extract the label image at the coded label location in the original image data, and decode it according to the pollution status of the label image to obtain the identification code. The identification code is at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identification codes. The basic code is a globally unique item identifier for consumables, and the dynamic batch code is used to identify the attributes of consumables that change over time. The tag verification module is used to determine the similarity between the current environment data and the environment verification code. The environment verification code is generated based on the environment data when the code tag is generated, and the environment verification code has a unique mapping relationship with the dynamic batch code. The data output module is used to confirm the lifecycle data of the corresponding consumables based on the mapping relationship between the identification codes when the similarity is higher than a preset value.

[0013] Thirdly, this application provides a pollution-resistant medical consumables multi-code fusion traceability device, which adopts the following technical solution: A pollution-resistant medical consumable multi-code fusion traceability device includes: a memory and a processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0014] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible embodiments of the first aspect as described above.

[0015] The embodiments of this application provide a multi-code fusion traceability method and system for anti-contamination medical consumables. By first calculating the contamination area ratio and classifying the levels based on pixel-level segmentation of potentially contaminated label images, and then using an intelligent matching decoding strategy based on the levels to comprehensively cover different levels of contamination, a basic code or a dynamic batch code is obtained. By judging the similarity between the current environmental data and the environmental check code, and by leveraging the pre-built mapping relationship between the basic code, dynamic batch code, and environmental check code, the information can be accurately verified and checked. Ultimately, this achieves accurate traceability of consumable lifecycle data, effectively improving the reliability and security of medical consumable management. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a multi-code fusion traceability method for anti-pollution medical consumables, provided as an embodiment of this application.

[0017] Figure 2 This is a flowchart illustrating a multi-code fusion traceability method for anti-pollution medical consumables, provided as an embodiment of this application.

[0018] Figure 3 This is a schematic diagram of the structure of a multi-code fusion traceability system for anti-pollution medical consumables provided in an embodiment of this application.

[0019] Figure 4 This is a schematic diagram of the structure of a pollution-resistant medical consumable multi-code fusion traceability device provided in one embodiment of this application. Detailed Implementation

[0020] To better understand the purpose, technical solutions, and advantages of this application, it has been described and illustrated below with reference to the accompanying drawings and embodiments. However, those skilled in the art should understand that this application can be implemented without these details. In some cases, to avoid obscuring various aspects of this application due to unnecessary description, well-known methods, processes, systems, components, and / or circuits already described at a higher level will not be elaborated upon. It will be apparent to those skilled in the art that various modifications can be made to the embodiments disclosed in this application, and the general principles defined in this application can be applied to other embodiments and application scenarios without departing from the principles and scope of this application. Therefore, this application is not limited to the illustrated embodiments, but conforms to the broadest scope consistent with the scope of protection claimed in this application.

[0021] It should be noted that the descriptions of these embodiments are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0022] It should be understood that the embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or any combination thereof. For hardware implementations, the processor may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic units designed to perform the functions described herein, or combinations thereof.

[0023] When an embodiment is implemented as software, firmware, middleware, or microcode, program code, or code segments, it may be stored in a machine-readable medium, such as a storage component. A code segment may represent a procedure, function, subroutine, program, routine, subroutine, module, software package, class, or any combination of instructions, data structures, or program statements. One code segment can be coupled to another code segment or hardware circuitry by passing and / or receiving information, data, arguments, parameters, or memory contents. Information, arguments, parameters, data, etc., can be passed, forwarded, or transmitted using any suitable means, including memory sharing, messaging, token passing, network transmission, etc.

[0024] For software implementations, the techniques described herein can be implemented using modules (e.g., programs, functions, etc.) that perform the functions described herein. The software code can be stored in memory units and executed by a processor. The memory units can be implemented within or outside the processor; in the latter case, the memory units can be communicatively coupled to the processor via various methods known in this art.

[0025] In the description of this application, "several" means one or more, "more than" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0026] In the description of this application, the terms "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples.

[0027] Currently, the production, packaging, and distribution of medical consumables are handled by different entities, with each stage generating its own code. Typically, the production end uses the manufacturer's internal item code, the packaging end uses the UDI code, and the distribution end uses the batch code. The three types of codes have inconsistent formats and lack a clear correlation.

[0028] For example, a certain surgical gauze is identified as "MAT-20240501" in the production system and entered as UDI code "GS10012345678901234567" in the hospital's SPD system (medical supply chain management system). The two cannot be automatically linked, resulting in a situation of "inconsistent codes", which leads to information gaps when tracing the same consumable across systems.

[0029] When one of the codes is contaminated or lost, the information corresponding to the consumable will be missing. In order to solve the above technical problem, this application uses multi-code automatic mapping to enable the reading of the entire life cycle data of the consumable through multi-code collaboration when the consumable is contaminated.

[0030] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0031] Figure 1 This is a flowchart illustrating the method for accurate prediction and optimization of surgical resource consumption based on multimodal data and deep learning, as described in this application; Figure 1 As shown in the embodiments of this application, a method for accurate prediction and optimization of surgical consumables based on multimodal data and deep learning is disclosed. The method includes: S101. Collect multimodal analysis data of consumables.

[0032] The multimodal analysis data includes raw image data and current environmental data, wherein the raw image data includes at least one visually visible coded label.

[0033] Regarding the acquisition of raw image data: All visually visible coded labels on medical consumables, including inner labels, outer labels, and UDI compliance labels.

[0034] Inner labels are coded labels on the inner packaging of consumables, such as individual pouches and glass bottle stickers. When capturing the original image, the entire area of ​​the coded label, including the edge blank areas, must be covered to avoid cropping key code elements. The outer label is the coded label on the outer packaging box or large packaging bag of consumables. When collecting the original image, it is necessary to completely photograph the printed area of ​​the coded label and the surrounding related markings. UDI compliance labels, as carriers of UDI codes, must ensure that the DI (Medical Device Identification Code), PI (Production Identification Code, including batch, date, and serial number) and extended field areas on the label are fully visible in the image without obstruction or tilting / cropping during the collection of raw image data.

[0035] Image acquisition equipment typically uses industrial cameras and near-infrared cameras (NIR cameras). By combining the two cameras, it can be adapted to different pollution scenarios. The industrial camera has a resolution of ≥2MP, supports automatic exposure and autofocus, and has a frame rate of ≥25fps. The NIR camera is fixed in the 700-900nm near-infrared frequency band. This frequency band has penetrability to liquid stains (such as alcohol and disinfectants) and oil stains, and can effectively avoid pollution blockage interference under visible light. It has a resolution of ≥1MP (1280×720 and above), a frame rate of ≥30fps, and supports synchronous triggering acquisition with the industrial camera to ensure image timing consistency.

[0036] During image acquisition, 3-5 frames of RGB images are captured using an industrial camera, covering different exposure parameters to avoid overexposure or underexposure in a single frame. Additionally, a near-infrared image is simultaneously captured by combining a NIR camera with the industrial camera for synchronous triggering. Noise is reduced and image quality is improved through multi-frame fusion.

[0037] Regarding the collection of current environmental data: The barcode scanning device integrates temperature and humidity sensors, light sensors, and reflectivity sensors to obtain the current ambient temperature, humidity, light intensity, and label surface reflectivity.

[0038] The temperature and humidity sensor has a temperature measurement range of 0-60℃ and a humidity measurement range of 20%-90%RH. By collecting the ambient temperature and humidity data, it can determine whether there is a risk of high humidity (≥80% RH) or high temperature (≥40℃), providing a basis for subsequent liquid contamination inference. The light sensor has a measurement range of 0-10000 Lux and is used to detect ambient light intensity, identify strong light (≥2000 Lux) and weak light (≤50 Lux) scenes, and help judge the risk of reflection interference. The reflectivity sensor measures wavelengths of 550-650nm and is used to collect the reflectivity of the coded label surface. The reflectivity fluctuation pattern is used to determine whether there is liquid coverage (such as water stains or disinfectant) or reflection.

[0039] The scanning device has a sampling frequency of 1Hz, which means it collects one set of environmental data per second. Each set of environmental data needs to be bound to the timestamp of the corresponding image acquisition to ensure that the original image data and the current environmental data correspond one-to-one. When collecting the current environmental data, outliers are filtered out, such as temperatures exceeding the 0-60℃ range. Only valid data is retained for subsequent environmental verification.

[0040] By simultaneously collecting raw image data of medical consumables and current environmental data, a multi-dimensional analysis dataset is constructed to provide basic data support for subsequent pollution detection, coding label recognition, data recovery, and environmental verification, ensuring that the data covers all information of the consumable coding labels and the environmental characteristics of the collection scenario.

[0041] S102. Extract the label image from the original image data at the location of the assigned label, and decode it according to the pollution status of the label image to obtain the identification code.

[0042] The original image usually needs to include the complete label image. Therefore, when acquiring images, the acquired original image includes other images such as the background. Thus, by locating the bounding box, the system can automatically extract and crop the label image. The extracted label image may be complete or it may be contaminated.

[0043] By analyzing the contamination levels of the label images, the identification codes carried by the labels are obtained through decoding. Differentiated decoding strategies are developed for different contamination levels, while general adaptation rules are defined to ensure efficient and accurate extraction of identification codes even in complex medical environments.

[0044] For example, when the code elements in the label image are clear, it is generally a newly opened consumable or a coded label in a dry storage environment, which can be directly decoded; If there are minor stains or slight reflections on the edges of the label image, but the key code elements are not obstructed, it is generally due to slight water splashes or finger residue. This can be addressed by combining image optimization with decoding. When some code elements in the label image are obscured, or there is local fading or concentrated reflection, it is generally due to disinfectant spray residue or local oil stains. Image enhancement and multi-channel auxiliary methods can be used for decoding. When the core code area in the label image is severely obscured or damaged, or the overall color fades, resulting in an imbalance of code element grayscale, it is generally due to a large amount of liquid immersion, tearing of the coded label, or long-term high-humidity oxidation. In such cases, data recovery combined with database reverse lookup is required for decoding.

[0045] The identifier obtained after decoding can be at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identifier codes.

[0046] UDI codes typically consist of two parts: DI (Device Identifier) ​​and PI (Production Identifier). In some scenarios, extended fields are included. In the medical field, UDI codes are often carried out using QR codes.

[0047] The DI field, including the GS1 prefix (manufacturer identification code), product model code, specification code, packaging grade code, etc., is used to identify the constant attributes of consumables at the product level. The DI of consumables of the same model / specification remains unchanged. The PI field contains production batch number, sterilization batch number, expiration date, serial number, production date, etc., and is used to identify the dynamic attributes of consumables in the production / distribution stage. Different batches / serial numbers under the same DI have different PIs. Extended fields include manufacturer-defined information such as internal material IDs and factory codes.

[0048] The basic code is a globally unique item identifier for consumables, consisting of the static characteristics of the consumables, which can typically be the product model, specifications, internal material ID, or factory ID.

[0049] Dynamic batch codes are used to identify the attributes of consumables that change over time. They are composed of the dynamic characteristics of the consumables and can typically be production batch number, sterilization batch number, timestamp, or counter.

[0050] The base code is generated using a hash algorithm based on the product model, specifications, internal material ID, and factory ID. The base code is present in the fields of the first coding tag, RFID chip, internal packaging imprint of consumables, and UDI carrier.

[0051] For example, if the SHA-256 algorithm is used to generate the base code, then: Basic code = Hash(Product model + Specification + Internal material ID + Factory ID) For example: Hash(“surgical gauze dressing-7.5*7.5cm”+“8 layers 2 pieces / pack”+“MAT-2024-001”+“FAC-003”)=“a3f2d4e5... (64-bit string) The dynamic batch code is generated using an encryption algorithm that combines the production batch number, sterilization batch number, timestamp, and batch counter. The dynamic batch code is present on the second coding label, variable QR code layer, and production identification field of UDI on the outer packaging of the consumables.

[0052] For example, if the AES-128 encryption algorithm is used to generate dynamic batch codes, then: Dynamic batch code = Encrypt(Production batch number + Sterilization batch number + Timestamp + Counter) For example: Encrypt(“B20240501”+“S20240501001”+“20240501143000”+“0001”,“Key123456”)=“b7c8e9f0... (128-bit string) The basic code and dynamic batch code can exist in multiple coding carriers, and the first coding tag and the second coding tag can represent different physical carriers.

[0053] Since the DI field of the UDI code already contains core information such as product model, specifications, and manufacturer identification code, the UDI code and the base code can be mapped 1:1. When the association between the UDI code and the base code is realized, the system automatically queries the UDI database, compares the product model and specifications after parsing the DI field with the parameters generated by the base code, and writes the data into the multi-code mapping database (MappingTable) after a successful match, thus forming a fixed association between DI and base code.

[0054] The PI field in the UDI code already contains information of the same dimension, such as production batch number and sterilization batch number. In the production process, for each dynamic batch code generated, the system automatically extracts its production batch number and sterilization batch number, compares them with the corresponding information in the PI field of the UDI, and writes them into the multi-code mapping database after matching to form the association of DI+PI→dynamic batch code, which supports the association of multiple dynamic batch codes under the same DI (same product).

[0055] Before consumable production, a unique basic code is generated based on rules related to product model, specifications, internal material ID, and factory ID to ensure that each item has a unique basic code. After each batch of production / sterilization is completed, the system generates a unique dynamic batch code for that batch based on rules related to the production batch number, sterilization batch number, timestamp, and counter. The counter is used to distinguish different packages within the same batch. For example, under the same sterilization batch S001, the first package has a counter of 001, and the second package has a counter of 002. After the dynamic batch code is generated, the system immediately queries the multi-code mapping database, matches the corresponding basic code with the product model and specification fields, and writes the basic code-dynamic batch code association relationship into the multi-code mapping database to form the initial association link.

[0056] By decoding the label image and obtaining the identification code, as well as the mapping relationship between different types of identification codes, we can obtain multiple identification code information associated with the consumable and improve the multi-dimensional information chain corresponding to the consumable.

[0057] S103. Determine the similarity between the current environment data and the environment check code.

[0058] The environment check code is generated based on the environment data at the time the code label is generated, and the environment check code has a unique mapping relationship with the dynamic batch code.

[0059] During the medical consumables production stage, after the dynamic batch code is generated, the system synchronously collects environmental data at the time the corresponding coding label is generated. The environmental data includes temperature, humidity, light intensity, and label surface reflectivity. A unique environmental verification code is generated according to the corresponding rules, and the mapping relationship between the dynamic batch code and the environmental verification code is written into the multi-code mapping database to ensure that one dynamic batch code corresponds to only one environmental verification code, with no duplicate mappings. Furthermore, the generation parameters of the environmental verification code are strictly aligned with the production records of the dynamic batch code to avoid mapping failure due to parameter deviations.

[0060] The environmental verification code is generated from the environmental data at the time of dynamic batch code generation. It has a 1:1 unique mapping with the dynamic batch code. In essence, it is a solidified record of the environmental and physical characteristics at the time of code tag generation, and is a digital fingerprint of the initial state of the code tag.

[0061] For example, the SHA-1 algorithm can be used to generate an environment checksum. Environmental verification code = Hash(temperature + humidity + illumination parameters + tag edge features) For example: Hash(“25.3”+“45”+“300”+“[10,20,300,400]”)=“c9d8e7f6... (40-character string) Environmental verification codes can be written into the high error correction area of ​​the UDI code's QR code, such as the H-level error correction field of the QR code, or they can be stored in the internal storage area of ​​the RFID / NFC tag.

[0062] The current environmental data consists of environmental parameters collected in real time during the scanning of the coded tags, reflecting the current state of the tags. If the coded tags have not been tampered with, contaminated, or replaced, the difference between the current environmental data and the environmental data at the time of generation should be within a reasonable range, and the corresponding similarity should meet the threshold requirements.

[0063] By calculating the similarity between the two, the consistency of the labels can be verified in reverse. If the similarity is high, such as ≥0.85, it means that the label has not been replaced, the fake label does not have a corresponding initial environment check code, and the characteristics have not changed drastically due to severe pollution or damage. For example, liquid immersion will cause abnormal reflectivity and lower the similarity. If the similarity is low, it is determined that the label status is abnormal and the erroneous tracing is blocked.

[0064] S104. When the similarity is higher than the preset value, the life cycle data of the corresponding consumable is confirmed based on the mapping relationship between the identification codes.

[0065] By calculating the similarity between the current environmental data and the environmental check code associated with the dynamic batch code, when the similarity is greater than a preset threshold, it is determined that the label has not been tampered with, the pollution impact is controllable, the decoding result is reliable, and a multi-code mapping query is triggered.

[0066] Based on a multi-code mapping database, other identifier codes can be derived or associated using any of the decoded identifier codes, such as UDI codes, basic codes, and dynamic batch codes, through preset mapping relationships.

[0067] When a UDI code is identified, the basic code and dynamic batch code can be obtained by parsing the DI code and PI code. The three types of identification codes can be confirmed by verifying the similarity between the current environment data and the environment check code. When the DI code of the UDI code is identified, the base code can be obtained by parsing the DI code. Since the base code corresponds to multiple dynamic batch codes, the similarity between the current environmental data and the environmental check codes corresponding to the multiple dynamic batch codes can be calculated to filter the environmental check codes and the corresponding dynamic batch codes that are higher than the threshold, thereby realizing the confirmation of the three types of identification codes. When the basic code is identified, since the basic code corresponds to multiple dynamic batch codes, the environmental check codes corresponding to the current environmental data and multiple dynamic batch codes can be used to calculate the similarity and filter out environmental check codes that are higher than the threshold and their corresponding dynamic batch codes, thereby determining the UDI code and realizing the confirmation of the three types of identification codes. When a dynamic batch code is identified, the three types of identification codes can be confirmed because each dynamic batch code uniquely corresponds to a basic code and a UDI code.

[0068] By using the associated complete identification code, the full lifecycle data of the corresponding consumables can be retrieved from the database to ensure the integrity of the traceability chain.

[0069] This application provides a multi-code fusion traceability method for contamination-resistant medical consumables. By constructing a multi-layered coding system of basic code, dynamic batch code, and environmental verification code, combined with a multi-code mapping database, it achieves automatic association between UDI code, basic code, and dynamic batch code, solving the problem of inconsistent codes in traditional methods and ensuring uninterrupted traceability information. Furthermore, by collecting raw image data and simultaneously integrating temperature, humidity, light, and reflectivity sensors to collect current environmental data, a multi-dimensional analysis dataset is constructed to support contamination detection and coded label identification. This method adapts to complex contamination environments in medical settings. Based on the contamination status of the coded labels, a differentiated decoding strategy is adopted for different contamination levels to decode the identification code. Combined with the environmental verification code and the similarity judgment of the current environmental data, label consistency is verified, preventing fake label replacement and misidentification, ensuring the authenticity of traceability information. Through multiple types of identification codes, it achieves full lifecycle data traceability of consumables, reducing manual intervention, lowering costs, and improving medical safety and the reliability of regulatory traceability.

[0070] Figure 2 A flowchart illustrating a multi-code fusion traceability method for anti-contamination medical consumables provided in an embodiment of this application is shown below. Figure 2 As shown, this embodiment, based on the above embodiments, includes the following method: S201. Collect multimodal analysis data of consumables.

[0071] For a detailed description of step S201, please refer to step S101.

[0072] S202. Extract the label image from the coded label location in the original image data.

[0073] Based on the geometric features of the coded labels, such as rectangular boundaries, locators, and texture features, image processing algorithms are used to achieve automatic positioning and cropping of the label region.

[0074] The coded label region is accurately cropped from the original image, eliminating irrelevant background interference and reducing the computational load of subsequent segmentation and decoding.

[0075] S203. Perform pixel-level segmentation based on the labeled image, and label the mask region according to the segmented pixel-level mask.

[0076] The system uses lightweight U-Net with MobileNetV2 encoder and symmetrical decoder structure. The input is a label image, which is an RGB image. The output is a pixel-level mask (Mask_damaged), and the mask area includes the background, code area, contaminated area, and occluded area.

[0077] S204. Calculate the contaminated area ratio based on the number of pixels in the contaminated area and the total number of pixels in the code area.

[0078] The contaminated area ratio r = number of pixels in the contaminated area / total number of pixels in the code area.

[0079] S205. Classification based on the ratio of polluted area.

[0080] The pollution levels are classified into four categories based on a preset range of pollution area ratios: no pollution, light pollution, moderate pollution, and heavy pollution.

[0081] Specifically: r < 0.03: No pollution level; 0.03 ≤ r < 0.10: Light pollution level; 0.10≤r<0.30: Moderate pollution level; r≥0.30: Severe pollution level.

[0082] S206. Select the corresponding decoding strategy according to the level classification to obtain the identification code.

[0083] When there is no pollution level, the label image is scanned using a standard decoder to obtain the identification code; Image preprocessing is performed on the label image by performing basic geometric correction and single-frame Gaussian filtering for noise reduction. Then, standard barcode or QR code decoders such as ZXing and Libdmtx are used to directly read the encoded information in the RGB image of the scanned label.

[0084] At the level of light pollution, enhance the label image, acquire and decode the near-infrared image of the enhanced label image to obtain the identification code; By performing multi-frame fusion and adaptive histogram equalization (CLAHE) on the javelin images to enhance contrast and eliminate slight reflections or dirt interference, the tag images are then acquired simultaneously using a NIR camera to penetrate shallow liquids, oil stains, and other contaminants. The RGB enhanced images and near-infrared images are decoded in parallel. If the results are consistent, they are output directly; if there are differences, the near-infrared decoding result is used because it has a stronger resistance to contamination.

[0085] At the moderate pollution level, the pollution area of ​​the label image is located by semantic segmentation, the missing code elements in the pollution area are filled, and the missing code elements are completed based on the preset encoding rules of the identifier code to obtain the identifier code; A lightweight U-Net model is used to perform pixel-level semantic segmentation, marking the location and extent of contaminated areas, distinguishing between two types of contamination states: "complete occlusion" and "blurred". Local repair algorithms are performed on contaminated areas. Small-area contamination is filled with the Navier-Stokes PDE algorithm, and diffuse contamination is matched and replaced with valid textures using the PatchMatch algorithm. Based on preset encoding rules, such as the DI segment length of the UDI code and the factory ID prefix of the base code, blurred or missing fields are filled in, and a unique identifier code is output.

[0086] The severe pollution level is determined by using depth prediction to complete the labeled image and obtain the identification code, as follows: Based on the fixed grid characteristics of QR codes, a standard grid is constructed by detecting positioning points, and the state of each symbol is marked to obtain the sequence of damaged symbols; Input the current environmental data and the sequence of damaged symbols into the trained deep prediction model to complete the label image and obtain the identification code.

[0087] Specifically, the finder pattern and alignment pattern of the QR code are identified to determine the geometric boundaries of the label image. Based on the fixed grid characteristics of the QR code, an N×N standard grid (such as the 21×21 or 41×41 grid of the QR code) is constructed, the pixels of the label image are mapped, and the status of the code elements in the grid is determined one by one, marked as "valid (0 / 1)" or "damaged (missing / blurred)", and a damaged code element sequence is generated.

[0088] The damaged symbol sequence is flattened and encapsulated with standardized current environmental data (temperature, humidity, light intensity, reflectivity) into an input vector. This vector is then input into a trained Vision Transformer+CNN hybrid model. The model predicts the binary values ​​of the missing symbols through global sequence modeling and local feature extraction, generates a complete symbol sequence, completes the label image, and converts it into a readable identifier code. This is then verified by cross-referencing multiple databases in a multi-code mapping database to output the final valid identifier code.

[0089] The types of identification codes are determined based on their explicit fixed features and multi-dimensional implicit features.

[0090] Explicit fixed features include encoding length, prefix format, character set, and check bit rules, while multi-dimensional implicit features include segmentation structure, character distribution, check bit association, and semantic association.

[0091] Encoding length: Counts the total number of identifier characters and matches them within a preset length range, such as UDI code (24-34 bits) or basic code (32 / 64 bits); Prefix format: Detects the starting character of the encoding, such as the GS prefix "01" or "10" in UDI codes, and internal prefixes such as "FAC-" or "MAT-" in basic codes; Character set: Determines the types of characters contained in the encoding. For example, UDI code consists of numbers and a small number of separators, while basic code consists of letters (AF) and numbers in a hash format. Check bit rules: Verify the validity of the code check bit, such as the Mod10 check of the DI segment of the UDI code, and the Hash check bit matching logic of the basic code.

[0092] Segmented structure: Determine whether the encoding has fixed segments, such as the DI segment of the UDI code being 14 bits and the PI segment being of variable length, while the basic code has no explicit segments; Character distribution: Statistics on the proportion of letters, numbers, and separators. For example, in basic codes, letters and numbers are evenly distributed, while in UDI codes, numbers are predominant. Check bit association: Calculate the association degree between the check bit and the preceding character to verify whether it conforms to the check algorithm logic of the corresponding encoding, such as UDI code Mod10 and basic code Hash check. Semantic association: Detect whether the encoding contains semantic fields such as timestamps and batch numbers. For example, the date format "20240501" in the PI segment of the UDI code has no semantic fields in the basic code.

[0093] Explicit features are filtered using regular expressions, and latent features are verified using a feature vector classifier. A two-stage discrimination process is used to first exclude obviously non-compliant types, then calculate the feature matching degree, and output the encoding type and confidence score. For example, if the confidence score of "basic code" is 0.95, it means that the identifier code is a basic code. When the confidence score is lower than the threshold, it can be set to trigger manual review.

[0094] S207. Determine the similarity between the current environment data and the environment check code.

[0095] S208. When the similarity is higher than the preset value, the life cycle data of the corresponding consumable is confirmed based on the mapping relationship between the identification codes.

[0096] For a detailed description of steps S207-S208, please refer to steps S103-S104.

[0097] This application provides a multi-code fusion traceability method for contamination-resistant medical consumables. It acquires multimodal images using industrial and near-infrared cameras, combined with environmental sensor data. After precise cropping and pixel-level segmentation of the label image region, it classifies the contamination area into four levels and employs differentiated strategies for different contamination scenarios. Furthermore, it accurately identifies the coding type of the identification code through two-level discrimination using explicit and implicit features. Finally, it combines environmental verification and multi-code mapping to ensure the authenticity and completeness of traceability data, adapting to various contamination scenarios, guaranteeing the reliability of identification code acquisition, and achieving seamless traceability throughout the entire lifecycle of consumables, thus solving the problem of traceability gaps caused by contamination.

[0098] Figure 3 This application provides a schematic diagram of the structure of a multi-code fusion traceability system for anti-contamination medical consumables, as shown in one embodiment. Figure 3 As shown, the anti-contamination medical consumables multi-code fusion traceability system 30 provided in this embodiment includes: The data acquisition module 301 is used to acquire multimodal analysis data of consumables. The multimodal analysis data includes raw image data and current environmental data. The raw image data includes at least one visually visible coded label. The decoding processing module 302 is used to extract the label image at the coded label location in the original image data, and decode it according to the pollution status of the label image to obtain the identification code. The identification code is at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identification codes. The basic code is the globally unique item identifier of the consumable, and the dynamic batch code is used to identify the attributes of the consumable that change over time. The tag verification module 303 is used to determine the similarity between the current environmental data and the environmental verification code. The environmental verification code is generated based on the environmental data when the coded tag is generated, and the environmental verification code has a unique mapping relationship with the dynamic batch code. The data output module 304 is used to confirm the life cycle data of the corresponding consumable based on the mapping relationship between the identification codes when the similarity is higher than a preset value.

[0099] This embodiment provides a multi-code fusion traceability system for anti-pollution medical consumables, which can execute the methods provided in the above-described method embodiments. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0100] Figure 4 This application provides a schematic diagram of the structure of a multi-code fusion traceability device for anti-pollution medical consumables, as shown in one embodiment. Figure 4 As shown, the anti-contamination medical consumable multi-code fusion traceability device 40 provided in this embodiment includes: The device 40 includes at least one processor 401 and a memory 402. Optionally, the device 40 also includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.

[0101] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.

[0102] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0103] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0104] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0105] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0106] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0107] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0108] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0109] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0110] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0111] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0113] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0115] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for multi-code fusion traceability of pollution-resistant medical consumables, characterized in that, The method includes: Collect multimodal analysis data of consumables, the multimodal analysis data including raw image data and current environmental data, wherein the raw image data includes at least one visually visible coded label; The label image at the coded label location is extracted from the original image data, and the label image is decoded according to the pollution status to obtain the identification code. The identification code is at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identification codes. The basic code is a globally unique item identifier for consumables, and the dynamic batch code is used to identify the attributes of consumables that change over time. Determine the similarity between the current environmental data and the environmental check code. The environmental check code is generated based on the environmental data when the code assignment label is generated, and the environmental check code has a unique mapping relationship with the dynamic batch code. When the similarity is higher than a preset value, the lifecycle data of the corresponding consumable is confirmed based on the mapping relationship between the identification codes.

2. The method according to claim 1, characterized in that, The label image is decoded based on its contamination level to obtain the identification code, including: Pixel-level segmentation is performed based on the labeled image, and the mask region is marked according to the segmented pixel-level mask. The mask region includes background, code region, contaminated region, and occlusion region. The contaminated area ratio is calculated based on the number of pixels in the contaminated area and the total number of pixels in the code area; The pollution levels are classified according to the percentage of contaminated area. Based on the classification, the corresponding decoding strategy is selected first to obtain the identification code.

3. The method according to claim 2, characterized in that, The step of prioritizing the selection of the corresponding decoding strategy based on the level classification to obtain the identifier code includes: The pollution level classification is set according to a preset range of pollution area ratios, including no pollution level, light pollution level, moderate pollution level, and heavy pollution level. When there is no pollution level, the label image is scanned using a standard decoder to obtain the identification code; At a light pollution level, the label image is enhanced, a near-infrared image of the enhanced label image is acquired and decoded to obtain the identification code; At the moderate pollution level, the pollution area of ​​the label image is located by semantic segmentation, the missing code elements in the pollution area are filled, and the missing code elements are completed based on the preset encoding rules of the identifier code to obtain the identifier code; The severe pollution level is determined by using depth prediction to complete the label image and obtain the identification code.

4. The method according to claim 3, characterized in that, The identification code is obtained by completing the label image using depth prediction, including: Based on the fixed grid characteristics of QR codes, a standard grid is constructed by detecting positioning points, and the state of each symbol is marked to obtain the sequence of damaged symbols; Input the current environment data and the damaged code sequence into the trained deep prediction model to complete the label image and obtain the identification code.

5. The method according to claim 4, characterized in that, Obtaining the identifier code includes: The type of the identifier is obtained based on the explicit fixed features and multi-dimensional implicit features of the identifier. The explicit fixed features include encoding length, prefix format, character set, and check bit rules. The multi-dimensional implicit features include segmentation structure, character distribution, check bit association, and semantic association.

6. The method according to claim 1, characterized in that, The basic code is generated using a hash algorithm based on the product model, specifications, internal material ID, and factory ID, and the basic code exists in the fields of the first coding tag, RFID chip, internal packaging imprint of consumables, and UDI carrier.

7. The method according to claim 1, characterized in that, The dynamic batch code is generated using an encryption algorithm that combines the production batch number, sterilization batch number, timestamp, and batch counter. The dynamic batch code is present on the second coding label, variable QR code layer, and production identification field of UDI on the outer packaging of the consumables.

8. A multi-code fusion traceability system for pollution-resistant medical consumables, characterized in that, The system includes: The data acquisition module is used to acquire multimodal analysis data of consumables. The multimodal analysis data includes raw image data and current environmental data. The raw image data includes at least one visually visible coded label. The decoding processing module is used to extract the label image at the coded label location in the original image data, and decode it according to the pollution status of the label image to obtain the identification code. The identification code is at least one of UDI code, basic code, and dynamic batch code, and there is a mapping relationship between the various identification codes. The basic code is a globally unique item identifier for consumables, and the dynamic batch code is used to identify the attributes of consumables that change over time. The tag verification module is used to determine the similarity between the current environment data and the environment verification code. The environment verification code is generated based on the environment data when the code tag is generated, and the environment verification code has a unique mapping relationship with the dynamic batch code. The data output module is used to confirm the lifecycle data of the corresponding consumables based on the mapping relationship between the identification codes when the similarity is higher than a preset value.

9. A pollution-resistant multi-code fusion traceability device for medical consumables, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images