Microorganism intelligent detection system and method based on big data analysis

Abstract

The invention provides an intelligent microorganism detection system and method based on big data analysis, and relates to the field of intelligent analys.The method comprises the steps that firstly, an in-vitro to-be-detected specimen of a patient is obtained, then, drug resistance detection is conducted through a microorganism detector and a matched self-flowing type microorganism detection card to obtain a microorganism drug resistance detection data set, and then, the detection data set is analyzed; the data set is uploaded to a cloud server by means of an internet communication module, the uploaded microorganism drug resistance detection data set is subjected to data analysis at the cloud server side, a corresponding analysis result is obtained, and the analysis result is displayed. Therefore, the analysis of the drug resistance data of the microorganisms can be more accurate, and public health management can be better served.

Description

Technical Field

[0001] The present application relates to the field of intelligent analysis, and more specifically, to a microbial intelligent detection system and method based on big data analysis. Background Art

[0002] In today's medical field, diseases caused by microbial infections are still a severe challenge. With the acceleration of globalization and the increasing mobility of population, the spread of microorganisms has become more extensive. At the same time, the continuous emergence of new drug-resistant microorganisms has made clinical treatment face great difficulties. In order to meet these challenges, it is particularly important to improve the accuracy and efficiency of microbial detection.

[0003] In this regard, patent CN107545138B proposes a microbial detection method, a computer-readable storage medium and a microbial detection system based on big data. First, the drug resistance test is performed by obtaining in vitro specimens, and then the data is uploaded to the cloud server. During the analysis process, a local weighted linear regression algorithm is used to obtain the analysis results. After that, the algorithm is adjusted through the intervention module, and finally, the corrected analysis results are displayed.

[0004] The above patent uses a local weighted linear regression algorithm for data analysis. However, the interactions between different pathogens and antibiotics are complex and diverse. Since local weighted linear regression is essentially still based on the assumption of a linear model, for data sets that exhibit significant nonlinear relationships, especially complex microbial resistance data, this algorithm may not accurately capture the true patterns and complex interactions of the data, resulting in prediction bias.

[0005] Therefore, an optimized microbial intelligent detection solution based on big data analysis is expected, which can improve the understanding and prediction capabilities of complex nonlinear data relationships to adapt to diverse data distribution patterns. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a microbial intelligent detection system and method based on big data analysis.

[0007] According to one aspect of the present application, a microbial intelligent detection system based on big data analysis is provided, which includes: a detection data acquisition module, which is used to obtain a sample to be detected from a patient, and perform drug resistance detection on the microorganisms in the sample to be detected through a microbial detector and a self-flowing microbial detection card matched therewith to obtain a microbial drug resistance detection data set; a data transmission module, which is used to upload the microbial drug resistance detection data set to a cloud server through an Internet communication module; a data analysis module, which is used to perform data analysis on the microbial drug resistance detection data set on the cloud server to obtain an analysis result; and a display module, which is used to display the analysis result; wherein the data analysis module includes:

[0008] A detection data embedding unit, used for embedding and encoding each microbial drug resistance detection data in the microbial drug resistance detection data set on the cloud server to obtain a set of microbial drug resistance detection data embedding coding vectors;

[0009] A data aggregation unit, configured to perform graph structure adaptive feature significant aggregation on the set of microbial drug resistance detection data embedding coding vectors on the cloud server to obtain a microbial drug resistance detection data set significant aggregation feature vector;

[0010] An attention guiding fusion unit is used to extract a microorganism category embedding coding vector from the set of microorganism drug resistance detection data embedding coding vectors on the cloud server, and perform attention interaction response on the microorganism category embedding coding vector and the significant aggregation feature vector of the microorganism drug resistance detection data set to obtain a microorganism category guided drug resistance detection data representation vector;

[0011] The analysis result generating unit is used to obtain the analysis result on the cloud server by guiding the drug resistance detection data characterization vector based on the microorganism category.

[0012] According to another aspect of the present application, a microbial intelligent detection method based on big data analysis is provided, which is used to execute the microbial intelligent detection system based on big data analysis as described above, and comprises:

[0013] On the cloud server, each microbial drug resistance detection data in the microbial drug resistance detection data set is embedded and encoded to obtain a set of microbial drug resistance detection data embedded encoding vectors;

[0014] On the cloud server, performing graph structure adaptive feature significant aggregation on the set of microbial drug resistance detection data embedding coding vectors to obtain a microbial drug resistance detection data set significant aggregation feature vector;

[0015] On the cloud server, extracting a microorganism category embedding coding vector from the set of microorganism drug resistance detection data embedding coding vectors, and performing attention interaction response on the microorganism category embedding coding vector and the significant aggregation feature vector of the microorganism drug resistance detection data set to obtain a microorganism category guided drug resistance detection data representation vector;

[0016] On the cloud server, the drug resistance detection data characterization vector is guided based on the microorganism category to obtain the analysis result.

[0017] Compared with the prior art, the microbial intelligent detection system and method based on big data analysis provided by the present application first obtains the patient's in vitro specimen to be tested, then uses the microbial detector and the matching self-flowing microbial detection card to perform drug resistance detection to obtain a microbial resistance detection data set, and then uploads the data set to the cloud server with the help of the Internet communication module, and performs data analysis on the uploaded microbial resistance detection data set on the cloud server to obtain the corresponding analysis results, and displays the analysis results. In this way, the analysis of microbial resistance data can be more accurate, and thus better serve public health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0019] Figure 1 It is a system block diagram of a microbial intelligent detection system based on big data analysis according to an embodiment of the present application.

[0020] Figure 2 It is a block diagram of a data analysis module in a microbial intelligent detection system based on big data analysis according to an embodiment of the present application.

[0021] Figure 3 It is a block diagram of a data aggregation unit in a microbial intelligent detection system based on big data analysis according to an embodiment of the present application.

[0022] Figure 4 The present invention is a flowchart of a method for intelligent detection of microorganisms based on big data analysis according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0024] In the current medical field, microbial infections still pose a severe challenge. The frequent population movement caused by globalization has exacerbated the spread of microorganisms, and the continuous emergence of new drug-resistant microorganisms has brought huge difficulties to clinical treatment. In order to effectively respond to these challenges, it is crucial to improve the accuracy and efficiency of microbial detection.

[0025] Patent CN107545138B proposes a microbial detection method, a computer-readable storage medium, and a microbial detection system based on big data. It first collects in vitro specimens for drug resistance detection, and then uploads the data to a cloud server. In the data analysis stage, a local weighted linear regression algorithm is used to obtain the analysis results. Subsequently, the algorithm is adjusted through an intervention module, and finally the corrected analysis results are displayed.

[0026] However, it should be noted that due to the complex and diverse interactions between different pathogens and antibiotics, and the fact that the local weighted linear regression algorithm is based on the linear model assumption, for data sets that exhibit significant nonlinear relationships, especially complex microbial resistance data, the algorithm may not be able to accurately capture the true patterns and complex interactions of the data, thus leading to deviations in the prediction results.

[0027] Based on this, this application proposes a microbial intelligent detection system based on big data analysis. Figure 1 : is a system block diagram of a microbial intelligent detection system based on big data analysis according to an embodiment of the present application. Figure 1 As shown, in the microbial intelligent detection system 100 based on big data analysis, it includes: a detection data acquisition module 110, which is used to obtain a sample to be detected from a patient, and perform drug resistance detection on the microorganisms in the sample to be detected through a microbial detector and a self-flowing microbial detection card matched therewith to obtain a microbial resistance detection data set; a data transmission module 120, which is used to upload the microbial resistance detection data set to a cloud server through an Internet communication module; a data analysis module 130, which is used to perform data analysis on the microbial resistance detection data set on the cloud server to obtain an analysis result; and a display module 140, which is used to display the analysis result.

[0028] In an embodiment of the present application, the detection data acquisition module 110 is used to obtain an in vitro specimen to be detected from a patient, and to perform drug resistance detection on the microorganisms in the specimen to be detected through a microbial detector and a self-flowing microbial detection card matched therewith to obtain a microbial resistance detection data set. Specifically, in an embodiment of the present application, the microbial resistance detection data includes the gender and age of the specimen patient, the type of microorganism, the antibiotic resistance, the specimen detection time, and the region where the specimen is detected. It should be understood that in order to conduct an effective microbial resistance analysis, accurate and detailed original data must first be obtained. This includes collecting in vitro specimens from the patient's body and using professional equipment (microbial detector and self-flowing microbial detection card) to detect them. By using a self-flowing microbial detection card in conjunction with a microbial detector, the consistency and accuracy of specimen processing can be ensured, human errors can be reduced, and the reliability of microbial resistance detection data results can be improved. The microbial resistance detection data obtained here through resistance testing includes multiple information such as the gender, age, microbial species, antibiotic resistance, specimen testing time, and the region where the specimen was tested. It can reflect the characteristics and distribution of microorganisms from multiple dimensions, such as the differences in microbial infections among people of different genders and ages, and the prevalence trend of microbial resistance in specific regions. This can provide a rich data basis for subsequent data analysis, and based on the analysis results obtained from data analysis, targeted public health measures can be taken, thereby enabling better public health management.

[0029] The specific implementation process of obtaining the patient's in vitro specimens to be tested and conducting microbial resistance testing to obtain a microbial resistance test data set involves multiple links, which requires professionals to strictly follow the specifications and cooperate with effective quality control measures to ensure the accuracy and reliability of the test results. The following is a detailed description of the implementation process:

[0030] The first is the specimen collection process. This process must be performed by professionally trained medical staff, who need to strictly follow aseptic operation specifications to prevent the specimens from being contaminated by the outside world, thereby affecting the test results. When collecting different types of specimens, the corresponding correct methods should be adopted according to the specific situation. For example, when collecting blood specimens, medical staff use disposable sterile syringes to extract an appropriate amount of blood from the patient's veins; for sputum specimens, patients should be instructed to cough up sputum correctly to ensure that the specimens come from the lower respiratory tract, because miscellaneous bacteria from the mouth may interfere with the test results; when collecting urine specimens, patients need to clean the vulva first, and then collect midstream urine, which can reduce the contamination of urine specimens by bacteria in the urethral orifice. The specimen types are rich and varied. In addition to the common blood, sputum, and urine, they also cover feces, cerebrospinal fluid, wound secretions, and genital secretions. The specific choice of specimen collection depends on the initial judgment of the patient's infection. For example, when intestinal infection is suspected, fecal specimens are the key; if central nervous system infection is suspected, cerebrospinal fluid specimens are indispensable.

[0031] After completing the specimen collection, the test preparation stage begins. For the microbial detector, to ensure that it is in normal working condition, a series of preparations need to be carried out in advance, including calibration, cleaning and debugging of the instrument. The calibration work requires the use of standard substances to ensure the normal operation of various functions such as unobstructed detection light path and stable temperature control system. At the same time, it is necessary to select a self-flowing microbial detection card that matches the microbial detector, and carefully check whether its packaging is intact and whether the validity period is within the range. Before use, place the test card at room temperature for a period of time to balance it to a suitable working temperature. This will help the reagents and reaction system in the test card to be in the best condition, thereby improving the accuracy of the test.

[0032] Next is the core step of drug resistance detection. First, the collected in vitro specimens to be tested are properly processed. For example, blood specimens may need to be centrifuged to separate serum or plasma for subsequent testing; sputum specimens need to be liquefied to make them easier to detect. The processed specimens are accurately loaded into the corresponding sample wells of the test card according to the operating instructions of the self-flowing microbial test card. Subsequently, the self-flowing microbial test card loaded with the specimen is placed in the microbial detector, and the appropriate incubation temperature and time are set according to the characteristics of the test card. During the incubation period, complex biochemical reactions occur in the test card, and microorganisms interact with specific reagents. The resistance to antibiotics is judged by monitoring the growth of microorganisms, changes in metabolites and other indicators. After the incubation is completed, the microbial detector automatically reads the test results on the test card, which include important information such as the type of microorganism and antibiotic resistance. At the same time, relevant data such as the specimen detection time and the region where the specimen is detected need to be recorded. This information together constitutes the microbial resistance detection data set.

[0033] In order to ensure the accuracy and reliability of the entire testing process, quality control measures need to be taken. In terms of specimen collection quality control, regular training of medical staff is the key. Through training, they can strengthen their understanding of the correct methods and importance of specimen collection and improve the accuracy and standardization of collection operations. At the same time, a supervision mechanism should be established to regularly check the specimen collection process to ensure that every operation link meets the standards. Post-collection specimen quality assessment is also essential, such as checking whether blood specimens have hemolysis or coagulation, whether sputum specimens are qualified, etc. Unqualified specimens should be notified in time for recollection.

[0034] In the quality control of the testing process, the calibration and maintenance of the microbial detector must be carried out regularly and strictly in accordance with the manufacturer's recommendations to ensure that the instrument test results are always accurate and reliable. Each batch of self-flowing microbial test cards must be subject to quality sampling, and known positive and negative control samples must be used for testing to verify their sensitivity and specificity, and the batch information of the test cards must be recorded for traceability. In-house quality control products must also be added to daily testing, and quality control charts must be drawn regularly to monitor the accuracy and repeatability of test results. Once the quality control results exceed the expected range, the cause must be immediately found and corrective measures must be taken, such as recalibrating the instrument and replacing the test reagents.

[0035] Data recording and management quality control should also not be neglected. Detailed data recording specifications should be formulated, requiring recorders to accurately and completely record all information in the microbial resistance detection data set, including specimen information, test results, test time, etc. The records must be clear and traceable to avoid alterations and omissions. A data review position should be established to strictly review the recorded data to ensure the accuracy and consistency of the data. In addition, data should be backed up regularly, using a variety of methods such as local backup and off-site backup to prevent data loss, and backup data should be regularly restored and tested to ensure its availability.

[0036] Through the above rigorous implementation process and comprehensive quality control measures, high-quality microbial resistance detection data sets can be obtained, which can provide a reliable data basis for subsequent data analysis.

[0037] In an embodiment of the present application, the data transmission module 120 is used to upload the microbial resistance detection data set to the cloud server through the Internet communication module. It should be understood that local devices are limited in computing power and storage capacity, and cannot efficiently process large-scale data sets or execute complex algorithms. Cloud servers usually have powerful computing resources and massive storage space, which can support the efficient processing of large amounts of microbial resistance detection data. Specifically, cloud servers have stronger data processing capabilities, can complete data analysis more quickly, and can apply more advanced statistical methods and artificial intelligence technologies to mine deep information in microbial resistance detection data sets. Moreover, by uploading data sets to cloud servers, it is possible to centrally manage and uniformly process data from different time points, ensure data consistency and integrity, and avoid data loss or inconsistency problems that may be caused by decentralized management.

[0038] Uploading the microbial resistance detection data set to the cloud server through the Internet communication module is a key process to achieve efficient data analysis. This process involves multiple considerations and operations to ensure the security and rapid transmission of data and effective management on the cloud server side. The following is a detailed description of its implementation process:

[0039] The first prerequisite for data transmission is to ensure a stable connection between the local device and the Internet communication module. In medical institutions or testing centers, it is necessary to equip them with high-performance network equipment, such as routers, switches, etc., to ensure a stable wired network connection; for mobile testing equipment or wireless networks that may be used in special scenarios, their signal strength, stability and security must be ensured.

[0040] After the connection is established, the microbial resistance detection data set must be preprocessed to optimize data transmission efficiency. This includes data compression and format conversion. Data compression can use efficient compression algorithms, such as lossless compression algorithms (such as LZW algorithm, Deflate algorithm, etc.), to reduce the size of data files without losing data accuracy, thereby speeding up transmission. At the same time, according to the data reception requirements of the cloud server, the data set is converted into a suitable format, such as the common CSV (comma separated value) format or JSON (JavaScript Object Notation) format, to ensure that the data can be successfully recognized and processed by the cloud server.

[0041] Choosing the right Internet communication protocol is crucial for data uploading. Commonly used protocols include HTTP (Hypertext Transfer Protocol) or HTTPS (Secure Hypertext Transfer Protocol). HTTPS adds an encryption layer on the basis of HTTP, which can effectively ensure the security of data transmission and prevent data from being stolen or tampered with during network transmission. It is especially suitable for microbial resistance detection data sets containing sensitive patient information.

[0042] During the data upload process, in order to ensure the integrity and accuracy of the data, a data verification mechanism needs to be adopted. For example, the hash value of the data set is calculated (such as hash algorithms such as MD5 and SHA-1), and after the data is uploaded, the hash value of the received data is calculated again on the cloud server side, and the two are compared. If the hash values ​​are consistent, it means that there is no error or tampering in the data transmission process; if they are inconsistent, it is necessary to re-upload the data or find the problem in the transmission process.

[0043] The cloud server should be equipped with high-performance storage space and provide fast data writing speed. At the same time, a complete data management system should be established to classify and store the uploaded microbial resistance detection data set, such as by detection time, detection area, microbial species, etc., to facilitate subsequent data retrieval and analysis.

[0044] In order to cope with possible data transmission interruptions or errors, the system should have a breakpoint resume function. When a network failure or other unexpected event causes data transmission interruption during data upload, the system can record the location of the transmitted data and continue uploading data from the breakpoint after the network is restored, without having to retransmit the entire data set, saving time and network resources.

[0045] Through the above series of rigorous implementation processes, the microbial resistance detection data set can be safely and efficiently uploaded to the cloud server through the Internet communication module.

[0046] In the embodiment of the present application, the data analysis module 130 is used to perform data analysis on the microbial resistance detection data set on the cloud server to obtain analysis results.

[0047] Correspondingly, in the data analysis module, the technical concept of the present application is to use big data-based data analysis and coding technology to embed and encode each microbial resistance detection data in the microbial resistance detection data set in the cloud server, then perform graph structure adaptive aggregation on the embedded features of each microbial resistance detection data, and then extract the features corresponding to the microbial category from the embedded features of each microbial resistance detection data, so as to realize the intelligent generation of analysis results based on the attention category guided representation between the microbial category embedded features and the significant aggregation features of the microbial resistance detection data set. In this way, the subtle differences and complex patterns in the data can be better captured, and the adaptability and generalization ability of the model to different data distributions are enhanced, which is especially suitable for complex nonlinear data sets and can provide more accurate analysis results compared with simple linear models.

[0048] Specifically, Figure 2 FIG. 1 is a block diagram of a data analysis module in a microbial intelligent detection system based on big data analysis according to an embodiment of the present application. Figure 2 As shown, the data analysis module 130 includes: a detection data embedding unit 131, which is used to embed and encode each microbial resistance detection data in the microbial resistance detection data set on the cloud server to obtain a set of microbial resistance detection data embedded coding vectors; a data aggregation unit 132, which is used to perform graph structure adaptive feature significant aggregation on the set of microbial resistance detection data embedded coding vectors on the cloud server to obtain a microbial resistance detection data set significant aggregation feature vector; an attention guided fusion unit 133, which is used to extract a microbial category embedded coding vector from the set of microbial resistance detection data embedded coding vectors on the cloud server, and perform attention interactive response on the microbial category embedded coding vector and the microbial resistance detection data set significant aggregation feature vector to obtain a microbial category guided resistance detection data representation vector; an analysis result generating unit 134, which is used to obtain the analysis result on the cloud server based on the microbial category guided resistance detection data representation vector.

[0049] In an embodiment of the present application, the detection data embedding unit 131 is used to embed and encode each microbial resistance detection data in the microbial resistance detection data set in the cloud server to obtain a set of microbial resistance detection data embedded coding vectors. It should be understood that the microbial resistance detection data covers many aspects of information, and these information belong to these data of different formats and different properties. Therefore, in order to be able to convert these information into a unified representation, in the technical solution of the present application, each microbial resistance detection data in the microbial resistance detection data set is embedded and encoded to integrate various types of attribute information into a unified numerical vector to obtain a set of microbial resistance detection data embedded coding vectors. In this way, it can be effectively analyzed within the same framework, and better to mine and characterize potential key information and hidden patterns. In particular, in a specific embodiment of the present application, each microbial resistance detection data in the microbial resistance detection data set can be embedded and encoded by using word embedding technology (such as Word2Vec) to construct a vocabulary and training model in a suitable manner, and effectively convert these attribute information into microbial resistance detection data embedded coding vectors.

[0050] In an embodiment of the present application, the data aggregation unit 132 is used to perform graph structure adaptive feature significant aggregation on the set of microbial resistance detection data embedded coding vectors in the cloud server to obtain a significantly aggregated feature vector of the microbial resistance detection data set. It should be understood that although the microbial resistance detection data has a unified vector representation after embedding coding, there are still complex internal associations between the data represented by each vector. For example, the age and gender of the patient may be associated with the drug resistance of the microorganism, and the drug resistance of microorganisms in different regions may also affect each other. Therefore, in order to effectively capture the long-distance, nonlinear correlation between different vectors (i.e., different microbial resistance detection data), so as to better understand the behavior within the microbial community, the present application models these vectors in the form of a graph by performing graph structure adaptive feature significant aggregation on the set of microbial resistance detection data embedded coding vectors, and by simulating the wandering process in the graph structure, these complex associations are excavated, and at the same time, the inherent laws of the data are more accurately reflected in the process of aggregating features, so as to obtain a significantly aggregated feature vector of the microbial resistance detection data set. In this way, the core content of multiple dimensions of information in the microbial resistance detection data can be comprehensively reflected, and it can be shown which combinations of factors have a more significant impact on the microbial resistance situation, which will help to grasp the overall characteristics and key influencing factors of microbial resistance more clearly in the future.

[0051] Specifically, Figure 3 FIG. 1 is a block diagram of a data aggregation unit in a microbial intelligent detection system based on big data analysis according to an embodiment of the present application. Figure 3As shown, the data aggregation unit 132 includes: a microbial drug resistance detection data mapping subunit 1321, which is used to map each microbial drug resistance detection data embedded coding vector in the set of microbial drug resistance detection data embedded coding vectors to the hyperbolic space to obtain a set of microbial drug resistance detection data embedded coding vectors after hyperbolic space mapping; a microbial drug resistance detection data topology calculation subunit 1322, which is used to obtain a microbial drug resistance detection data graph walk topology feature matrix based on the set of microbial drug resistance detection data embedded coding vectors after hyperbolic space mapping; a microbial drug resistance detection data weight calculation subunit 1323, which is used to calculate the microbial drug resistance detection data graph walk topology feature matrix based on the set of microbial drug resistance detection data embedded coding vectors after hyperbolic space mapping. A set of drug resistance detection data embedded coding vectors and the microbial resistance detection data graph walk topology feature matrix are used to calculate the autocorrelation gated significant confidence weight factor of each microbial resistance detection data embedded coding vector in the set of microbial resistance detection data embedded coding vectors to obtain a set of microbial resistance detection data autocorrelation gated significant confidence weight factors; a microbial resistance detection data aggregate feature generation subunit 1324 is used to calculate the position-weighted sum of the set of microbial resistance detection data embedded coding vectors based on the set of microbial resistance detection data autocorrelation gated significant confidence weight factors to obtain a significant aggregate feature vector of the microbial resistance detection data set.

[0052] Specifically, first, each microbial drug resistance detection data embedding coding vector in the set of microbial drug resistance detection data embedding coding vectors is mapped to the hyperbolic space to obtain a set of microbial drug resistance detection data embedding coding vectors after hyperbolic space mapping. The above process can be expressed as:

[0053]

[0054] Wherein, V is the set of embedding coding vectors of the microbial drug resistance detection data, v 1 、v 2 、v i 、v j and v n are the first, second, i-th, j-th and n-th microbial resistance detection data embedding coding vectors in the set of microbial resistance detection data embedding coding vectors, respectively, W 1 and W 2 are the first weight matrix in hyperbolic space and the second weight matrix in hyperbolic space, respectively, h i It is the i-th embedding coding vector of microorganism resistance detection data after hyperbolic space mapping in the set of embedding coding vectors of microorganism resistance detection data after hyperbolic space mapping.

[0055] It should be understood that microbial resistance detection data often have a complex hierarchical structure. For example, there may be a relationship between different microbial species, forming a classification hierarchy. Antibiotic resistance may also be graded, such as mild resistance, moderate resistance, and severe resistance. By mapping the embedding coding vectors of each microbial resistance detection data to a hyperbolic space, the characteristics of hyperbolic geometry can be used to better maintain the hierarchical structure of these data, while also helping to enhance the relative distance representation capability between embedded coding vectors. In hyperbolic space, similar microbial resistance detection data embedding coding vectors are closer. This means that for data points with similar resistance characteristics, from similar patient groups (such as the same age, gender, region), or similar microbial species, they will be closely focused together, while those microbial resistance detection data embedding coding vectors that are farther away can be effectively separated.

[0056] Specifically, in the embodiment of the present application, the microbial resistance detection data topology calculation subunit is used to calculate the Poincare distance between any two microbial resistance detection data embedded coding vectors after hyperbolic space mapping in the set of microbial resistance detection data embedded coding vectors after hyperbolic space mapping to obtain the microbial resistance detection data graph walk topology matrix. The process can be expressed as:

[0057]

[0058] Among them, h i and h j are respectively the i-th and j-th embedded coding vectors of microbial drug resistance detection data after hyperbolic space mapping in the set of embedded coding vectors of microbial drug resistance detection data after hyperbolic space mapping, ||·|| 2 is the square of the vector's modulus, arccosh is the inverse hyperbolic cosine function, d P (h i ,h j ) is used to calculate h i and h j The Poincare distance between ij is the eigenvalue of the (i, j)th position in the walk topology matrix of the microbial resistance detection data graph;

[0059] The walk topology matrix of the microbial drug resistance detection data graph is subjected to dilated convolution encoding to obtain the walk topology feature matrix of the microbial drug resistance detection data graph. The process can be expressed as:

[0060] D t =DilatedConv(D)

[0061] Where D is the walk topology matrix of the microbial resistance detection data graph, DilatedConv is the dilated convolutional code, and D tIt is the topological feature matrix of the microbial resistance detection data graph.

[0062] It should be understood that the Poincare distance measures the actual distance between two points in the hyperbolic space. By calculating the Poincare distance between any two embedded coding vectors of the microbial resistance detection data after hyperbolic space mapping in the set of embedded coding vectors of the microbial resistance detection data after hyperbolic space mapping, it can help identify the topological relationship between the embedded coding vectors after different hyperbolic space mappings, and on this basis, construct a microbial resistance detection data graph walk topology matrix. The constructed graph walk topology matrix not only reflects the direct connection between the microbial resistance detection data points, but also implies higher-order structural information through distance calculation. Specifically, in the microbial resistance data set, there may be multi-level structures, such as the classification hierarchy between microbial species, the clustering structure of drug resistance in different regions, etc. The graph walk topology matrix can capture these multi-level structural information and present it in the form of a matrix, which can provide a more comprehensive data relationship description for subsequent analysis. Subsequently, the microbial resistance detection data graph walk topology matrix is ​​subjected to void convolution coding to obtain the microbial resistance detection data graph walk topology feature matrix. Dilated convolution can expand the receptive field without increasing the number of parameters, which is very useful for capturing long-range dependencies. When applied to the walk topology matrix of microbial resistance detection data graph, dilated convolution can simultaneously capture the close association between microbial resistance data points in a small local range and the overall trend between data points in a larger range, which can provide richer information for a comprehensive analysis of microbial resistance.

[0063] Specifically, in an embodiment of the present application, the microbial resistance detection data weight calculation subunit is used to: perform global context walk of the microbial resistance detection data graph walk topological feature matrix and the set of microbial resistance detection data embedded coding vectors after hyperbolic space mapping to obtain a set of microbial resistance detection data context semantic enhancement feature vectors. The process can be expressed as:

[0064]

[0065] Among them, h i is the i-th embedded coding vector of the microbial drug resistance detection data after hyperbolic space mapping in the set of embedded coding vectors of the microbial drug resistance detection data after hyperbolic space mapping, D t is the topological feature matrix of the microbial resistance detection data graph, GCN is the graph convolutional coding, s i is the i-th microbial resistance detection data context semantics enhanced feature vector in the set of microbial resistance detection data context semantics enhanced feature vectors, and n is the number of vectors in the set of microbial resistance detection data embedded coding vectors;

[0066] The autocorrelation gating calculation is performed on each corresponding group of microbial resistance detection data context semantics enhancement feature vectors and microbial resistance detection data embedding coding vectors in the set of the microbial resistance detection data context semantics enhancement feature vectors and the set of the microbial resistance detection data embedding coding vectors to obtain a set of microbial resistance detection data autocorrelation gating significant confidence factors. The process can be expressed as:

[0067]

[0068] Among them, s i is the i-th microbial resistance detection data context semantic enhancement feature vector in the set of microbial resistance detection data context semantic enhancement feature vectors, v i is the i-th microbial resistance detection data embedding coding vector in the set of microbial resistance detection data embedding coding vectors, softmax is a normalization function, To subtract by position point, c i is the i-th microbial resistance detection data autocorrelation gated significant confidence factor in the set of microbial resistance detection data autocorrelation gated significant confidence factors, and n is the number of vectors in the set of microbial resistance detection data embedded coding vectors;

[0069] The set of the autocorrelation gated significant confidence factors of the microbial resistance detection data is normalized by the Softmax function to obtain the set of the autocorrelation gated significant confidence weight factors of the microbial resistance detection data. The process can be expressed as:

[0070]

[0071] Among them, c i and c j are the i-th and j-th microbial resistance detection data autocorrelation gated significant confidence factors in the set of microbial resistance detection data autocorrelation gated significant confidence factors, exp represents the exponential function value with the natural constant e as the base, m is the number of factors in the set of microbial resistance detection data autocorrelation gated significant confidence factors, w i It is the i-th microbial resistance detection data autocorrelation gated significant confidence weight factor in the set of microbial resistance detection data autocorrelation gated significant confidence weight factors.

[0072] It should be understood that by performing global contextual walks of graph convolution on the microbial resistance detection data graph walk topological feature matrix and the set of microbial resistance detection data embedded coding vectors after hyperbolic space mapping, it is possible to break through local limitations and integrate information within the entire graph range. This makes the final set of microbial resistance detection data context semantics enhanced feature vectors contain more comprehensive information, can grasp the microbial resistance related feature information from a macro perspective, and thus can provide a broader perspective for accurate subsequent analysis. Next, it is necessary to perform autocorrelation gating calculations on each set of microbial resistance detection data context semantics enhanced feature vectors and microbial resistance detection data embedded coding vectors in the set of microbial resistance detection data context semantics enhanced feature vectors and the set of microbial resistance detection data embedded coding vectors. In this way, the consistency and correlation between the microbial resistance detection data context semantics enhanced feature vectors and the microbial resistance detection data embedded coding vectors can be measured, so that the model can focus on those features that are more critical to the final analysis results, while suppressing irrelevant features. Then, the set of microbial resistance detection data autocorrelation gated significant confidence factors is normalized by the Softmax function. The Softmax function ensures that the sum of the weight factors is 1, and gives each embedded coding vector a weight that reflects the contribution of the embedded coding vector to the significant aggregation feature vector of the final microbial resistance detection data set. That is, by normalizing the Softmax function, the importance of features is accurately quantified, which helps the model focus more on in-depth analysis of key data.

[0073] Finally, based on the set of autocorrelation gated significant confidence weight factors of the microbial resistance detection data, the position-weighted sum of the set of embedded coding vectors of the microbial resistance detection data is calculated to obtain the significant aggregation feature vector of the microbial resistance detection data set. The above process can be expressed as:

[0074]

[0075] Among them, v i is the i-th microbial drug resistance detection data embedding coding vector in the set of microbial drug resistance detection data embedding coding vectors, w i is the i-th microbial resistance detection data autocorrelation gated significant confidence weight factor in the set of microbial resistance detection data autocorrelation gated significant confidence weight factors, n is the number of vectors in the set of microbial resistance detection data embedded coding vectors, ⊙ is the position point multiplication, v f It is a significant aggregation feature vector of the microbial resistance detection data set.

[0076] It should be understood that the process of calculating the weighted sum by position comprehensively considers the characteristics of the entire microbial resistance detection data set. Each data point (embedded coding vector) contributes its own information in the weighted sum, and the role of the weight factor emphasizes important features, which makes the final generated microbial resistance detection data set significantly aggregated feature vector able to well characterize the key information in the original microbial resistance detection data set.

[0077] In an embodiment of the present application, the attention-guided fusion unit 133 is used to extract the microbial category embedded coding vector from the set of the microbial resistance detection data embedded coding vectors in the cloud server, and perform attention interaction response on the microbial category embedded coding vector and the significantly aggregated feature vector of the microbial resistance detection data set to obtain the microbial category-guided resistance detection data representation vector. Specifically, in an embodiment of the present application, the attention-guided fusion unit is used to: extract the microbial resistance detection data embedded coding vector corresponding to the microbial species from the set of the microbial resistance detection data embedded coding vectors as the microbial category embedded coding vector; pass the microbial category embedded coding vector and the significantly aggregated feature vector of the microbial resistance detection data set through a microbial category-guided infection response expressor based on an interactive response attention network to obtain the microbial category-guided resistance detection data representation vector.

[0078] It should be understood that the microbial species plays a key guiding role in the entire microbial resistance detection and analysis process. Different microorganisms have very different resistance mechanisms, transmission patterns, and responses to treatment. Based on this, when comprehensively analyzing multiple factors, the microbial category is used as the core guide to consider how it and other factors synergistically affect the resistance detection results. In the technical solution of the present application, the microbial category embedding coding vector and the microbial resistance detection data set significant aggregation feature vector are passed through the microbial category guided infection response expressor based on the interactive response attention network to obtain the microbial category guided resistance detection data representation vector. In this way, when integrating various types of information, the dominant position of the microbial category can be highlighted, so that the final representation vector can more accurately reflect the resistance of different microbial species under different conditions, and avoid ignoring the influence of the important dimension of microbial species when fusing multi-source information. In particular, in one embodiment of the present application, the microbial category embedding coding vector and the microbial resistance detection data set significant aggregation feature vector are first linearly changed to obtain the query vector, value vector and key vector corresponding to the microbial category and the query vector, value vector and key vector corresponding to the microbial resistance detection data. Next, the query vector corresponding to the microbial category and the value vector and key vector corresponding to the microbial resistance detection data are Transformer encoded. At the same time, the query vector corresponding to the microbial resistance detection data and the value vector and key vector corresponding to the microbial category are Transformer encoded to dynamically pay attention to the correlation and importance between different vectors and capture the interactive relationship between the two different sources of data. Finally, the two interactive features are fused to form a unified representation vector that can fully reflect the resistance detection situation guided by the microbial category.

[0079] In an embodiment of the present application, the analysis result generating unit 134 is used to obtain the analysis result based on the microorganism category-guided drug resistance detection data characterization vector in the cloud server. Specifically, in an embodiment of the present application, the analysis result generating unit is used to: pass the microorganism category-guided drug resistance detection data characterization vector through a microorganism infection situation generator based on AIGC to obtain the analysis result. More specifically, in an embodiment of the present application, the analysis result includes the outbreak of microorganisms within a preset time, the distribution of people susceptible to microorganism infection, the geographical distribution of microorganisms susceptible to infection, and the resistance of antibiotics to a certain disease. It should be understood that the microorganism category-guided drug resistance detection data characterization vector is difficult for non-professionals to understand. In order to convert the information in the microorganism category-guided drug resistance detection data characterization vector into easy-to-understand content, it is necessary to input the microorganism category-guided drug resistance detection data characterization vector into a microorganism infection situation generator based on AIGC. Here, AIGC (artificial intelligence generated content) refers to a technology that automatically generates content using artificial intelligence technology. A microorganism infection situation generator refers to a model established by learning a large amount of microorganism resistance detection data and related analysis result data based on an artificial intelligence algorithm. The microbial infection situation generator can guide the drug resistance detection data representation vector based on the input microbial category, and generate microbial infection situation analysis results that conform to the data distribution law and actual situation. Specifically, after the microbial category-guided drug resistance detection data representation vector is input into the generator, it is first feature mapped through the neural network layer to convert the information in the vector into a form suitable for model processing. Then, the model analyzes the input features based on the relationship pattern between the pre-learned input vector and the output result, and finally obtains analysis results including the outbreak of microorganisms within a preset time, the distribution of people susceptible to microbial infection, the geographical distribution of microbial infection, and the resistance of antibiotics to a certain disease.

[0080] Here, when the microbial category embedding coding vector and the microbial resistance detection data set significant aggregation feature vector represent the single-sample microbial resistance detection data embedding coding features and the graph walk autocorrelation gated sequence aggregation features of the full-sample microbial resistance detection data, respectively, when performing microbial category-guided infection response expression based on the interactive response attention network, the interactive response attention misalignment of the embedded coding feature distribution pattern in the local-global sample domain will also cause the category-guided response expression richness of the microbial category-guided resistance detection data representation vector, resulting in repetition of unstructured probability mapping, affecting the accuracy of the generated results.

[0081] Therefore, it is preferred that when the microorganism category-guided drug resistance detection data characterization vector is passed through the AIGC-based microorganism infection situation generator to obtain the analysis result, the microorganism category-guided drug resistance detection data characterization vector is optimized. Specifically, the optimization includes the following steps:

[0082] Determine the number of non-zero eigenvalues ​​in the microbial category-guided drug resistance detection data characterization vector, and subtract the number of non-zero eigenvalues ​​from the length of the microbial category-guided drug resistance detection data characterization vector to obtain the zero-dimensional value n of the microbial category-guided drug resistance detection data. 1 ;

[0083] Calculate the zero-dimensional value of the microbial category-guided drug resistance detection data minus one (n 1 -1) are multiplied and divided by the zero-dimensional value n of the microbial category-guided drug resistance detection data 1 The first microbial category obtained after the drug resistance detection data field fitting value (n 1 -1)×n 1 and the second microbial category guided resistance detection data field fitting value (n 1 -1) / n 1 ;

[0084] Calculate the square root of the sum of squares of all eigenvalues ​​of the microbial category-guided drug resistance detection data characterization vector to obtain the microbial category-guided drug resistance detection data pattern characterization value

[0085] The zero differentiable value (n 1 -1) as an index to calculate each eigenvalue v of the characterization vector of the microbial category-guided drug resistance detection data m The power function And multiply it by the first microbial category to guide the resistance detection data field fitting value (n 1 -1)×n 1 To obtain the local representation value v of the microbial category-guided drug resistance detection data 1m ;

[0086] Calculate each eigenvalue v of the characterization vector of the microbial category-guided drug resistance detection data i The second microbial category guided resistance detection data field fitting value (n 1 -1) / n 1 and the microbial categories to guide the characterization of resistance detection data patterns The product of the microbial category guides the drug resistance detection data center pointing value v 2m ;

[0087] The microbial category guides the local representation value v of the drug resistance detection data 1m The microbial category guides the data center for drug resistance detection and points to the value v 2m The weighted difference w 1 v 1m -w 2 v 2m Composition-optimized microbial class-guided characterization vectors for drug resistance detection data.

[0088] That is, based on the vector field fitting law of high-dimensional manifolds, the isolated zeros in the microorganism category-guided drug resistance detection data characterization vector are used as the differentiable dimension of the high-dimensional manifold, so as to fix the normal direction pointing to the distribution center near the local coordinates represented by the eigenvalues ​​of the microorganism category-guided drug resistance detection data characterization vector, thereby avoiding invalid repetition in the probability mapping process caused by the unstructured characteristics of the distribution of the microorganism category-guided drug resistance detection data characterization vector with rich expression by effectively aligning the characteristic distribution pattern of the microorganism category-guided drug resistance detection data characterization vector to the probability density differentiable field, thereby improving the accuracy of the generated results obtained by the microorganism category-guided drug resistance detection data generator based on AIGC.

[0089] In summary, the data analysis module 130 is clearly explained, which uses big data-based data analysis and coding technology on the cloud server side to embed and encode each microbial resistance detection data in the microbial resistance detection data set, and then performs graph structure adaptive aggregation on the embedded features of each microbial resistance detection data, and then extracts the features corresponding to the microbial category from the embedded features of each microbial resistance detection data, thereby realizing intelligent generation of analysis results based on the attention category guided representation between the microbial category embedded features and the significant aggregation features of the microbial resistance detection data set. In this way, the subtle differences and complex patterns in the data can be better captured, and the adaptability and generalization ability of the model to different data distributions are enhanced, which is especially suitable for complex nonlinear data sets, and can provide more accurate analysis results compared to simple linear models.

[0090] In the embodiment of the present application, the display module 140 is used to display the analysis results. It should be understood that by displaying the analysis results, complex data and analysis conclusions can be presented in an intuitive form, and professionals in medical institutions or public health managers can quickly understand the meaning behind the data through this result. By clearly displaying information such as microbial outbreak trends, susceptible populations and regional distribution, it is also helpful for health departments to take preventive measures in a timely manner, such as adjusting the allocation of medical resources, implementing targeted prevention and control strategies, or issuing early warning notices, thereby effectively responding to possible public health events.

[0091] The following is a detailed description of the specific implementation method of "displaying analysis results":

[0092] Visualization technology is usually used to display the analysis results. There are many different visualization methods, among which chart display is one of the most commonly used methods. For example, the outbreak of microorganisms within a preset time can be presented by drawing a line graph or a bar graph. The line graph can clearly show the trend of the number of microorganisms over time, allowing users to intuitively observe the growth rate, peak period and trend changes of microbial outbreaks, so as to promptly determine whether there is an increased risk of outbreak; the bar graph can compare the number of microorganisms in different time periods to highlight the differences in the degree of outbreak.

[0093] When showing the distribution of people susceptible to microbial infection, you can use pie charts or stacked bar charts. Pie charts can intuitively present the proportion of different age, gender and other population categories in the overall susceptible population, helping health managers quickly identify the characteristics of high-risk groups; stacked bar charts can further subdivide the distribution of people under different combinations of factors, such as the comparison of infection proportions of men and women in different age groups.

[0094] Map visualization is an ideal choice for the geographical distribution of microbial infections. By marking the infection levels of different regions on the map, such as using different colored areas or the size of the marking points to indicate the infection rate, it can clearly show the high-incidence and low-incidence areas of microbial infection and the clustering or diffusion trend of the geographical distribution, which will help the health department to allocate medical resources in a targeted manner and implement regionalized prevention and control strategies.

[0095] The antibiotic resistance status of a certain disease can be displayed through radar charts or tables. Radar charts can simultaneously present the resistance indicators of multiple antibiotics to the disease, such as resistance rate, sensitivity rate, etc., which is convenient for comparing the effects of different antibiotics; the table form can list the specific resistance data of each antibiotic in detail for professionals to conduct in-depth analysis.

[0096] In addition, the display module should also have interactive functions, so that users can choose to view the analysis results of specific data dimensions, time ranges or regional ranges according to their own needs, so as to achieve flexible exploration of data. At the same time, the system can update data in real time to ensure that the displayed analysis results are always based on the latest test information, which has provided timely and accurate basis for public health management decisions.

[0097] In summary, the microbial intelligent detection system 100 based on big data analysis according to the embodiment of the present application is explained, which first obtains the in vitro specimen to be tested of the patient, then uses the microbial detector and the matching self-flowing microbial detection card to perform drug resistance detection to obtain a microbial resistance detection data set, and then uploads the data set to the cloud server with the help of the Internet communication module, and performs data analysis on the uploaded microbial resistance detection data set on the cloud server to obtain the corresponding analysis results, and displays the analysis results. In this way, the analysis of microbial resistance data can be made more accurate, and thus better serve public health management.

[0098] In an embodiment of the present application, a microbial intelligent detection method based on big data analysis is also provided, which is used to execute the microbial intelligent detection system based on big data analysis as described above. Figure 4 Flow chart of the microbial intelligent detection method based on big data analysis according to an embodiment of the present application. Figure 4 As shown, in the microbial intelligent detection method based on big data analysis, it includes: S110, in the cloud server, embedding and encoding each microbial resistance detection data in the microbial resistance detection data set to obtain a set of microbial resistance detection data embedded coding vectors; S120, in the cloud server, performing graph structure adaptive feature significant aggregation on the set of microbial resistance detection data embedded coding vectors to obtain a microbial resistance detection data set significant aggregation feature vector; S130, in the cloud server, extracting a microbial category embedded coding vector from the set of microbial resistance detection data embedded coding vectors, and performing attention interaction response on the microbial category embedded coding vector and the microbial resistance detection data set significant aggregation feature vector to obtain a microbial category guided resistance detection data representation vector; S140, in the cloud server, obtaining the analysis result based on the microbial category guided resistance detection data representation vector.

[0099] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned microbial intelligent detection method based on big data analysis have been referred to above. Figures 1 to 3 The invention has been introduced in detail in the description of the microbial intelligent detection system based on big data analysis, and therefore, its repeated description will be omitted.

[0100] In summary, the intelligent microbial detection method based on big data analysis based on the embodiment of the present application is explained, which first obtains the patient's in vitro specimen to be tested, then uses the microbial detector and the matching self-flowing microbial detection card to perform drug resistance detection to obtain a microbial resistance detection data set, and then uploads the data set to the cloud server with the help of the Internet communication module, and performs data analysis on the uploaded microbial resistance detection data set on the cloud server to obtain the corresponding analysis results, and the analysis results are displayed. In this way, the analysis of microbial resistance data can be made more accurate, and thus better serve public health management.

Claims

1. A microbial intelligent detection system based on big data analysis, comprising: A test data acquisition module is used to obtain a sample to be tested from a patient, and to perform drug resistance detection on the microorganisms in the sample to be tested by using a microorganism detector and a self-flowing microorganism detection card matched therewith to obtain a microorganism drug resistance detection data set; A data transmission module, used to upload the microbial resistance detection data set to a cloud server via an Internet communication module; A data analysis module, used for performing data analysis on the microbial resistance detection data set on the cloud server to obtain an analysis result; A display module is used to display the analysis results; characterized in that the data analysis module includes: A detection data embedding unit, used for embedding and encoding each microbial drug resistance detection data in the microbial drug resistance detection data set on the cloud server to obtain a set of microbial drug resistance detection data embedding coding vectors; A data aggregation unit, configured to perform graph structure adaptive feature significant aggregation on the set of microbial drug resistance detection data embedding coding vectors on the cloud server to obtain a microbial drug resistance detection data set significant aggregation feature vector; An attention guiding fusion unit is used to extract a microorganism category embedding coding vector from the set of microorganism drug resistance detection data embedding coding vectors on the cloud server, and perform attention interaction response on the microorganism category embedding coding vector and the significant aggregation feature vector of the microorganism drug resistance detection data set to obtain a microorganism category guided drug resistance detection data representation vector; The analysis result generating unit is used to obtain the analysis result on the cloud server by guiding the drug resistance detection data characterization vector based on the microorganism category.

2. The microbial intelligent detection system based on big data analysis according to claim 1 is characterized in that: The microbial resistance detection data includes the gender and age of the specimen patient, the type of microorganism, antibiotic resistance, the specimen detection time, and the region where the specimen detection is located.

3. The microbial intelligent detection system based on big data analysis according to claim 2 is characterized in that: The data aggregation unit comprises: A microbial drug resistance detection data mapping subunit, used for mapping each microbial drug resistance detection data embedding coding vector in the set of microbial drug resistance detection data embedding coding vectors to a hyperbolic space to obtain a set of microbial drug resistance detection data embedding coding vectors after hyperbolic space mapping; A microbial drug resistance detection data topology calculation subunit, used to obtain a microbial drug resistance detection data graph walk topology feature matrix based on a set of microbial drug resistance detection data embedding coding vectors after hyperbolic space mapping; A microbial drug resistance detection data weight calculation subunit, used to calculate the autocorrelation gated significant confidence weight factor of each microbial drug resistance detection data embedded coding vector in the set of microbial drug resistance detection data embedded coding vectors based on the set of microbial drug resistance detection data embedded coding vectors after hyperbolic space mapping and the microbial drug resistance detection data graph walk topology feature matrix to obtain a set of microbial drug resistance detection data autocorrelation gated significant confidence weight factors; The microbial resistance detection data aggregate feature generation subunit is used to calculate the position-weighted sum of the set of microbial resistance detection data embedded coding vectors based on the set of microbial resistance detection data autocorrelation gated significant confidence weight factors to obtain the significant aggregate feature vector of the microbial resistance detection data set.

4. The microbial intelligent detection system based on big data analysis according to claim 3 is characterized in that: The microbial drug resistance detection data topology calculation subunit is used to: Calculating the Poincare distance between any two embedding coding vectors of the microbial drug resistance detection data after hyperbolic space mapping in the set of embedding coding vectors of the microbial drug resistance detection data after hyperbolic space mapping to obtain a walk topology matrix of the microbial drug resistance detection data graph; The walk topology matrix of the microbial drug resistance detection data graph is subjected to dilated convolution encoding to obtain the walk topology feature matrix of the microbial drug resistance detection data graph.

5. The microbial intelligent detection system based on big data analysis according to claim 4 is characterized in that: The microbial drug resistance detection data weight calculation subunit is used to: Performing global context walk of graph convolution on the microbial resistance detection data graph walk topology feature matrix and the set of microbial resistance detection data embedding coding vectors after hyperbolic space mapping to obtain a set of microbial resistance detection data context semantics enhancement feature vectors; Performing autocorrelation gating calculation on each corresponding group of microbial resistance detection data context semantics enhancement feature vectors and microbial resistance detection data embedding coding vectors in the set of the microbial resistance detection data context semantics enhancement feature vectors and the set of the microbial resistance detection data embedding coding vectors to obtain a set of microbial resistance detection data autocorrelation gating significant confidence factors; The set of the microbial drug resistance detection data autocorrelation gated significant confidence factors is normalized by a Softmax function to obtain a set of the microbial drug resistance detection data autocorrelation gated significant confidence weight factors.

6. The microbial intelligent detection system based on big data analysis according to claim 5 is characterized in that: The attention guidance fusion unit is used to: Extracting the microbial drug resistance detection data embedding code vector corresponding to the microbial species from the set of microbial drug resistance detection data embedding code vectors as the microbial category embedding code vector; The microorganism category embedding encoding vector and the microorganism resistance detection data set significant aggregation feature vector are passed through a microorganism category guided infection response expressor based on an interactive response attention network to obtain the microorganism category guided resistance detection data representation vector.

7. The microbial intelligent detection system based on big data analysis according to claim 6 is characterized in that: The analysis result generating unit is used to: pass the microorganism category-guided drug resistance detection data characterization vector through an AIGC-based microorganism infection situation generator to obtain the analysis result.

8. The microbial intelligent detection system based on big data analysis according to claim 7 is characterized in that: The analysis results include the outbreak of microorganisms within a preset time, the distribution of people susceptible to microbial infection, the geographical distribution of microbial infection, and the resistance of antibiotics to a certain disease.

9. A microbial intelligent detection method based on big data analysis, used to execute the microbial intelligent detection system based on big data analysis as claimed in claim 1, characterized in that: The microbial intelligent detection method based on big data analysis includes: On the cloud server, each microbial drug resistance detection data in the microbial drug resistance detection data set is embedded and encoded to obtain a set of microbial drug resistance detection data embedded encoding vectors; On the cloud server, performing graph structure adaptive feature significant aggregation on the set of microbial drug resistance detection data embedding coding vectors to obtain a microbial drug resistance detection data set significant aggregation feature vector; On the cloud server, extracting a microorganism category embedding coding vector from the set of microorganism drug resistance detection data embedding coding vectors, and performing attention interaction response on the microorganism category embedding coding vector and the significant aggregation feature vector of the microorganism drug resistance detection data set to obtain a microorganism category guided drug resistance detection data representation vector; On the cloud server, the drug resistance detection data characterization vector is guided based on the microorganism category to obtain the analysis result.

10. The method for intelligent detection of microorganisms based on big data analysis according to claim 9, characterized in that: On the cloud server, the analysis result is obtained based on the microbial category-guided drug resistance detection data characterization vector, including: passing the microbial category-guided drug resistance detection data characterization vector through an AIGC-based microbial infection situation generator to obtain the analysis result.

Images