Medical material intelligent management and control traceability method and device based on artificial intelligence

Abstract

The invention discloses a medical material intelligent control traceability method and device based on artificial intelligence, and the method comprises the steps: collecting the face image and operation information of an operator during the management operation of various medical materials, and carrying out the face recognition of the face image; therefore, the personnel information of the medical materials in various material management operations can be obtained. Then, real traceability information is generated by using the operator information and the operation information; and finally, the traceability code is associated with the traceability code of the medical material, and traceability management and control of the medical material can be completed. Therefore, complete and accurate traceability information can be formed, and traceability management of each operation and personnel in the whole life cycle of the medical materials can be realized, so that the method is very suitable for large-scale application and popularization in the technical field of medical material traceability management and control.

Description

Technical Field

[0001] The present invention belongs to the technical field of traceability control of medical supplies, and particularly relates to an intelligent control and traceability method and device for medical supplies based on artificial intelligence. Background Art

[0002] At present, there are many types and large quantities of hospital supplies. Strengthening the management of hospital material supplies is of great significance. Improving the management level of supplies, managing the supply work well, rationalizing the inventory quantity and capital occupancy, reducing the situation of materials expiring, avoiding a large amount of waste of materials, and improving the effective utilization rate of various hospital materials are the top priorities for realizing the optimal allocation of hospital materials and promoting the sustainable development of the hospital.

[0003] Currently, hospitals usually manage supplies by automatically generating in-hospital barcodes according to the barcode rules set in the in-hospital system. However, such a management system can only perform simple barcode scanning registration during use (such as in processes like warehousing and outbound), so it can only manage the use of the supplies themselves and cannot associate the users of the supplies. And even if personnel management is carried out during registration, it is only manual registration, which has the problem of inaccurate information. Thus, when it is necessary to trace the use of supplies, complete and accurate traceability information is often not formed, and the complete traceability of medical supplies cannot be achieved. Based on this, how to provide an intelligent control and traceability method for medical supplies that can form complete and accurate traceability information has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent control and traceability method and device for medical supplies based on artificial intelligence to solve the problem that the prior art cannot form complete and accurate traceability information, thus unable to achieve the complete traceability of medical supplies.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] In the first aspect, an intelligent control and traceability method for medical supplies based on artificial intelligence is provided, including:

[0007] Obtain the product information and captured images of the target medical supplies, and generate a traceability code for the target medical supplies based on the product information and the captured images, where the product information includes a product unique identification code;

[0008] Obtain the original traceability information of the target medical supplies during each material management operation, where the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include warehousing operations, application operations, and logistics operations;

[0009] Perform face recognition processing on the face images of the operators in each original traceability information to obtain the operator information of the target medical supplies when performing each material management operation;

[0010] Generate the true traceability information of the target medical supplies when performing each material management operation by using the operator information and operation information of the target medical supplies when performing each material management operation;

[0011] Perform an association process between each true traceability information and the traceability code, and store the associated true traceability information in a database, so that after the user terminal scans the traceability code, all the true traceability information of the target medical supplies during use can be found from the database to complete the traceability control of the target medical supplies.

[0012] Based on the above disclosed content, the present invention first obtains the product unique identification code and the captured image of the target medical supplies, and generates a traceability code based on the two; then, collects the original traceability information of the target medical supplies when performing each material management operation; then, performs face recognition on the face images of the operators in each collected original traceability information to obtain the information of each operator; then, the true traceability information of the target medical supplies when performing material management operations such as warehousing, application, and logistics can be generated by using the operator information and operation information during each material management operation; finally, associate each true traceability information with the aforementioned traceability code to complete the traceability control of the target medical supplies; thus, when it is necessary to trace the target medical supplies, by scanning the traceability code, all the true traceability information associated with the traceability code in the database can be found, so as to obtain all the operation information and the corresponding personnel information of the target medical supplies during use.

[0013] Through the above design, the present invention collects the face images and operation information of the operators when the medical supplies perform various material management operations, and performs face recognition on the face images to obtain the personnel information of the medical supplies when performing various material management operations; then, uses the operator information and operation information to generate true traceability information; finally, associates it with the traceability code of the medical supplies to complete the traceability control of the medical supplies; thus, the present invention can form complete and accurate traceability information, and can realize the traceability management of each operation and personnel during the entire life cycle of the medical supplies. Therefore, it is very suitable for large-scale application and promotion in the technical field of medical supplies traceability control.

[0014] In a possible design, generating the traceability code of the target medical supplies based on the product information and the captured image includes:

[0015] Perform image recognition processing on the captured image to obtain the image feature information of the captured image, where the image feature information is used to characterize the classification information of the target medical supplies;

[0016] Obtain a key generation function and use the key generation function to generate a first encryption key;

[0017] Obtain the capture time of the captured image and perform encryption processing on the capture time using an irreversible encryption algorithm to obtain a second encryption key;

[0018] Generate a third encryption key according to the first encryption key and the second encryption key;

[0019] Use the third encryption key to encrypt the product unique identifier in the product information to obtain encrypted information;

[0020] Generate a traceability code for the target medical supplies based on the encrypted information and the image feature information.

[0021] In a possible design, the third encryption key is a binary sequence. Among them, using the third encryption key to encrypt the product unique identifier in the product information to obtain encrypted information includes:

[0022] Divide the product unique identifier into character segments of length k, where k is a positive integer greater than 1;

[0023] Convert the binary sequence into a decimal value, and based on the decimal value, circularly shift each character segment to the right or left to obtain updated character segments;

[0024] Perform an exclusive OR operation on the third encryption key and the updated character segments to obtain multiple encrypted character segments;

[0025] Perform a splicing process on multiple encrypted character segments to obtain the encrypted information;

[0026] Correspondingly, generating a traceability code for the target medical supplies based on the encrypted information and the image feature information includes:

[0027] Generate an anti-counterfeiting pattern based on the encrypted information;

[0028] Generate an initial traceability code using the image feature information;

[0029] Add the anti-counterfeiting pattern to the initial traceability code to obtain the traceability code for the target medical supplies.

[0030] In a possible design, the key generation function is:

[0031]

[0032] In formula (1), represents the key generation function, m i represents the first random variable, p represents the number of encryption iterations, P represents the maximum number of encryption iterations, x p represents the second random variable at the p-th time, and i represents the first encryption parameter;

[0033] wherein, x p = α × x p-1 × (1 - x p-1 ) (2);

[0034] In formula (2), α represents the second encryption parameter, wherein when p is 1, x p-1 is the initial value, and x p-1 ∈ [0, 1];

[0035]

[0036] In formula (3), represents the random parameter function,

[0037] In a possible design, face recognition processing is performed on the face images of the operators in each original traceability information to obtain the operator information when the target medical supplies are performing each material management operation, including:

[0038] For any face image, extract the low-frequency features and high-frequency features of the any face image;

[0039] Based on the low-frequency features, high-frequency features and the any face image, generate the face feature vector of the any face image;

[0040] Obtain an improved face recognition model, wherein the improved face recognition model is trained with the sample face feature vectors of multiple sample faces as inputs and the face recognition results of each sample face as outputs;

[0041] Input the face feature vector into the improved face recognition model to obtain the operator information when the target medical supplies are performing the specified material management operation, wherein the specified material management operation is the material management operation associated with the original traceability information corresponding to the any face image.

[0042] In a possible design, the improved face recognition model includes: a first convolutional structure layer, a first max-pooling layer, a second convolutional structure layer, a second max-pooling layer, a third convolutional structure layer, a third max-pooling layer, a fourth convolutional structure layer, a fourth max-pooling layer, a fifth convolutional structure layer, a fifth max-pooling layer, a fully-connected layer, and an output layer that are connected in sequence. Moreover, the improved face recognition model further includes: a sixth convolutional structure layer, a first feature fusion layer, a sixth max-pooling layer, a seventh convolutional structure layer, and a second feature fusion layer;

[0043] The output end of the third max-pooling layer is connected to the input end of the sixth convolutional structure layer, and the output ends of the fourth convolutional structure layer and the sixth convolutional structure layer are respectively connected to the input end of the first feature fusion layer for feature fusion through the first feature fusion layer;

[0044] The output end of the first feature fusion layer is connected to the input end of the sixth max-pooling layer, and the output end of the sixth max-pooling layer is connected to the input end of the seventh convolutional structure layer. Among them, the output end of the seventh convolutional structure layer and the output end of the fifth convolutional structure layer are connected to the input end of the second feature fusion layer, and the output end of the second feature fusion layer is connected to the input end of the fifth max-pooling layer.

[0045] In a possible design, both the sixth convolutional structure layer and the seventh convolutional structure layer include a first convolutional layer, a second convolutional layer, a third convolutional layer, a fourth convolutional layer, and a fifth convolutional layer, and the number of convolutional kernels of the first convolutional layer, the second convolutional layer, and the third convolutional layer is 64, and the number of convolutional kernels of the fourth convolutional layer and the fifth convolutional layer is 128;

[0046] For the sixth convolutional structure layer, the output end of the third max-pooling layer is respectively connected to the input ends of the first convolutional layer, the second convolutional layer, and the third convolutional layer. Among them, the output end of the second convolutional layer is connected to the input end of the fourth convolutional layer, the output end of the third convolutional layer is connected to the input end of the fifth convolutional layer, and the output ends of the first convolutional layer, the fourth convolutional layer, and the fifth convolutional layer are all connected to the input end of the first feature fusion layer.

[0047] In a second aspect, there is provided an intelligent control and traceability device for medical supplies based on artificial intelligence, including:

[0048] An acquisition unit for acquiring product information and a captured image of a target medical supply, and generating a traceability code for the target medical supply based on the product information and the captured image, where the product information includes a product unique identification code;

[0049] An acquisition unit, configured to acquire the original traceability information of the target medical supplies during each material management operation, where the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include warehousing operations, application operations, and logistics operations;

[0050] A face recognition unit, configured to perform face recognition processing on the face images of the operators in each original traceability information to obtain the operator information of the target medical supplies during each material management operation;

[0051] A traceability information generation unit, configured to generate the true traceability information of the target medical supplies during each material management operation by using the operator information and operation information of the target medical supplies during each material management operation;

[0052] A traceability management unit, configured to perform an association process between each true traceability information and the traceability code, and store the associated true traceability information in a database, so that after the user terminal scans the traceability code, all the true traceability information of the target medical supplies during the use process can be found from the database to complete the traceability control of the target medical supplies.

[0053] In a third aspect, another intelligent control and traceability device for medical supplies based on artificial intelligence is provided. Taking the device as an electronic device as an example, it includes a memory, a processor, and a transceiver that are communicatively connected in sequence. Among them, the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the intelligent control and traceability method for medical supplies based on artificial intelligence as described in the first aspect or any possible design in the first aspect.

[0054] In a fourth aspect, a storage medium is provided, on which instructions are stored. When the instructions run on a computer, the intelligent control and traceability method for medical supplies based on artificial intelligence as described in the first aspect or any possible design in the first aspect is executed.

[0055] In a fifth aspect, a computer program product containing instructions is provided. When the instructions run on a computer, the computer is made to execute the intelligent control and traceability method for medical supplies based on artificial intelligence as described in the first aspect or any possible design in the first aspect.

[0056] Beneficial effects:

[0057] (1) The present invention collects the face images and operation information of the operators when various medical supplies management operations are carried out, and performs face recognition on the face images to obtain the personnel information of the medical supplies during various management operations. Then, using the operator information and operation information, real traceability information is generated. Finally, by associating it with the traceability code of the medical supplies, the traceability control of the medical supplies can be completed. Thus, the present invention can form complete and accurate traceability information, and can realize the traceability management of various operations and personnel during the entire life cycle of medical supplies. Therefore, it is very suitable for large-scale application and promotion in the technical field of medical supplies traceability control.

[0058] (2) The present invention takes the image feature information of the target medical supplies as verification information, and the encrypted product unique identification code of the target medical supplies as anti-counterfeiting information and adds them to the traceability code. In this way, the verifiability and anti-counterfeiting property of the traceability code can be guaranteed, thereby improving the reliability of traceability. Brief Description of the Drawings

[0059] Figure 1 It is a schematic diagram of the steps of the intelligent control and traceability method for medical supplies based on artificial intelligence provided by an embodiment of the present invention;

[0060] Figure 2 It is a network structure diagram of the improved face recognition model provided by an embodiment of the present invention;

[0061] Figure 3 It is a schematic diagram of the structure of the intelligent control and traceability device for medical supplies based on artificial intelligence provided by an embodiment of the present invention;

[0062] Figure 4 It is a schematic diagram of the structure of the electronic device provided by an embodiment of the present invention. Detailed Embodiment

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the present invention in combination with the drawings and the descriptions of the embodiments or the prior art. Obviously, the following descriptions of the structures of the drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation to the present invention.

[0064] It should be understood that although terms such as first, second, etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, the first unit may be referred to as the second unit, and similarly, the second unit may be referred to as the first unit, without departing from the scope of the exemplary embodiments of the present invention.

[0065] It should be understood that for the term "and / or" that may appear herein, it is merely an association relationship describing associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and A and B exist simultaneously; for the term " / and" that may appear herein, it describes another association object relationship, indicating that two relationships may exist. For example, A / and B may represent: A exists alone, and A and B exist alone; in addition, for the character " / " that may appear herein, generally it represents that the associated objects before and after are an "or" relationship.

[0066] Embodiment:

[0067] See Figure 1 As shown, for the intelligent control and traceability method of medical supplies based on artificial intelligence provided in this embodiment, first, a traceability code is generated based on the unique product identification code of the medical supplies and the captured image. Then, when performing material management operations such as warehousing, requisition, and logistics of medical supplies, the corresponding operation information is obtained, and face recognition is performed on the corresponding operating personnel to obtain the operator information. Finally, by associating the operation information and the operator information when the medical supplies are performing various material management operations with their corresponding traceability codes, the traceability control of the medical supplies can be completed. In this way, the present invention can realize the traceability management of various operations and personnel during the entire life cycle of medical supplies, and thus is very suitable for large-scale application and promotion in the technical field of medical supply traceability control. Among them, for example, this method can but is not limited to running on the server side. It can be understood that the foregoing execution subject does not constitute a limitation to the embodiments of the present application. Correspondingly, the running steps of this method can but are not limited to the following steps S1 to S5.

[0068] S1. Obtain the product information and the captured image of the target medical supplies, and generate a traceability code for the target medical supplies based on the product information and the captured image, where the product information includes a unique product identification code; in this embodiment, for example, the captured image of the target medical supplies may be, but is not limited to, the image captured during the first storage, and at the same time, the product information may also be collected during the first storage; optionally, this embodiment generates the traceability code for the target medical supplies based on the feature information in the captured image for characterizing the classification information of the target medical supplies and the unique product identification code; in this way, the generated traceability code can be bound to the target medical supplies, thus ensuring the accuracy of the correspondence between the traceability code and the supplies.

[0069] In practical applications, for example, but not limited to, the following steps S11 to S16 may be adopted to generate the aforementioned traceability code.

[0070] S11. Perform image recognition processing on the captured image to obtain the image feature information of the captured image, where the image feature information is used to characterize the classification information of the target medical supplies; in specific applications, for example, but not limited to, input the aforementioned captured image into an image recognition model to obtain the aforementioned image feature information; specifically, the image feature information is the feature vector output by the layer before the output layer (i.e., the softmax classification layer) of the image recognition model; that is, the output layer of the image recognition model classifies the feature vector to obtain the classification result of the target medical supplies; further, for example, the image recognition model may include, but is not limited to, a convolutional neural network model, in this case, the feature vector output by the fully connected layer before the output layer in the convolutional neural network model is used as the image feature information of the captured image.

[0071] After obtaining the image feature information of the captured image, it can be used as a verification information and added to the traceability code to be generated for subsequent verification of the traceability code; at the same time, to ensure the anti-counterfeiting property of the traceability code, this embodiment also encrypts the unique product identification code of the target medical supplies to generate encrypted information, and adds the encrypted information as anti-counterfeiting information to the traceability code to be generated, so as to achieve the anti-counterfeiting of the traceability code; where the encryption process may be, for example, as shown in the following steps S12 to S15.

[0072] S12. Obtain a key generation function and use the key generation function to generate a first encryption key; in this embodiment, for example, the key generation function may be, but is not limited to, as shown in the following formula (1).

[0073]

[0074] In the above formula (1), represents the key generation function, mi represents the first random variable, p represents the number of encryption iterations, P represents the maximum number of encryption iterations, and x p represents the second random variable at the p-th time, and i represents the first encryption parameter; wherein, in this embodiment, the first encryption parameter and the maximum number of encryption iterations are randomly obtained each time encryption is performed. Optionally, for this encryption, i can be taken as 3 and P can be taken as 60.

[0075] Optionally, for example, x p = α × x p-1 × (1 - x p-1 ) (2);

[0076]

[0077] In the above formula (2), α represents the second encryption parameter. Among them, when p is 1, x p-1 is the initial value, and x p-1 ∈ [0, 1]; similarly, in formula (3), represents the random parameter function,

[0078] Thus, based on the foregoing key generation function, the generation of the first encryption key can be carried out. That is, take the first encryption parameter as 3 and the maximum number of encryption iterations as 60; then substitute the foregoing two parameters into the foregoing formula (1), and a binary sequence (i.e., ) can be obtained, and this binary sequence is used as the first encryption key.

[0079] After obtaining the first encryption key, the generation of the second encryption key can be carried out, as shown in the following step S13.

[0080] S13. Obtain the shooting time of the captured image, and use an irreversible encryption algorithm to encrypt the shooting time to obtain the second encryption key; in this embodiment, for example, but not limited to, algorithms such as MD5 can be used to encrypt the shooting time to obtain the second encryption key; of course, other irreversible encryption algorithms can also be used for encryption, and it is not limited to the foregoing example here.

[0081] After obtaining the second encryption key, the first encryption key can be combined to generate the third encryption key; among them, the generation process of the third encryption key can be, but not limited to, as shown in the following step S14.

[0082] S14. Generate a third encryption key according to the first encryption key and the second encryption key; in this embodiment, by concatenating the first encryption key and the second encryption key, the third encryption key can be obtained; alternatively, randomly select multiple characters from the first encryption key and the second encryption key to form the third encryption key; thus, the third encryption key is also essentially a binary sequence; therefrom, based on the third encryption key, the product unique identification code of the target medical supplies can be encrypted to obtain encrypted information; wherein, the encryption process can be but is not limited to the steps shown in S15 below.

[0083] S15. Use the third encryption key to encrypt the product unique identification code in the product information to obtain encrypted information; in this embodiment, for example, but not limited to, the following steps S15a to S15d can be adopted to encrypt the product unique identification code.

[0084] S15a. Divide the product unique identification code into character segments of length k, where k is a positive integer greater than 1; in this embodiment, for example, k is taken as 4, and when dividing for the last time, if the remaining length of the product unique identification code is less than 4, the remaining part can be directly used as the last character segment; of course, the division length can be specifically set according to actual use, and is not limited to the foregoing example here.

[0085] After completing the division of the product unique identification code, preprocessing of each character segment can be carried out, and the preprocessing process is as shown in the following step S15b.

[0086] S15b. Convert the binary sequence into a decimal value, and based on the decimal value, cyclically shift each character segment to the right or left to obtain updated character segments; in this embodiment, it is equivalent to cyclically shifting each character segment to the left or right by R bits, and R is the decimal value corresponding to the third encryption key.

[0087] After completing the preprocessing of each character segment, encryption processing can be carried out, as shown in the following step S15c and step S15d.

[0088] S15c. Perform an exclusive OR operation on the third encryption key and the updated character segments to obtain multiple encrypted character segments; in this embodiment, it is equivalent to performing an exclusive OR operation on each updated character segment with the third encryption key to obtain each encrypted character segment; then, concatenate them in the division order to obtain the encrypted information, and the concatenation process can be but is not limited to the steps shown in the following step S15d.

[0089] S15d. Perform a concatenation process on multiple encrypted character segments to obtain the encrypted information.

[0090] Thus, through the foregoing steps S15a to S15d, the encryption of the unique product identification code can be completed. Then, in combination with the foregoing image feature information, the traceability code of the target medical supplies can be generated. The generation process of the traceability code is as shown in the following step S16.

[0091] S16. Generate the traceability code of the target medical supplies based on the encrypted information and the image feature information. In this embodiment, for example, but not limited to, first generate an anti-counterfeiting pattern based on the encrypted information. Then, use the image feature information to generate an initial traceability code. Finally, add the anti-counterfeiting pattern to the initial traceability code to obtain the traceability code of the target medical supplies. At the same time, for example, the traceability code can be marked on the target medical supplies so that the target medical supplies can be traced by scanning the traceability code subsequently. Optionally, for example, the initial traceability code can be associated with traceability information (i.e., the following real traceability information), and the anti-counterfeiting image is associated with the foregoing encrypted information. In this way, a traceability code with anti-counterfeiting identification can be generated.

[0092] At the same time, for example, the foregoing encrypted information and image feature information are both associated with the traceability code and stored in the database.

[0093] Furthermore, for example, when making the identification, two kinds of inks or toners that are sensitive to different wavelength bands of light can be used to print the initial traceability code and the anti-counterfeiting pattern on the target medical supplies, so as to form the traceability code of the target medical supplies. Then, a barcode scanner with different emission wavelength bands can be used to scan the traceability code to obtain the anti-counterfeiting pattern and the initial traceability code in the traceability code. In this way, the encrypted information associated with the anti-counterfeiting pattern can be obtained, and then it is compared with the encrypted information stored in the database. If the two are consistent, it can be shown that the traceability code is a real traceability code; otherwise, it is a counterfeit traceability code. Similarly, through the real traceability information associated with the initial traceability code, the tracing of the target medical supplies can be completed.

[0094] In addition, for example, secondary authentication of the traceability code can also be performed, that is, when the user terminal performs tracing, it takes a picture of the target medical supplies to obtain a tracing image. Then, the tracing image is transmitted to the server. Then, the server inputs the tracing image into the foregoing image recognition model to obtain the image feature information of the tracing image (hereinafter referred to as the information to be verified). Among them, if the information to be verified is consistent with the image feature information of the foregoing captured image, it means that the target medical supplies during tracing are the same as the target medical supplies in the foregoing step S1, and there is no problem of replacement or counterfeiting between the two. In this way, the reliability of tracing can be ensured.

[0095] Through the foregoing step S1 and its sub-steps, the present invention takes the image feature information of the target medical supplies as verification information, and the encrypted product unique identification code of the target medical supplies as anti-counterfeiting information and adds them to the traceability code. In this way, the verifiability and anti-counterfeiting property of the traceability code can be ensured, thereby improving the reliability of tracing.

[0096] After obtaining the traceability code of the target medical supplies, the information during each material management operation (such as operation information and corresponding personnel information) can be obtained, so as to generate real traceability information based on the foregoing information and associate it with the traceability code, thereby realizing the traceability monitoring of the target medical supplies; among them, the generation process of the real traceability information can be but is not limited to the following steps S2 to S4.

[0097] S2. Obtain the original traceability information of the target medical supplies during each material management operation. Among them, the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include warehousing operation, application operation and logistics operation; in this embodiment, since barcodes are currently used in hospitals to manage medical supplies, this embodiment is based on the barcode to obtain the information corresponding to each material management operation; for example, when the target medical supplies are warehoused, the barcode on the target medical supplies is scanned by a scanner to obtain the corresponding operation information, and then, during the scanning, the face image of the operator is captured by a camera terminal, so as to obtain the original traceability information during warehousing.

[0098] Furthermore, the operation information includes the product information of this operation and the operation data of the product. For example, if the material management operation is a warehousing operation, the operation information includes the name of the warehoused product, quantity, warehousing warehouse number, shelf number, warehousing time, etc. Another example is that when the material management operation is an application operation, then the operation information includes the name of the applied product, quantity, the outbound warehouse number and shelf number where the applied product is located, application time, etc.; of course, the operation information corresponding to the rest of the material principle operations changes according to its specific operation and will not be elaborated one by one here.

[0099] After obtaining the original traceability information of the target medical supplies during each material management operation, face recognition can be performed on the face image in the original traceability information to associate the target medical supplies with the operator; among them, the face recognition process can be but is not limited to the following step S3.

[0100] S3. Perform face recognition processing on the face images of the operators in each piece of original traceability information to obtain the operator information when the target medical supplies are undergoing various material management operations; in this embodiment, an improved face recognition model is constructed. Among them, the improved face recognition model takes the sample face feature vectors of multiple sample faces as inputs (in this embodiment, the sample face feature vector of any sample face is obtained based on the low-frequency feature, high-frequency feature, and the any sample face), and the face recognition results of each sample face as outputs for training. Moreover, the improved face recognition model extracts features of different scales through parallel multi-layer convolutional layers, and fuses the extracted features through a feature fusion layer, and finally performs face recognition with the fused features; in this way, the recognition accuracy can be improved.

[0101] Optionally, one specific structure of the aforementioned improved face recognition model is provided below.

[0102] See Figure 2 As shown, for example, the improved face recognition model may include, but is not limited to: a first convolutional structure layer, a first max pooling layer, a second convolutional structure layer, a second max pooling layer, a third convolutional structure layer, a third max pooling layer, a fourth convolutional structure layer, a fourth max pooling layer, a fifth convolutional structure layer, a fifth max pooling layer, a fully connected layer, and an output layer connected in sequence; at the same time, the improved face recognition model may also include: a sixth convolutional structure layer, a first feature fusion layer, a sixth max pooling layer, a seventh convolutional structure layer, and a second feature fusion layer.

[0103] Specifically, the connection relationships of the aforementioned respective structure layers are as follows:

[0104] See Figure 2 As shown, the output end of the third max pooling layer is connected to the input end of the sixth convolutional structure layer, and the output ends of the fourth convolutional structure layer and the sixth convolutional structure layer are respectively connected to the input end of the first feature fusion layer for feature fusion through the first feature fusion layer.

[0105] Similarly, the output end of the first feature fusion layer is connected to the input end of the sixth max pooling layer, the output end of the sixth max pooling layer is connected to the input end of the seventh convolutional structure layer. Among them, the output end of the seventh convolutional structure layer and the output end of the fifth convolutional structure layer are connected to the input end of the second feature fusion layer, and the output end of the second feature fusion layer is connected to the input end of the fifth max pooling layer.

[0106] Further, for example, the network structures of the sixth convolutional structure layer and the seventh convolutional structure layer are the same, and both include a first convolutional layer, a second convolutional layer, a third convolutional layer, a fourth convolutional layer, and a fifth convolutional layer. Among them, for the sixth convolutional structure layer, the output ends of the third max-pooling layer are respectively connected to the input ends of the first convolutional layer, the second convolutional layer, and the third convolutional layer. Among them, the output end of the second convolutional layer is connected to the input end of the fourth convolutional layer, and the output end of the third convolutional layer is connected to the input end of the fifth convolutional layer. Moreover, the output ends of the first convolutional layer, the fourth convolutional layer, and the fifth convolutional layer are all connected to the input end of the first feature fusion layer; of course, the connection method of the seventh convolutional structure layer is the same as that of the sixth convolutional structure layer (that is, the sixth max-pooling layer is respectively connected to the input ends of the first, second, and third convolutional layers in the seventh convolutional structure layer), which will not be elaborated here.

[0107] Thus, the recognition process of the entire improved face recognition model is as follows:

[0108] The sample face feature vector first undergoes first feature extraction processing through the first convolutional structure layer to obtain first feature information; then, it undergoes feature dimensionality reduction through the first max-pooling layer to obtain second feature information; next, it undergoes second feature extraction processing through the second convolutional structure layer to obtain third feature information; then, based on the second max-pooling layer, the third feature information is subjected to feature dimensionality reduction to obtain fourth feature information; similarly, the fourth feature information is input into the third convolutional structure layer for third feature extraction processing to obtain fifth feature information; finally, the fifth feature information is input into the third max-pooling layer for feature dimensionality reduction to obtain sixth feature information; at this time, the preprocessing of the network is completed.

[0109] Then, the sixth feature information is input into each convolutional layer in the sixth convolutional structure layer for seventh feature extraction processing. Among them, the first, fourth, and fifth convolutional layers in the sixth convolutional structure layer input the output feature information into the first feature fusion layer; at the same time, the fourth convolutional structure layer also needs to perform eighth feature extraction processing on the sixth feature information output by the third max-pooling layer to obtain seventh feature information; then, the fourth convolutional structure layer inputs the seventh feature information into the first feature fusion layer, and the first feature fusion layer fuses the feature information output by the aforementioned first, fourth, and fifth convolutional layers with the seventh feature information to obtain first feature fusion information.

[0110] After the first feature fusion is completed, the first feature fusion information can be input into the sixth max pooling layer for feature dimensionality reduction to obtain the first dimensionality-reduced feature fusion information. Then, the first dimensionality-reduced feature fusion information is input into each convolutional layer in the seventh convolutional structure layer for the eighth feature extraction process. Similarly, the first, fourth, and fifth convolutional layers in the seventh convolutional structure layer also input the output feature information into the second feature fusion layer. At this time, the aforementioned seventh feature information also needs to be input into the fourth max pooling layer and the fifth convolutional structure layer for processing to obtain the ninth feature information, which is then input into the second feature fusion layer. Then, the second feature fusion layer fuses the ninth feature information with the features output by the first, fourth, and fifth convolutional layers in the aforementioned seventh convolutional structure layer to obtain the second feature fusion information.

[0111] After the above operations are completed, the second feature fusion information is input into the fifth max pooling layer for feature dimensionality reduction, and finally, the feature vector for final face classification is extracted by the fully connected layer. Finally, the softmax classification function in the output layer can be used to calculate the probabilities of each classification to which the feature vector belongs, and the category corresponding to the maximum probability is used as the recognition result.

[0112] Furthermore, for example, the first convolutional structure layer and the second convolutional structure layer may include but are not limited to two convolutional layers, and the third convolutional structure layer, the fourth convolutional structure layer, and the fifth convolutional structure layer may include but are not limited to three convolutional layers. At the same time, for example, the number of convolutional kernels of the first, second, and third convolutional layers in the aforementioned sixth convolutional structure layer and seventh convolutional structure layer is 64, and the number of convolutional kernels of the fourth and fifth convolutional layers is 128.

[0113] Thus, through the above description of the improved face recognition model, in actual use, face recognition can be performed based on this improved face recognition model. Herein, in this embodiment, any face image is taken as an example to specifically describe the face recognition process, which may include but is not limited to the following steps S31 to S34.

[0114] S31. For any face image, extract the low-frequency features and high-frequency features of the said face image. In this embodiment, for example, the high-frequency features are LBP statistical histogram features, and after performing DCT (Discrete Cosine Transform) on the said face image, the low-frequency features of the said face image can be obtained.

[0115] After obtaining the low-frequency features and high-frequency features of the said face image, the face feature vector of the said face image can be generated in combination with the original face image. The generation process of the face feature vector is as shown in the following step S32.

[0116] S32. Generate a face feature vector of the arbitrary face image based on the low-frequency feature, high-frequency feature, and the arbitrary face image; in this embodiment, the foregoing low-frequency feature, high-frequency feature, and the arbitrary face image (which can be a pixel matrix) are combined to obtain a vector sequence, which is used as the face feature vector; thus, using this face feature vector as the input of the model enables the network to extract the high-frequency local features and low-frequency local features of the face, and can also extract the global original image information; thereby, the features extracted by the network are more sufficient and more representative, so as to improve the recognition accuracy of the model.

[0117] After obtaining the face feature vector of the arbitrary face image, input it into the improved face recognition model to obtain the face recognition result, and according to the face recognition result, the corresponding operator information can be obtained; among them, the face recognition process is shown in the following steps S33 and S34.

[0118] S33. Obtain the improved face recognition model, where the improved face recognition model is trained with the sample face feature vectors of multiple sample faces as the input and the face recognition results of each sample face as the output.

[0119] S34. Input the face feature vector into the improved face recognition model to obtain the operator information when the target medical supplies are performing the specified material management operation, where the specified material management operation is the material management operation associated with the original traceability information corresponding to the arbitrary face image; in this embodiment, after obtaining the face recognition result, the employee information of the person corresponding to the face, such as the affiliated department, position, etc., can be obtained from the hospital employee database, and then the face recognition result and the employee information are combined to form the operator information.

[0120] In this way, through the foregoing steps S31 - S34, the face recognition results of each face image in each original traceability information can be obtained, so as to obtain the operator information when the target medical supplies are performing each material management operation; then, combined with the operation information, the true traceability information of the target medical supplies can be generated; among them, the generation process of the true traceability information can be but is not limited to the following step S4.

[0121] S4. Use the operator information and operation information when the target medical supplies are performing each material management operation to generate the true traceability information when the target medical supplies are performing each material management operation; in this embodiment, the form of the true traceability information can be: material management operation type (such as warehousing) + operation information + operator information; after obtaining each true traceability information, associate it with the traceability code to complete the traceability control of the target medical supplies, where the association process is shown in the following step S5.

[0122] S5. Associate each piece of real traceability information with the traceability code, and store the associated real traceability information in a database, so that after the user terminal scans the traceability code, all the real traceability information of the target medical supplies during use can be retrieved from the database to complete the traceability control of the target medical supplies.

[0123] In this embodiment, when the target medical supplies are first warehoused, the aforementioned traceability code can be generated and printed on the target medical supplies. Then, each time a material management operation is performed, corresponding real traceability information will be generated and associated with the traceability code. In this way, each time a material management operation is performed on the target medical supplies, a new piece of real traceability information will be added; based on this, the whole process traceability during its usage cycle can be realized.

[0124] Thus, through the intelligent control and traceability method for medical supplies based on artificial intelligence described in detail in the aforementioned steps S1 to S5, the present invention first generates a traceability code based on the unique product identification code of the medical supplies and the captured image. Then, when performing material management operations such as warehousing, outbound, and logistics of medical supplies, the corresponding operation information is obtained, and face recognition is performed on the corresponding operators to obtain operator information; finally, by associating the operation information and operator information of the medical supplies during each material management operation with their corresponding traceability codes, the traceability control of the medical supplies can be completed; in this way, the present invention can realize the traceability management of each operation and personnel during the entire life cycle of medical supplies, and thus is very suitable for large-scale application and promotion in the technical field of medical supplies traceability control.

[0125] As Figure 3 shown, the second aspect of this embodiment provides a hardware device for implementing the intelligent control and traceability method for medical supplies based on artificial intelligence described in the first aspect of the embodiment, including:

[0126] An acquisition unit, configured to acquire the product information and the captured image of the target medical supplies, and generate the traceability code of the target medical supplies based on the product information and the captured image, wherein the product information includes the unique product identification code.

[0127] An acquisition unit, configured to acquire the original traceability information of the target medical supplies during each material management operation, wherein the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include warehousing operations, application operations, and logistics operations.

[0128] A face recognition unit, configured to perform face recognition processing on the face images of the operators in each piece of original traceability information, so as to obtain the operator information of the target medical supplies during each material management operation.

[0129] A traceability information generation unit, configured to generate the true traceability information of the target medical supplies during each material management operation by using the operator information and operation information of the target medical supplies during each material management operation.

[0130] A traceability management unit, configured to perform an association process between each piece of true traceability information and the traceability code, and store the associated true traceability information in a database, so that after the user terminal scans the traceability code, all the true traceability information of the target medical supplies during use can be found from the database, so as to complete the traceability control of the target medical supplies.

[0131] For the working process, working details and technical effects of the device provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated here.

[0132] As Figure 4 shown, a third aspect of this embodiment provides another intelligent control and traceability device for medical supplies based on artificial intelligence. Taking the device as an electronic device as an example, it includes: a memory, a processor, and a transceiver that are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer programs and execute the intelligent control and traceability method for medical supplies based on artificial intelligence as described in the first aspect of the embodiment.

[0133] Specifically, the memory may include, but is not limited to, random access memory (RAM), read only memory (ROM), flash memory, first input first output (FIFO) and / or first in last out (FILO), etc.; specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), PLA (Programmable Logic Array). At the same time, the processor may also include a main processor and a coprocessor. The main processor is a processor for processing data in the wake state, also known as the CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state.

[0134] In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. For example, the processor may not be limited to using a microprocessor of the STM32F105 series, a reduced instruction set computer (RISC) microprocessor, an X86 architecture processor, or a processor integrated with an embedded neural-network processing unit (NPU); the transceiver may include, but is not limited to, a Wi-Fi wireless transceiver, a Bluetooth wireless transceiver, a General Packet Radio Service (GPRS) wireless transceiver, a ZigBee (a low-power local area network protocol based on the IEEE 802.15.4 standard) wireless transceiver, a 3G transceiver, a 4G transceiver, and / or a 5G transceiver, etc. In addition, the device may also include, but is not limited to, a power module, a display screen, and other necessary components.

[0135] For the working process, working details and technical effects of the electronic device provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated here.

[0136] In the fourth aspect of this embodiment, a storage medium storing instructions for the artificial intelligence-based intelligent control and traceability method of medical supplies described in the first aspect of the embodiment is provided, that is, instructions are stored on the storage medium, and when the instructions run on a computer, they execute the artificial intelligence-based intelligent control and traceability method of medical supplies described in the first aspect of the embodiment.

[0137] Among them, the storage medium refers to a carrier for storing data, and can include but is not limited to floppy disks, optical discs, hard disks, flash memories, USB flash drives, and / or memory sticks, etc. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0138] For the working process, working details, and technical effects of the storage medium provided in this embodiment, reference can be made to the first aspect of the embodiment, and details will not be repeated here.

[0139] In the fifth aspect of this embodiment, a computer program product containing instructions is provided. When the instructions run on a computer, the computer is made to execute the artificial intelligence-based intelligent control and traceability method of medical supplies described in the first aspect of the embodiment. Among them, the computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0140] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent control and traceability method for medical supplies based on artificial intelligence, characterized in that, Including: Obtain the product information of the target medical supplies and the captured images, and generate a traceability code for the target medical supplies based on the product information and the captured images, where the product information includes a unique product identification code; Obtain the original traceability information of the target medical supplies during each material management operation, where the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include warehousing operations, application operations, and logistics operations; Perform face recognition processing on the face images of the operators in each original traceability information to obtain the operator information of the target medical supplies during each material management operation; Generate the real traceability information of the target medical supplies during each material management operation by using the operator information and operation information of the target medical supplies during each material management operation; Perform an association process on each real traceability information and the traceability code, and store the associated real traceability information in a database, so that after the user terminal scans the traceability code, all the real traceability information of the target medical supplies during use can be found from the database to complete the traceability control of the target medical supplies.

2. The method according to claim 1, wherein Generating the traceability code for the target medical supplies based on the product information and the captured images includes: Perform image recognition processing on the captured images to obtain the image feature information of the captured images, where the image feature information is used to represent the classification information of the target medical supplies; Obtain a key generation function and generate a first encryption key by using the key generation function; Obtain the shooting time of the captured images, and perform encryption processing on the shooting time by using an irreversible encryption algorithm to obtain a second encryption key; Generate a third encryption key according to the first encryption key and the second encryption key; Use the third encryption key to perform encryption processing on the unique product identification code in the product information to obtain encrypted information; Generate the traceability code for the target medical supplies based on the encrypted information and the image feature information.

3. The method according to claim 2, wherein The third encryption key is a binary sequence, where using the third encryption key to perform encryption processing on the unique product identification code in the product information to obtain encrypted information includes: Divide the unique product identification code into character segments of length k, where k is a positive integer greater than 1; Convert the binary sequence into a decimal value, and cyclically shift each character segment to the right or left based on the decimal value to obtain updated character segments; Perform an exclusive OR operation on the third encryption key and the updated character segments to obtain multiple encrypted character segments; Perform a splicing process on the multiple encrypted character segments to obtain the encrypted information; Correspondingly, generating the traceability code for the target medical supplies based on the encrypted information and the image feature information includes: Generate an anti-counterfeiting pattern based on the encrypted information; Generate an initial traceability code by using the image feature information; Add the anti-counterfeiting pattern to the initial traceability code to obtain the traceability code of the target medical supplies.

4. The method according to claim 2, wherein The key generation function is as follows: In formula (1), represents the key generation function, m i represents the first random variable, p represents the number of encryption iterations, P represents the maximum number of encryption iterations, x p represents the second random variable at the p-th time, and i represents the first encryption parameter; where x p = α × x p-1 × (1 - x p-1 ) (2); In formula (2), α represents the second encryption parameter, where when p is 1, x p-1 is the initial value, and x p-1 ∈ [0, 1]; In formula (3), represents a random parameter function,[ 5. The method according to claim 1, wherein Perform face recognition processing on the face images of the operators in each original traceability information to obtain the operator information of the target medical supplies during each material management operation, including: For any face image, extract the low-frequency features and high-frequency features of the any face image; Generate a face feature vector of the any face image based on the low-frequency features, high-frequency features, and the any face image; Obtain an improved face recognition model, where the improved face recognition model is trained with the sample face feature vectors of multiple sample faces as inputs and the face recognition results of each sample face as outputs; Input the face feature vector into the improved face recognition model to obtain the operator information of the target medical supplies during the specified material management operation, where the specified material management operation is the material management operation associated with the original traceability information corresponding to the any face image.

6. The method according to claim 5, wherein The improved face recognition model includes: a first convolutional structure layer, a first max pooling layer, a second convolutional structure layer, a second max pooling layer, a third convolutional structure layer, a third max pooling layer, a fourth convolutional structure layer, a fourth max pooling layer, a fifth convolutional structure layer, a fifth max pooling layer, a fully connected layer, and an output layer connected in sequence, and the improved face recognition model further includes: a sixth convolutional structure layer, a first feature fusion layer, a sixth max pooling layer, a seventh convolutional structure layer, and a second feature fusion layer; The output end of the third max pooling layer is connected to the input end of the sixth convolutional structure layer, and the output ends of the fourth convolutional structure layer and the sixth convolutional structure layer are respectively connected to the input end of the first feature fusion layer for feature fusion through the first feature fusion layer; The output end of the first feature fusion layer is connected to the input end of the sixth max pooling layer, and the output end of the sixth max pooling layer is connected to the input end of the seventh convolutional structure layer, where the output end of the seventh convolutional structure layer and the output end of the fifth convolutional structure layer are connected to the input end of the second feature fusion layer, and the output end of the second feature fusion layer is connected to the input end of the fifth max pooling layer.

7. The method according to claim 6, wherein Both the sixth convolutional structure layer and the seventh convolutional structure layer include a first convolutional layer, a second convolutional layer, a third convolutional layer, a fourth convolutional layer, and a fifth convolutional layer, and the number of convolution kernels of the first convolutional layer, the second convolutional layer, and the third convolutional layer is 64, and the number of convolution kernels of the fourth convolutional layer and the fifth convolutional layer is 128; For the sixth convolutional structure layer, the output end of the third max pooling layer is respectively connected to the input ends of the first convolutional layer, the second convolutional layer, and the third convolutional layer. Among them, the output end of the second convolutional layer is connected to the input end of the fourth convolutional layer, the output end of the third convolutional layer is connected to the input end of the fifth convolutional layer, and the output ends of the first convolutional layer, the fourth convolutional layer, and the fifth convolutional layer are all connected to the input end of the first feature fusion layer.

8. An intelligent control and traceability device for medical supplies based on artificial intelligence, characterized in that, Comprising: An acquisition unit, configured to acquire product information and a captured image of a target medical supply, and generate a traceability code for the target medical supply based on the product information and the captured image, where the product information includes a product unique identification code; An acquisition unit, configured to acquire the original traceability information of the target medical supply during each material management operation, where the original traceability information corresponding to any material management operation includes the operation information corresponding to the any material management operation and the face image of the operator, and the material management operations include an inbound operation, a requisition operation, and a logistics operation; A face recognition unit, configured to perform face recognition processing on the face images of the operators in each piece of original traceability information to obtain the operator information of the target medical supply during each material management operation; A traceability information generation unit, configured to generate the true traceability information of the target medical supply during each material management operation by using the operator information and the operation information of the target medical supply during each material management operation; A traceability management unit, configured to perform an association process on each piece of true traceability information with the traceability code, and store the associated true traceability information in a database, so that after a user terminal scans the traceability code, all the true traceability information of the target medical supply during use can be found from the database to complete the traceability control of the target medical supply.

9. An intelligent control and traceability device for medical supplies based on artificial intelligence, characterized in that, Comprising: A memory, a processor, and a transceiver that are communicatively connected in sequence, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the artificial intelligence-based intelligent control and traceability method for medical supplies according to any one of claims 1 to 7.

10. A storage medium, characterized in that, Instructions are stored on the storage medium, and when the instructions are run on a computer, the artificial intelligence-based intelligent control and traceability method for medical supplies according to any one of claims 1 to 7 is executed.

Images