Method and system for processing various meningitis data of machine learning assisted sensor array

Abstract

The embodiment of the invention provides a method and a system for processing various meningitis data of a machine learning auxiliary sensor array, and belongs to the technical field of data processing and sensor arrays. The method comprises the following steps: acquiring a sensing element; the sensing element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework; constructing a sensing array according to the sensing elements; obtaining a fluorescence signal according to the sensing array and the meningitis cerebrospinal fluid; and obtaining meningitis type information according to the fluorescence signal and a machine learning algorithm. According to the invention, data processing of different infection types of meningitis can be realized.

Description

Technical Field

[0001] The present application relates to the technical field of data processing and sensor arrays, and in particular to a method and system for processing multiple meningitis data using a machine learning-assisted sensor array. Background Art

[0002] Meningitis is a clinical syndrome characterized by inflammation of the meninges and is one of the most common infectious diseases of the central nervous system. The typical symptoms of meningitis are fever, stiff neck, headache, seizures, and cognitive impairment. Depending on the type of infectious pathogen, meningitis can be divided into bacterial, tuberculous, viral, cryptococcal, fungal, etc.

[0003] In the process of processing various types of meningitis data, despite continuous technological advances, there are still some significant shortcomings and challenges. In data analysis and modeling, the model is complex: advanced analytical techniques (such as machine learning and deep learning) can process complex data, but the training and optimization of the model requires a lot of computing resources and data, and is easily affected by data bias. In computing resources and efficiency, there are high computing requirements: processing high-dimensional data (such as imaging data, genomic data) requires a lot of computing resources and storage, which may lead to high costs and long processing times. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to provide a method and system for processing multiple meningitis data using a machine learning-assisted sensor array.

[0005] The technical solution adopted by the present invention is:

[0006] In one aspect, an embodiment of the present invention provides a method for processing multiple meningitis data using a machine learning-assisted sensor array, the method comprising the following steps:

[0007] Obtaining a sensing element; wherein the sensing element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework;

[0008] According to the sensing element, a sensing array is constructed;

[0009] Obtaining a fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid;

[0010] According to the fluorescent signal and the machine learning algorithm, meningitis type information is obtained.

[0011] Furthermore, the obtaining of the sensor element comprises the following steps:

[0012] synthesize metal-organic frameworks, covalent organic frameworks, and hydrogen-bonded organic frameworks using synthetic methods;

[0013] The synthesis methods include one-pot method, post-synthesis method and hydrothermal method;

[0014] In carrying out the synthesis method, a variety of organic ligands and metal salts are used to react in a solvent;

[0015] The metal organic framework, the covalent organic framework and the hydrogen bond organic framework are placed in a 3+ soaked in a solution to obtain a sensing element.

[0016] Furthermore, constructing a sensor array according to the sensor element comprises the following steps:

[0017] A plurality of the sensing elements are used to construct a sensing array.

[0018] Further, obtaining the fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid comprises the following steps:

[0019] Obtain meningitis cerebrospinal fluid for several types of infection;

[0020] Respectively reacting each of the sensing elements in the sensing array with samples of the meningitis cerebrospinal fluid of different infection types to obtain sensing signal information;

[0021] A fluorescence signal is obtained according to the sensing signal information.

[0022] Furthermore, the step of reacting each of the sensor elements in the sensor array with samples of meningitis cerebrospinal fluid of different infection types to obtain sensor signal information comprises the following steps:

[0023] 20 μL of the meningitis cerebrospinal fluid of different infection types was reacted with 50 μL of each of the sensing elements to obtain sensing signal information.

[0024] Furthermore, obtaining the fluorescence signal according to the sensor signal information comprises the following steps:

[0025] The sensing signal information is used to obtain a fluorescence response signal generated by the sensing element when the excitation wavelength is 313 nm;

[0026] According to the fluorescence response signal, Eu 3+ The emission wavelength of the fluorescence signal is 616 nm.

[0027] Furthermore, obtaining meningitis type information according to the fluorescence signal and the machine learning algorithm also includes the following steps:

[0028] Processing and analyzing the collected fluorescence signals using a machine learning algorithm to obtain analysis data;

[0029] Build a classification model;

[0030] According to the analysis data and the classification model, pattern recognition and feature extraction are performed on the fluorescence signal to obtain meningitis type information.

[0031] On the other hand, an embodiment of the present invention provides a machine learning-assisted sensor array multiple meningitis data processing system, the system comprising:

[0032] The first module is used to prepare a sensor element; the sensor element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework;

[0033] A second module is used to construct a sensor array according to the sensor elements;

[0034] A third module is used to obtain a fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid;

[0035] The fourth module is used to obtain meningitis type information based on the fluorescent signal and the machine learning algorithm.

[0036] On the other hand, an embodiment of the present invention provides a device for processing multiple meningitis data using a machine learning-assisted sensor array, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for processing multiple meningitis data using a machine learning-assisted sensor array as described above is implemented.

[0037] On the other hand, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method as described above.

[0038] The embodiments of the present application include at least the following beneficial effects: The present application provides a method and system for processing multiple meningitis data using a machine learning-assisted sensor array. The present invention obtains a sensor element; the sensor element is synthesized by a metal organic framework, a covalent organic framework, and a hydrogen bond organic framework; a sensor array is constructed based on the sensor element; a fluorescent signal is obtained based on the sensor array and meningitis cerebrospinal fluid; and meningitis type information is obtained based on the fluorescent signal and a machine learning algorithm. The present invention can realize data processing of different types of meningitis infections and assist in determining the type of disease. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flow chart of a method for processing multiple meningitis data using a machine learning-assisted sensor array provided by an embodiment of the present invention;

[0040] Figure 2is a scanning electron microscope SEM image of 10 sensor elements in Example 3 provided by the present invention;

[0041] Figure 3 is the X-ray diffraction pattern (XRD) of 10 sensor elements in Example 3 provided by the present invention;

[0042] Figure 4 is a Fourier transform infrared spectrum diagram of 10 sensor elements in Example 3 provided by the present invention;

[0043] Figure 5 It is the fluorescence spectrum of the representative sensor element before and after the reaction with different volumes of cerebrospinal fluid at an excitation wavelength of 310 nm in Example 4 provided by the present invention;

[0044] Figure 6 It is the fluorescence spectrum before and after the reaction between the sensor element and 20 μL cerebrospinal fluid at an excitation wavelength of 310 nm in Example 4 provided by the present invention;

[0045] Figure 7 It is a schematic diagram of the box plot of the fluorescent signal of the reaction characteristic of TBM cerebrospinal fluid and sensor array, the correlation, the confusion matrix of the classification model and the importance analysis of the sensor element in the test example 1 provided by the present invention;

[0046] Figure 8 It is a box plot of the fluorescence signal of the cerebrospinal fluid of the non-intracranial infection control group and the reaction characteristic of the sensor array, a loss function curve, and a ROC curve of the TBM cerebrospinal fluid and non-intracranial infection in the test example 1 provided by the present invention;

[0047] Fig. 9 It is a schematic diagram of the box plot, correlation, confusion matrix of the classification model and importance analysis of the sensor element for detecting BM cerebrospinal fluid and characteristic fluorescence signals of the sensor array in Experimental Example 2 provided by the present invention;

[0048] Fig.10 It is a schematic diagram of the box plot, correlation, confusion matrix of the classification model and importance analysis of the sensor element for detecting CM cerebrospinal fluid and characteristic fluorescence signals of the sensor array in Test Example 3 provided by the present invention;

[0049] Fig.11 It is a schematic diagram of the box plot, correlation, confusion matrix of the classification model and importance analysis of the sensor element for detecting VM cerebrospinal fluid and characteristic fluorescence signals of the sensor array in Test Example 4 provided by the present invention;

[0050] Fig.12 The LDH graph after the reaction between cerebrospinal fluid of different infection types and the sensor array in Experimental Example 5 provided by the present invention, the confusion matrix of the classification model, the importance analysis of the sensor element, and the loss function curve;

[0051] Fig.13 It is a schematic diagram of ROC analysis of BM, CM, VM and TBM cerebrospinal fluid under the classification model provided by the present invention;

[0052] Fig.14 It is a technical circuit diagram of the test example provided by the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are only examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the attached claims.

[0054] It is understood that the terms "first", "second", etc. used in this application can be used to describe various concepts in this article, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another concept. For example, without departing from the scope of the embodiment of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determination".

[0055] The terms "at least one", "multiple", "each", "any", etc. used in this application, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0057] Before describing the embodiments of the present application in detail, some nouns and terms involved in the embodiments of the present application are first described. The nouns and terms involved in the embodiments of the present application are subject to the following explanations.

[0058] 1) MOF (Metal-Organic Frameworks): Metal-organic frameworks refer to three-dimensional porous materials formed by metal ions or metal clusters and organic ligands through coordination bonds, which are used in applications such as gas storage and catalysis.

[0059] 2) COF (Covalent Organic Frameworks): Covalent organic frameworks are two-dimensional or three-dimensional structures formed by organic molecules connected by covalent bonds. They are mainly used in hydrogen storage, catalysis and separation.

[0060] 3) HOF (Hydrogen-bonded Organic Frameworks): Hydrogen-bonded organic frameworks, organic framework materials self-assembled through hydrogen bonding, used in applications such as gas storage, catalysis and separation.

[0061] 4) Eu 3+ : It is the trivalent europium ion in the praseodymium element and has significant fluorescent properties.

[0062] 5) BM (Bacterial Meningitis): Bacterial meningitis, meningitis caused by bacterial infection.

[0063] 6) CM (Cryptococcal Meningitis): Cryptococcal meningitis, meningitis caused by infection with Cryptococcus neoformans.

[0064] 7) VM (Viral Meningitis): Viral meningitis, meningitis caused by viral infection.

[0065] 8)TBM (Tuberculous Meningitis): tuberculous meningitis, meningitis caused by infection with Mycobacterium tuberculosis.

[0066] 9) ROC (Receiver Operating Characteristic) analysis: A method used to evaluate the performance of classification models.

[0067] The present invention is based on a comprehensive analysis strategy for meningitis using a machine learning-assisted cross-reactive sensor array. In order to improve the recognition ability of complex samples, the present invention constructs 10 organic framework-doped Eu 3+ The sensor array composed of sensor elements cross-reacted with four different types of meningitis samples to output different response patterns. The output signals of 10 sensor elements were used as feature signals, and machine learning was used to train and predict feature signals to achieve data processing for different types of meningitis.

[0068] The machine learning-assisted sensor array method for processing multiple meningitis data of the present invention proposes using the sensor array to cross-react with cerebrospinal fluid to generate a specific fluorescence response pattern. Combined with the machine learning algorithm, it demonstrates excellent classification performance and differentiation capabilities, and can perform image recognition and classification to assist in determining the type of disease.

[0069] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0070] On the one hand, the embodiment of the present invention provides a method for processing multiple meningitis data using a machine learning-assisted sensor array. Specifically, referring to Figure 1 , the method comprises the following steps:

[0071] S100, obtaining a sensing element; the sensing element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework;

[0072] S200, constructing a sensor array according to the sensor elements;

[0073] S300, obtaining a fluorescence signal according to the sensor array and the meningitis cerebrospinal fluid;

[0074] S400. Obtain meningitis type information based on the fluorescence signal and the machine learning algorithm.

[0075] The present invention uses machine learning algorithms to analyze and classify these fluorescent signals, which can effectively assist in distinguishing different types of meningitis.

[0076] The embodiment of the present invention discloses step S100 of acquiring a sensor element, including the following steps:

[0077] S110. Synthesize metal organic frameworks, covalent organic frameworks, and hydrogen bonded organic frameworks using synthetic methods;

[0078] S120, synthesis methods include one-pot method, post-synthesis method, and hydrothermal method;

[0079] S130, in the process of performing the synthesis method, using a plurality of organic ligands and metal salts to react in a solvent;

[0080] S140, placing metal organic frameworks, covalent organic frameworks and hydrogen bonded organic frameworks in a 3+ soaked in a solution to obtain a sensing element.

[0081] As an optional embodiment, the preparation of the metal organic framework (MOF) of the present invention:

[0082] Select appropriate metal salts (such as aluminum salts, zinc salts, copper salts, etc.) and organic ligands (such as phthalic acid, furanic acid, etc.) as reaction raw materials. Carry out the reaction in a specific solvent (such as water, ethanol, DMF, etc.), and use the one-pot method, post-synthesis method or hydrothermal method to synthesize MOF materials. The hydrothermal method is usually synthesized under high temperature and high pressure conditions, while the one-pot method and post-synthesis method are carried out at room temperature and during post-synthesis treatment, respectively. The obtained MOF material should be properly washed, dried and activated to ensure its pore structure and stable performance.

[0083] Preparation of Covalent Organic Frameworks (COFs):

[0084] Select appropriate organic monomers (such as p-phenylenediamine, 2,6-lutidine, etc.) and react with chemical reagents such as acetic acid in an organic solvent (such as chloroform, tetrahydrofuran, etc.) to generate COF.

[0085] The post-synthesis method is used to prepare COF to ensure that the monomers are successfully connected by covalent bonds to form a two-dimensional or three-dimensional framework structure.

[0086] The synthesized COF material is washed and dried to remove unreacted substances and solvent residues.

[0087] Preparation of Hydrogen Bonding Organic Framework (HOF):

[0088] Select appropriate organic molecules that can self-assemble into framework structures through hydrogen bonds.

[0089] In a specific solvent environment, organic molecules can undergo hydrogen bond self-assembly by adjusting conditions such as temperature and pH value.

[0090] The obtained HOF materials need to be properly processed to enhance their stability and performance, including drying and activation steps.

[0091] Eu 3+ Doping:

[0092] The prepared MOF, COF and HOF materials were immersed in a solution containing Eu 3+ solution, or directly add Eu during the synthesis process 3+ .

[0093] Through appropriate reaction conditions (such as temperature, time, pH value, etc.), Eu 3+ Uniform doping in organic frameworks.

[0094] The doped material needs to be washed and dried to remove excess Eu 3+ and solvents.

[0095] The embodiment of the present invention discloses that step S200 constructs a sensor array according to the sensor elements, including the following steps:

[0096] S210, using a plurality of sensor elements to construct a sensor array.

[0097] The embodiment of the present invention discloses step S300 of obtaining a fluorescent signal according to a sensor array and meningitis cerebrospinal fluid, including the following steps:

[0098] S310, obtain meningitis cerebrospinal fluid for several types of infection;

[0099] S320, reacting each sensor element in the sensor array with samples of cerebrospinal fluid of meningitis of different infection types to obtain sensor signal information;

[0100] S330, obtaining a fluorescence signal according to the sensing signal information.

[0101] The embodiment of the present invention discloses that step S320 reacts each sensor element in the sensor array with samples of cerebrospinal fluid of different infection types to obtain sensor signal information, including the following steps:

[0102] S321. React 20 μL of meningitis cerebrospinal fluid of different infection types with 50 μL of each sensor element to obtain sensor signal information.

[0103] The embodiment of the present invention discloses that step S330 obtains a fluorescent signal according to the sensor signal information, including the following steps:

[0104] S331, sensing signal information, when the excitation wavelength is 313 nm, obtaining the fluorescence response signal generated by the sensing element;

[0105] S332, according to the fluorescence response signal, obtain Eu 3+ The emission wavelength of the fluorescence signal is 616 nm.

[0106] As an optional embodiment, the present invention takes 20 μL of meningitis cerebrospinal fluid samples of different infection types and reacts them with each sensor element in the sensor array respectively, ensuring uniform contact between the sample and the sensor element, and controlling the reaction time to ensure data accuracy.

[0107] Fluorescence signal acquisition: Use a light source with an excitation wavelength of 313 nm to excite the sensor array, and measure the fluorescence emission intensity of each sensor element at a wavelength of 616 nm using a spectrometer or fluorescence detector. Record the fluorescence response signal of each sensor element in different meningitis cerebrospinal fluid samples for subsequent analysis.

[0108] The embodiment of the present invention discloses that step S400 obtains meningitis type information according to the fluorescent signal and the machine learning algorithm, and further includes the following steps:

[0109] S410, using a machine learning algorithm to process and analyze the collected fluorescence signals to obtain analysis data;

[0110] S420, constructing a classification model;

[0111] S430. Perform pattern recognition and feature extraction on the fluorescent signal according to the analysis data and the classification model to obtain meningitis type information.

[0112] The data preprocessing of the present invention: normalize and standardize the collected fluorescence signals to eliminate system errors and sample differences, and organize the fluorescence signal data into the input format required by the machine learning algorithm, such as feature vectors or matrices.

[0113] The present invention constructs a classification model by selecting a suitable machine learning algorithm (such as support vector machine, random forest, neural network, etc.) and using fluorescence signal data for training. Through cross-validation and model optimization, the accuracy and robustness of the classification model are ensured.

[0114] The type identification of the present invention: the processed fluorescent signal is input into the trained classification model to predict the infection type of the meningitis sample. The specific information of different infection types of meningitis is obtained through the classification results output by the model.

[0115] As an optional embodiment, the present invention proposes a comprehensive analysis strategy for meningitis based on machine learning-assisted cross-reactive sensor arrays, and constructs 10 organic framework-doped Eu 3+ In the present invention, the sensor array reacts with potential targets of cerebrospinal fluid of meningitis of different infection types to generate specific fluorescence response signals, and finally the fluorescence signals are processed by machine learning algorithms, which effectively verifies that the sensor array combined with the machine learning algorithm can effectively distinguish meningitis data of different infection types.

[0116] In one aspect of the present invention, a sensor array consisting of 10 sensor elements is provided, wherein the sensor elements are MOF, COF and HOF doped with Eu 3+ form.

[0117] The present invention generates a specific fluorescence response signal by reacting 20 μL of cerebrospinal fluid of different infection types with 50 μL of each sensor element at an excitation wavelength of 313 nm. u3+ The strongest emission at 616nm is used as the fluorescence signal output of the sensor element. A classification model is constructed for the fluorescence signal using a machine learning algorithm to achieve data processing of different types of meningitis infection and obtain infection type information.

[0118] Another aspect of the present invention provides a method for preparing the sensor array according to the previous aspect of the present invention, comprising the following steps:

[0119] (1) Preparation of MOF, COF and HOF: They can be synthesized by one-pot method, post-synthesis method or hydrothermal method, including various MOFs generated by reaction of different organic ligands and metal salts in solvents and HOF formed by self-assembly of hydrogen bonds. COF is generated by post-synthesis method by reacting p-phenylenediamine and 2,6-lutidine with acetic acid in an organic mixed solution.

[0120] (2)Eu 3+ Doping: The synthesized MOF, COF and HOF are doped with Eu 3+ solution, or directly add Eu during the reaction 3+ , realizing Eu doping in various organic frameworks 3+ , enhancing its fluorescence properties.

[0121] The present invention proposes to use MOF, HOF, COF doped with Eu 3+ A sensor array consisting of 10 sensor elements was constructed, and specific fluorescence response signals were generated by reacting with cerebrospinal fluid of different pathogenic types of meningitis. Finally, a classification model was constructed for the fluorescence response signals through a machine learning algorithm, which enabled data processing of different infection types of meningitis and obtained type result information.

[0122] As an optional embodiment, the present invention has the following embodiments:

[0123] Example 1 Materials and Chemical Reagents

[0124] Zirconium chloride octahydrate (ZrOCl2 8H2O), copper chloride hexahydrate (CuCl2 6H2O), terephthalic acid (H2BDC), benzene-1,3,5-tricarboxylic acid (H3BTC), 2,6-pyridinedicarboxylic acid (DPA), isophthalic acid, melamine, 1,2,4,5-benzenetetracarboxylic acid (H4BTEC), copper nitrate trihydrate (Cu(NO3)2 3H2O, Zinc nitrate hexahydrate (Zn(NO3)2 6H2O), aluminum chloride hexahydrate (AlCl3 6H2O), europium nitrate hexahydrate (Eu(NO3)3 6H2O), 2-methylimidazole, 4,4-bipyridine and 2,2′-bipyridine-3,3′-dicarboxylic acid, N,N-dimethylformamide (DMF), anhydrous sodium acetate and anhydrous ethanol, p-phenylenediamine (Pda), m-trimethylbenzene, 2,6-diformylpyridine (Dfp), 1,4-dioxane, acetonitrile and acetic acid can all be obtained through conventional commercial channels, and deionized water (18.2MΩ cm) can be obtained through the Milli-Q water system.

[0125] Example 2 Instruments and Equipment

[0126] SEM was performed on a scanning electron microscope, and Fourier transform infrared spectra (FT-IR) were recorded on a Tensor 27 FT-IR spectrometer. Fluorescence intensity was recorded on a Fluoromax-4 fluorescence spectrometer. Fluorescence values ​​were recorded on a SpectraMaxR ID3 multi-mode microplate reader.

[0127] Example 3 Synthesis of 10 sensor elements in a sensor array

[0128] In this embodiment, a method for synthesizing a sensor array composed of 10 sensor elements is provided, which specifically includes the following steps:

[0129] Eu 3+ / UiO-66(EM1):EM1 was synthesized by one-pot method. First, ZrOCl2 8H2O (84 mg) and H2BDC (180 mg) were dissolved in 12 mL and 4 mL N,N-dimethylformamide (DMF) respectively. After stirring for 5 minutes, 3 mL CH3COOH was added during stirring. Eu(NO3)3 6H2O (1M, 100 μL) was quickly added to the above solution. Then, the solution was placed in a 50 mL Teflon-lined stainless steel container and reacted at 90°C for 15 h. The precipitate was collected by centrifugation (8000 rpm, 10 min) and washed 3 times with DMF. Finally, the precipitate was dried at 60°C for 6 h to obtain EM1.

[0130] Eu 3+ / HKUST-1(EM2): EM2 was synthesized by post-synthesis method. H3BTC (430 mg) methanol solution (50 mL) was slowly added to Cu(NO3)2 3H2O (900 mg) and PVP (400 mg) in a methanol solution (50 mL) and stirred evenly. After stirring for 10 minutes, the solution was placed in the dark for one day. The resulting blue solid was collected, centrifuged (9000 rpm, 15 min), washed with methanol several times, and dried in an oven (50 ° C, 5 h). Then the resulting product (100 mg) was immersed in Eu(NO3)3 6H2O (0.1M, 10mL) for 12h. Finally, EM2 was collected by centrifugation (9000rpm, 15min), washed with deionized water for several times, and dried in an oven (60℃, 10h) to obtain EM2.

[0131] Eu 3+ / MIL-53(EM3): EM3 was synthesized by one-pot method. Briefly, AlCl3 6H2O (45.9 mg), H2BDC (66.4 mg) and Eu(NO3)3 6H2O (5 mg) was mixed in 30 mL of ethanol, DMF and deionized water (DMF:H2O:ethanol = 20:5:5). After stirring for 30 minutes, the solution was heated at 130 °C for 8 h in a 50 mL Teflon-lined stainless steel container. Finally, it was centrifuged (8000 rpm, 15 min), washed with deionized water and ethanol three times, and then dried in an oven at 60 °C for 10 h to obtain EM3.

[0132] Eu-MOF-1(EM4): 410mg Eu(NO3)3 6H2O and 150 mg of anhydrous sodium acetate were ultrasonically dissolved in 45 mL of deionized water to obtain solution A, and 200 mg of benzene 1,3,5-tricarboxylic acid (H3BTC) was ultrasonically dissolved in 45 mL of ethanol to obtain solution B. Next, solution A was added to solution B, and after vigorous stirring at room temperature for 60 min, the precipitate was collected by centrifugation, washed with ethanol and water three times, and then dried in an oven at 60 °C for 24 h to obtain EM4.

[0133] Eu-MOF-2(EM5): In a hydrothermal tube, Eu(NO3)3 6H2O (0.1 mmol, 0.0435 g) and H4btec (0.1 mmol, 0.0254 g) were dissolved in 5 mL of deionized water. The solution was then heated at 120°C for 12 h and the reaction was cooled to room temperature to obtain EM5.

[0134] Eu 3+ / Zn-MOF-1(EM6):Zn(NO3)2 6H2O (0.3 mmol), H3BTC (0.3 mmol) and 4,4-bipyridine (0.15 mmol) were dissolved in 10 mL DMF. Then Eu(NO3)3 was added to the above solution during stirring. 6H2O (0.1 mmol). Then, the reaction was carried out in a 20 mL polytetrafluoroethylene-lined stainless steel container at 120°C for 48 hours. Finally, Zn-MOF-1 was collected by centrifugation, washed with DMF several times, and dried in an oven to obtain EM6.

[0135] Eu 3+ / COF(EC): First, Pda (0.5mmol, 54mg) was dissolved in a mixture of xylene (2mL) and 1,4-dioxane (3mL). Then, Dfp (0.75mmol, 101mg) was also dissolved in a mixture of xylene (2mL) and 1,4-dioxane (3mL). Subsequently, one of the mixtures was slowly poured into the other mixture. Then acetic acid (0.5mL, 6M) was added dropwise. After the mixture was left at room temperature for two days, a yellow precipitate was obtained by centrifugation. Wash twice with tetrahydrofuran and once with acetone, and then dry the powder at 65°C. The obtained COF was stored at room temperature for further use. To load Eu 3+ , Eu(NO3)3·6H2O (60 mg) was dissolved in acetonitrile. Then 10 mg COF was weighed and added to the terbium nitrate acetonitrile solution. After soaking for 1 day, the mixture was transferred to a Teflon-lined stainless steel autoclave, heated in an oven at 85°C for 24 h, and then centrifuged to obtain a dark red precipitate. Washed three times with acetonitrile, the product was dried at 65°C to obtain EC.

[0136] Eu 3+ / MA-IPA (EH1): Melamine (MA, 0.1262 g, 0.0010 mol) was added to 30 mL of distilled water while stirring in an oil bath. The MA solution was then heated to 150 °C until the MA powder was completely dissolved. Subsequently, the MA solution was moved to a 70 °C oil bath and stirred for 10 min, and isophthalic acid (0.1662 g, 0.0010 mol) was added. After stirring for 3 h, the transparent mixed solution was taken out and cooled to room temperature. Pure white MA-IPA crystals could be separated from the mixture. Then, the MA-IPA crystals were dispersed in 50 mL of ethanol solution containing Eu(NO3)3·6H2O (0.1200 g, 0.2690 mmol). Stir at room temperature for 6 h, and collect the white solid by centrifugation to obtain EH1.

[0137] Eu 3+ / MA-TPA(EH2): Terephthalic acid (TPA, H2BDC; 0.0498 g, 0.3 mmol) and melamine (MA; 0.0378 g, 0.3 mmol) were stirred in 50 mL of deionized water at room temperature for 30 min. The mixed liquid was then refluxed in an oil bath at 170 °C for 2 h and heated at 20 °C for 1 h at room temperature. -1The mixture was cooled at a rate of 100 ℃ to obtain colorless layered crystals. The colorless crystals were washed with ethanol and dried in air. MA-TPA (50 mg) was then dispersed in 50 mL of ethanol solution containing Eu(NO3)3·6H2O (4 mg). Stir at room temperature for 6 h and collect the white solid by centrifugation. The product was washed with ethanol 3 times to remove excess Eu(NO3)3·6H2O physically adsorbed on the surface of Eu3+ / MA-TPA. The mixed material (Eu3+ / MA-TPA) was vacuum dried at 80°C overnight to obtain EH2.

[0138] The 10 sensor elements prepared in the example were characterized.

[0139] (1) The morphology of 10 sensor elements under low-power and high-power microscope fields were observed using a HITACHI SU8220 scanning electron microscope to obtain SEM images. The results are as follows: Figure 2 As shown, it can be found that the sensor array consists of 7 Eu 3+ -MOF composition, 2 Eu 3+ -doped HOFs and 1 Eu 3+ -doped COF, where Eu 3+ / UiO-66 is composed of ZrOCl2 8H2O and H2BDC synthesis. 3+ / HKUST-1 is composed of H3BTC and Cu(NO3)2 3H2O synthesis. Eu 3+ / MIL-53 is made of AlCl3 6H2O and H2BDC. Eu-MOF-1 is composed of Eu(NO3)3 6H2O and H3BTC. Eu-MOF-2 was synthesized from 6H2O and H4btec. Zn-MOF-1 was synthesized from Zn(NO3)2 6H2O, H3BTC and 4,4-bipyridine. Zn-MOF-3 is composed of Zn(NO3)2 6H2O, 2,2'-bipyridine-3,3'-dicarboxylic acid and 2-methylimidazole. They are denoted as EM1, EM2, EM3, EM4, EM5, EM6 and EM7 respectively. Figure 2 From the SEM images of AC, it can be seen that EM1 and EM3 are spherical structures with average sizes of 40±5 and 200±5nm. EM2 presents fine rods with a length of 1-2μm and a width of 100-150nm. The synthesized Eu-MOF-1 has a microsphere structure of 4μm ( Figure 2 D). Figure 2Figure E is the SEM image of Eu-MOF-2, showing a straw bundle structure. Zn-MOF-2 is an amorphous bulk morphology, but the morphology of nanosheets can still be identified ( Figure 2 F in the figure). Zn-MOF-3 exhibits a flake structure with a size of 300±50nm ( Figure 2 G in Eu 3+ -doped COF is composed of Pda and Dfp, i.e., EC. The hydrogen bonding interaction between MA and IPA / TPA may drive the two Eu2+ / -H2+ 3+ -doped HOFs. Figure 2 As shown in H, doped Eu 3+ The 2D COFs assembled together and formed flower-like clusters. Eu-HOF-1 was synthesized from MA and IPA, and Eu-HOF-2 was synthesized from MA and PTA. Figure 2 As shown in Figure 1J, the morphologies of synthesized Eu-HOF-1 and Eu-HOF-2 are flower-like crystals and needle-like crystals, respectively. This HOF is a highly ordered two-dimensional layered structure formed by hydrogen bond self-assembly.

[0140] (2) The 9 sensor elements were tested by X-ray diffraction analysis, and the X-ray diffraction (XRD) patterns of the 9 sensor elements were obtained. It can be found that the XRD patterns of 7 EMs ( Figure 3 AG) is consistent with its standard XRD pattern, indicating that Eu 3+ The doping of EH1 and EH2 will not affect the crystal structure of MOFs. Figure 3 HI in the reference), referring to the crystal X-ray diffraction results reported in previous literature, it is shown that the HOF material has been successfully prepared and 3+ The post-synthesis did not destroy the original HOF crystal structure.

[0141] (3) Use Tensor 27 Fourier transform infrared spectrometer to record the Fourier transform infrared (FT-IR) spectra of 10 sensor elements (such as Figure 4 As shown in Figure 2, it can be found that the FT-IR spectra of the 10 sensing materials correspond to their organic ligands, confirming the successful synthesis of the materials. For EM1, the FT-IR spectra at 1581 and 1396 cm -1 ν as (-COO-) and ν s The stretching vibration of (-COO-) belongs to the characteristic band of the carboxylic acid group in the PTA ligand, and the peak of the benzene ring is at 1506 cm -1 For EM2, EM4 and EM6, at 1630cm -1 The absorption band at 1450cm belongs to the stretching vibration of the -COOH group; -1The peak at 600 cm belongs to the stretching vibration of -C=C bond. -1 The peak at 1693 cm-1 should correspond to the Al-O stretching vibration. -1 The disappearance of the absorption band at 3+ -MOFs are formed by carboxyl groups and Eu 3+ ions are formed by coordination. EM5 is at 3401 and 3486 cm -1 The infrared peak at 1667 cm is attributed to -OH stretching. -1 An infrared peak was observed at 1610 cm -1 , 1540cm -1 , 1495cm -1 and 1379cm -1 For EM7, 1595cm -1 The absorbance at 1620 cm-1 is attributed to the vibration of NH in 2-methylimidazole. -1 ) peak, the greatly attenuated C=O stretching band (1716cm -1 ) indicates that there is a small amount of unreacted aldehyde on the periphery of COF. -1 The absorption peak of EH1 is attributed to the antisymmetric stretching vibration of C=O bond (-COO- group). -1 The absorption peaks in the range are attributed to the stretching vibration of the OH bond (OH of the carboxyl group in IPA). EH2 at 1637 cm -1 The C=O vibration band at 3395cm belongs to -COOH. -1 and 3341cm -1 The absorption peaks belong to the antisymmetric stretching vibration and symmetric stretching vibration of NH2 in MA. Example 4 Fluorescence signal detection of sensor element

[0142] (1) In a typical experiment, 50 μL of 10 sensor element solutions (2.5 mg mL -1 ) was mixed with 250 μL of deionized water and reacted at room temperature for 20 minutes. The fluorescence spectrum of the sensor element was recorded using a fluorescence spectrometer at Ex / Em (nm) 310:616 nm (scanning range: 550-720 nm).

[0143] (2) Fluoromax-4 spectrofluorometer was used to record the fluorescence intensity of different volumes (10, 20, 50, 200 μL) of cerebrospinal fluid (CSF) after reaction with the sensor element. Figure 5), Ex / Em (nm) 310:616nm (scanning range: 550-720nm), determine the optimal cerebrospinal fluid concentration and material concentration to obtain the best fluorescence intensity signal change and improve the signal-to-noise ratio. It can be found that 10μL CSF has little effect on the sensor element, while 200μL CSF cannot detect signal changes through completely quenchable fluorescence. At the same time, considering that cerebrospinal fluid samples are difficult to obtain, 20μL cerebrospinal fluid was selected as the optimal reaction concentration. 20μL cerebrospinal fluid and 50μL sensor element solution (2.5mg·mL -1 ) were mixed, and deionized water was added to keep the total volume at 300 μL. In the 8×10 area of ​​the 96-well plate, Ex / Em (nm) 310:616 nm was selected to record the fluorescence spectra before (I0) and after (I) the addition of cerebrospinal fluid samples. Figure 6 ).

[0144] (3) When the excitation wavelength is 310nm, Eu 3+ -MOF, Eu 3+ The spectra of -dopedHOFs show Eu 3+ Strong characteristic fluorescence, with emission peaks at 592, 616, 650 and 698 nm, is Eu 3+ Typical ff transition of Eu 3+ -dopedCOF has almost no fluorescence under excitation due to the 3+ Effectively cooperate to transfer energy to Eu under excitation 3+ , enhancing the Eu in COF 3+ Fluorescence, making Eu 3+ When the excitation wavelength is 310nm, Eu 3+ -dopedCOF has a main emission wavelength of Eu 3+ (616nm). Finally, the strongest emission at 616nm was selected as the fluorescence signal output of the sensor element.

[0145] Example 5 Machine Learning Algorithm to Build Classification Model

[0146] (1) Feature selection and machine learning algorithms were implemented based on the open source Python package scikit-learn. A two-sided 5×2CV f-test was performed using mlextend. The random forest algorithm (RF) was trained using nested stratified k-fold cross-validation and compared with two dummy classifiers (simulating random guessing).

[0147] Test Example 1 Detection and Analysis of Tuberculous Meningitis

[0148] refer to Figure 7In this experimental example, the inventors reacted the sensor element with different types of cerebrospinal fluid (tuberculous meningitis and non-intracranial infection control group) and used a machine learning algorithm to construct a classification model to explore the detection and discrimination ability of the sensor element prepared in the above embodiment for tuberculous meningitis.

[0149] A total of 113 cerebrospinal fluid samples were selected for this experiment, of which 56 were non-intracranial infection cerebrospinal fluid samples and 57 were TBM cerebrospinal fluid samples. The 113 cerebrospinal fluid samples reacted with 10 sensor elements respectively, repeated 6 times, and the fluorescence signal was detected. The data was preprocessed by subtracting the fluorescence value after the reaction (I) from the fluorescence value before the reaction (I0), and the difference between the two values ​​(ΔI=I0-I) was used as the signal for sensor analysis. In order to exclude the interaction between the 10 sensor elements, we calculated the Spearman rank correlation coefficient (ρ) of the 10 sensor elements, and it can be found that the correlation between the sensor elements is low, which indicates that the 10 sensor elements can provide important characteristic information.

[0150] refer to Figure 8 , a machine learning algorithm was used to construct a classification model for TBM and non-intracranial infection based on the sensor analysis signal. In order to select the simplest model with the best performance, the random forest algorithm was finally used to construct the classification model after testing. The training used nested stratified cross-validation, and the error rate between the prediction and the actual was used as the loss function. The classification model was constructed at the lowest point of the loss function. The average prediction accuracy of the classification model constructed by the random forest algorithm for the sensor analysis signal output by 10 sensor elements was 96.6%, of which the prediction accuracy for TBM samples was 97.1%. At the same time, the feature importance analysis of random forest was used to identify the sensor element with the greatest contribution. Finally, the receiver operating characteristic curve (ROC) of TBM and non-intracranial infection was established by the random forest algorithm to evaluate the classification effect of the model. The results showed that the area under the curve (AUC) of the model was 0.98, which showed that the sensor array had good detection accuracy and good discrimination ability. Then, the random forest algorithm was used to construct a classification model for the sensor analysis signals of the first five important features (EM1, EM2, EC, EH1, EH2), and the average prediction accuracy of the classification model was found to be 95.3%, of which the prediction accuracy for TBM samples was 95.9%. By comparing the prediction accuracy of the two models, it can be found that the performance of 10 sensor elements is better than that of 5 sensor elements, which shows that all 10 sensor elements in the sensor array are important.

[0151] In summary, it can be found that the sensor array combined with the random forest algorithm to construct a classification model has the ability to distinguish TBM from non-intracranial infection and has high detection accuracy, which has potential value for clinical detection of TBM.

[0152] Test Example 2 Detection and Analysis of Bacterial Meningitis

[0153] refer to Fig. 9 In this experimental example, the inventors reacted the sensor element with different types of cerebrospinal fluid (bacterial meningitis and non-intracranial infection control group) and used a machine learning algorithm to construct a classification model to explore the detection and discrimination ability of the sensor element prepared in the above embodiment for bacterial meningitis.

[0154] A total of 88 cerebrospinal fluid samples were selected in this experiment, of which 45 were non-intracranial infection cerebrospinal fluid samples and 43 were BM cerebrospinal fluid samples. The 88 cerebrospinal fluid samples reacted with 10 sensor elements respectively, repeated 6 times, and a total of 528 data were obtained by detecting the fluorescence signal. The data were preprocessed and analyzed as above, and the correlation coefficients between the 10 sensor elements were evaluated by the ρ coefficient. It can be found that the correlation between each sensor element group is low, which indicates that each sensor element of the sensor array can provide important feature information. The classification model was constructed using the random forest algorithm, and it can be found that the average prediction accuracy of the model was 97.7%, of which the prediction accuracy for BM samples was 97.3%. Through importance analysis, it can be found that EM4, EM5 and EM7 contribute more.

[0155] In summary, it can be found that the sensor array combined with the random forest algorithm to construct a classification model for BM and non-intracranial infection also has high detection accuracy and the ability to distinguish BM and non-intracranial infection. Experimental Example 3 Detection and Analysis of Cryptococcal Meningitis

[0156] refer to Fig.10 In this experimental example, the inventors reacted the sensor element with different types of cerebrospinal fluid (cryptococcal meningitis and non-intracranial infection control group) and used a machine learning algorithm to explore the detection and discrimination ability of the sensor element prepared in the above embodiment for cryptococcal meningitis.

[0157] A total of 44 cerebrospinal fluid samples were selected in this experiment, of which 21 were non-intracranial infection cerebrospinal fluid samples and 23 were CM cerebrospinal fluid samples. The fluorescence spectrometer was used to record 6 repeated fluorescence changes at an emission wavelength of 616nm. In the fluorescence response box plot, it can be found that CM shows completely different results, suggesting that CM has different fluorescence response patterns from the other three cerebrospinal fluid samples (TBM, BM, non-intracranial infection). In addition, the ρ coefficient also shows that the correlation between the 10 sensor element groups is low. By constructing a classification model using the random forest algorithm, it can be found that the average prediction accuracy of the model is 98.5%, of which the prediction accuracy for CM samples is 98.6%. The significance analysis shows that the sensor elements with the greatest contribution are EM1, EM2, EM3 and EH2, which are also quite different from the sensor elements with the greatest contribution in the above samples.

[0158] In summary, it can be found that the sensor array combined with the random forest algorithm to construct a classification model for CM and non-intracranial infection also has high detection accuracy and the ability to distinguish CM from non-intracranial infection. Experimental Example 4 Detection and Analysis of Viral Meningitis

[0159] refer to Fig.11 In this experimental example, the inventors reacted the sensor element with different types of cerebrospinal fluid (viral meningitis and non-intracranial infection control group) and used a machine learning algorithm to explore the detection and discrimination ability of the sensor element prepared in the above embodiment for viral meningitis.

[0160] A total of 118 cerebrospinal fluid samples were selected in this experiment, including 56 viral meningitis cerebrospinal fluid samples and 62 non-intracranial infection cerebrospinal fluid samples. Data inclusion and preprocessing As shown in the above experiment, after each sensor element reacted with VM cerebrospinal fluid, it was found that the fluorescence recognition pattern of the sensor was significantly different from that before. However, the ρ coefficient showed that the correlation between the 10 sensor elements in the VM group was relatively high, which means that multiple sensor elements provided similar feature information. The classification model was constructed using a random algorithm, and the average prediction accuracy of the model was 97.7%, of which the prediction accuracy for VM samples was 97.3%. In the feature importance analysis, it can be found that EM1, EM2, EM7 and EH2 are the sensor components that contribute the most.

[0161] In summary, it can be found that the sensor array combined with the random forest algorithm to construct a classification model also has the ability to distinguish VM from non-intracranial infection.

[0162] Test Example 5 Differential Detection of Meningitis with Different Etiologies

[0163] refer to Fig.12 In this test example, in order to understand whether the sensor array can be used to specifically distinguish meningitis caused by different pathogenic infections, the inventors further analyzed four types of meningitis. First, the inventors compared the fluorescence response signals of the four pathogenic cerebrospinal fluid samples by linear discriminant analysis (LDA). It can be found that although there are some overlaps, the four pathogenic meningitis cerebrospinal fluid samples still show separation results, which also shows that the sensor array obtains four meningitis-specific signals and can distinguish cerebrospinal fluid samples from different pathogenic infections. This result is also confirmed by constructing a confusion matrix using random forests.

[0164] refer to Fig.13In the confusion matrix result diagram of the classification model, it can be found that the classifier has a high prediction accuracy of VM (99.1%), but a low prediction accuracy of CM, only 92.8%. At the same time, in the importance analysis, it can be found that the sensor elements that contribute most to distinguishing the four types of meningitis are EM1, EM2, EM4 and EM5. The ROC curves of the cerebrospinal fluid samples of the four infection types were established using the random forest algorithm. It can be found that the AUCs of the classification models are 0.92 (BM), 0.95 (CM), 0.96 (VM) and 0.95 (TBM), respectively.

[0165] In summary, it can be found that the sensor array combined with the random forest algorithm to construct a classification model for 4 types of infection types of meningitis has the ability to distinguish 4 different types of infection types of meningitis, among which the prediction and detection accuracy of VM is higher, while the prediction and detection accuracy of CM is slightly lower, but the AUC of the 4 types of infection types of meningitis is greater than 0.9, indicating that the model has good detection and identification capabilities and has potential value for clinical detection and differential detection.

[0166] Fig.14 This is the technical circuit diagram of the above test example.

[0167] On the other hand, an embodiment of the present invention further provides a machine learning-assisted sensor array multiple meningitis data processing system, the system comprising:

[0168] The first module is used to prepare the sensing element; the sensing element is synthesized by metal organic framework, covalent organic framework and hydrogen bond organic framework;

[0169] The second module is used to construct a sensor array according to the sensor elements;

[0170] A third module is used for obtaining a fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid;

[0171] The fourth module is used to obtain meningitis type information based on fluorescence signals and machine learning algorithms.

[0172] An embodiment of the present invention provides a machine learning-assisted cross-reactive sensor array for detecting multiple types of meningitis, the system comprising: ① 10 sensor elements for reacting with cerebrospinal fluid samples and generating specific fluorescence response patterns; ② a machine learning algorithm module for constructing a classification for the generated fluorescence response signals to achieve identification of normal cerebrospinal fluid samples and samples of different types of meningitis.

[0173] Among them, the 10 sensing elements in ① include 7 doped Eu 3+ MOF sensor element, a Eu-doped 3+ COF sensing element and 2 Eu 3+ Doped HOF sensing element.

[0174] The machine learning algorithm in ② mainly refers to the random forest algorithm.

[0175] The detection of fluorescence signals requires an excitation wavelength of 310 nm, and includes a fluorescence signal with a maximum emission wavelength of 616 nm.

[0176] The system of the embodiment of the present invention is used to predict different infection types of meningitis.

[0177] As an optional embodiment, 7 Eu doped 3+ The MOF sensing elements were synthesized via a one-pot method, a post-synthesis method, or a hydrothermal method.

[0178] Eu doping 3+ The COF sensor element was prepared by reacting p-phenylenediamine and 2,6-lutidine with acetic acid in a mixed solution of toluene and 1,4-dioxane, followed by doping with Eu in acetonitrile. 3+ Heat treatment synthesis.

[0179] 2 Eu doped 3+ The HOF sensing element is formed by hydrogen bonding self-assembly and doped with Eu in acetonitrile. 3+ .

[0180] On the other hand, an embodiment of the present invention also provides a device for processing multiple meningitis data using a machine learning-assisted sensor array, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the device implements the method for processing multiple meningitis data using the machine learning-assisted sensor array described above.

[0181] The multiple meningitis data processing device of the machine learning-assisted sensor array of the embodiment of the present invention includes a memory and a processor.

[0182] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0183] The memory may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage. Among them, ROM can store static data or instructions required by the processor or other modules of the computer. The permanent storage device may be a readable and writable storage device. The permanent storage device may be a non-volatile storage device that does not lose the stored instructions and data even after the computer is powered off. In some embodiments, the permanent storage device uses a large-capacity storage device (such as a magnetic or optical disk, flash memory) as a permanent storage device. In some other embodiments, the permanent storage device may be a removable storage device (such as a floppy disk, optical drive). The system memory may be a readable and writable storage device or a volatile readable and writable storage device, such as a dynamic random access memory. The system memory may store some or all instructions and data required by the processor at run time. In addition, the memory may include any combination of computer-readable storage media, including various types of semiconductor memory chips (DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory may include a readable and / or writable removable storage device, such as a laser disc (CD), a read-only digital versatile disc (e.g., DVD-ROM, double-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. The computer-readable storage medium does not include carrier waves and transient electronic signals transmitted wirelessly or wired.

[0184] The memory stores executable codes, and when the executable codes are processed by the processor, the processor can execute part or all of the methods described above.

[0185] On the other hand, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above method.

[0186] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0187] The preferred embodiments of the present application are described above with reference to the accompanying drawings, but the scope of the rights of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present application should be within the scope of the rights of the present application.

Claims

1. A method for processing multiple meningitis data using a machine learning-assisted sensor array, characterized in that: The method comprises the following steps: Obtaining a sensing element; wherein the sensing element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework; According to the sensing element, a sensing array is constructed; obtaining a fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid; According to the fluorescent signal and the machine learning algorithm, meningitis type information is obtained.

2. The method according to claim 1, characterized in that The step of obtaining the sensor element comprises the following steps: synthesize metal-organic frameworks, covalent organic frameworks, and hydrogen-bonded organic frameworks using synthetic methods; The synthesis methods include one-pot method, post-synthesis method and hydrothermal method; In carrying out the synthesis method, a variety of organic ligands and metal salts are used to react in a solvent; The metal organic framework, the covalent organic framework and the hydrogen bond organic framework are placed in a 3+ soaked in a solution to obtain a sensing element.

3. The method according to claim 1, characterized in that The construction of a sensor array according to the sensor element comprises the following steps: A plurality of the sensing elements are used to construct a sensing array.

4. The method according to claim 1, characterized in that: The step of obtaining a fluorescent signal based on the sensor array and the meningitis cerebrospinal fluid comprises the following steps: Obtain meningitis cerebrospinal fluid for several types of infection; Respectively reacting each of the sensing elements in the sensing array with samples of the meningitis cerebrospinal fluid of different infection types to obtain sensing signal information; A fluorescence signal is obtained according to the sensing signal information.

5. The method according to claim 4, characterized in that The step of causing each of the sensor elements in the sensor array to react with samples of the meningitis cerebrospinal fluid of different infection types to obtain sensor signal information comprises the following steps: 20 μL of the meningitis cerebrospinal fluid of different infection types was reacted with 50 μL of each of the sensing elements to obtain sensing signal information.

6. The method according to claim 4, characterized in that The step of obtaining the fluorescence signal according to the sensor signal information comprises the following steps: The sensing signal information is used to obtain a fluorescence response signal generated by the sensing element when the excitation wavelength is 313 nm; According to the fluorescence response signal, Eu 3+ The emission wavelength of the fluorescence signal is 616 nm.

7. The method according to claim 1, characterized in that The method of obtaining meningitis type information according to the fluorescence signal and the machine learning algorithm further comprises the following steps: Processing and analyzing the collected fluorescence signals using a machine learning algorithm to obtain analysis data; Build a classification model; According to the analysis data and the classification model, pattern recognition and feature extraction are performed on the fluorescence signal to obtain meningitis type information.

8. A machine learning-assisted sensor array multi-meningitis data processing system, characterized in that: The system comprises: The first module is used to prepare a sensor element; the sensor element is synthesized by a metal organic framework, a covalent organic framework and a hydrogen bond organic framework; A second module is used to construct a sensor array according to the sensor elements; A third module is used to obtain a fluorescent signal according to the sensor array and the meningitis cerebrospinal fluid; The fourth module is used to obtain meningitis type information based on the fluorescent signal and the machine learning algorithm.

9. A device for processing multiple meningitis data using a machine learning-assisted sensor array, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for processing multiple meningitis data using a machine learning-assisted sensor array as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images