Serum metabolism fingerprint extraction method and application

Abstract

The invention discloses an extraction method and application of a serum metabolism fingerprint spectrum. According to the method, an iron-cobalt bimetallic oxide material is combined with laser desorption / ionization mass spectrometry, a metabolic fingerprint spectrum in serum is extracted, feature extraction and normalization are carried out, and a chronic obstructive pulmonary disease patient and a healthy person are distinguished in combination with a machine learning algorithm. The iron-cobalt bimetallic oxide material has strong ultraviolet absorption and high desorption / ionization efficiency, and can efficiently absorb mass spectrum laser energy and transfer the mass spectrum laser energy to metabolites to ionize the metabolites. By means of the method, the metabolic fingerprint spectrum of the serum sample can be obtained more rapidly, more simply and more conveniently in a high-throughput mode, rapid in-vitro diagnosis of the chronic obstructive pulmonary disease is successfully achieved in combination with a machine learning algorithm, high sensitivity and high accuracy are achieved, and the method has huge potential in the aspects of large-scale disease diagnosis and screening.

Description

Technical Field

[0001] The present invention relates to a method for extracting a serum metabolic fingerprint and an application thereof, and in particular to extracting a serum metabolic fingerprint using an iron-cobalt bimetallic oxide matrix material and applying the extraction to the diagnosis of chronic obstructive pulmonary disease, belonging to the technical field of biomedical detection. Background Art

[0002] Molecular diagnosis of body fluids provides a new approach to the diagnosis of diseases from the perspective of microscopic molecules. It is non-invasive, simple, economical and rapid. Metabolites are the final products of metabolism. Metabolic fingerprints can record the changes in a large number of metabolites and reveal the occurrence and development of diseases. Analysis of metabolic fingerprints can be used to perform molecular diagnosis of diseases, which is a very promising diagnostic method.

[0003] At present, the detection methods of metabolites mainly include nuclear magnetic resonance, mass spectrometry and chromatography. However, these technologies often require complex sample pretreatment steps or are combined with other technologies, making it difficult to efficiently detect a large number of molecules in actual samples with high throughput. Laser desorption / ionization mass spectrometry (LDI-MS) is a mass spectrometry with simple sample preparation, high-speed analysis and high throughput, and has the ability to process large-scale clinical samples. Recently, some nanomaterials have been reported as LDI-MS matrix materials, which have the advantages of high thermal stability, strong laser absorption and low background interference, and have good desorption / ionization efficiency, and are suitable as matrix materials for metabolic fingerprint extraction. Machine learning (ML) has shown significant effectiveness in dealing with the challenges brought by the surge in complex and massive data sets. If the high-throughput detection capability of LDI-MS is combined with machine learning algorithms, it is expected to achieve rapid detection of diseases and large-scale screening, thereby helping to formulate personalized treatment plans and precision medicine for patients. Summary of the invention

[0004] The purpose of the present invention is to achieve high-throughput analysis and detection of serum metabolic fingerprints. The present invention provides a serum metabolic fingerprint extraction method based on inorganic matrix materials and LDI-MS technology. The method uses iron and cobalt bimetallic oxide materials derived from metal organic framework materials as LDI-MS matrices, and has the advantages of low background interference, high ultraviolet absorption, high thermal stability and high desorption / ionization efficiency. The method does not require pretreatment or separation steps for serum samples, and only requires a trace amount of sample to achieve high-throughput extraction of serum metabolic information. The method is applied to the diagnosis of chronic obstructive pulmonary disease and has high clinical value and good application prospects.

[0005] In order to achieve the above object, the present invention provides a method for extracting serum metabolic profile, comprising the following steps:

[0006] Step 1: preparing a suspension of iron-cobalt bimetallic oxide;

[0007] Step 2: The serum sample is co-crystallized with a suspension of iron-cobalt bimetallic oxides, and then naturally dried and directly subjected to LDI-MS analysis to obtain a serum metabolic profile.

[0008] In some embodiments of the present invention, the method for preparing the iron-cobalt bimetallic oxide used in step 1 comprises the following steps:

[0009] Step 1.1: dissolving a divalent cobalt salt and a trivalent iron salt in a reaction solvent, stirring and ultrasonicating them fully to obtain a reaction solution A; dissolving an organic ligand in a reaction solvent to obtain a reaction solution B; adding the reaction solution A to the reaction solution B, stirring and ultrasonicating them fully to obtain a mixed reaction solution;

[0010] Step 1.2: transferring the mixed reaction solution obtained in step 1.1 to a reaction vessel for solvothermal reaction, cooling to room temperature after the reaction is completed, collecting the solid product, washing it, and drying it to obtain an iron-cobalt bimetallic organic framework material;

[0011] Step 1.3: The iron-cobalt bimetallic organic framework material obtained in step 1.2 is calcined in air. After the reaction is completed, it is cooled to room temperature and the product is collected to obtain an iron-cobalt bimetallic oxide.

[0012] In some embodiments of the present invention, the method for preparing the suspension in step 1 comprises: uniformly dispersing the iron-cobalt bimetallic oxide in deionized water to obtain a suspension of the iron-cobalt bimetallic oxide.

[0013] Furthermore, the divalent cobalt salt in step 1.1 is cobalt chloride and / or cobalt chloride hydrate, the trivalent iron salt is ferric chloride and / or ferric chloride hydrate; the reaction solvent is an organic solvent; and the organic ligand is 2-aminoterephthalic acid.

[0014] Furthermore, the organic solvent is N,N-dimethylformamide.

[0015] In some embodiments of the present invention, the molar ratio of the divalent cobalt salt, the trivalent iron salt and the 2-aminoterephthalic acid is 1:0.8-1.2:1-2.

[0016] In some embodiments of the present invention, the reaction temperature in step 1.2 is 100-150° C. and the reaction time is 5-15 hours.

[0017] In some embodiments of the present invention, the drying conditions in step 1.2 are: vacuum drying at 40-60°C.

[0018] In some embodiments of the present invention, the calcination temperature in step 1.3 is 200 to 600° C. and the calcination time is 1 to 3 hours.

[0019] In some embodiments of the present invention, the concentration of the suspension in step 1 is 0.5-1 mg / mL.

[0020] In some embodiments of the present invention, step 2 specifically includes: diluting the serum sample with physiological saline, dripping it on the array point of the stainless steel target plate, and naturally drying it, then dripping a suspension of iron-cobalt bimetallic oxide on the serum sample point and drying it naturally.

[0021] In some embodiments of the present invention, the specific conditions of the LDI-MS analysis in step 2 are: using a 355nm Nd:YAG laser light source, a laser frequency of 2000Hz, an acceleration voltage of 20kV, wherein the voltage at ion source 1 is 20kV, and the voltage at ion source 2 is 17.6kV; the acquisition mode is a reflection cation mode, and the acquisition range m / z is 100-1000Da; each sample is tested 5 times; mass spectrometry data is obtained from flexControl 3.4, and the data is exported in flexAnalysis 3.4.

[0022] The present invention also provides an application of an iron-cobalt bimetallic oxide in preparing a product for diagnosing chronic obstructive pulmonary disease. The preparation method of the iron-cobalt bimetallic oxide comprises the following steps:

[0023] Step 1.1: dissolving a divalent cobalt salt and a trivalent iron salt in a reaction solvent, stirring and ultrasonicating them fully to obtain a reaction solution A; dissolving an organic ligand in a reaction solvent to obtain a reaction solution B; adding the reaction solution A to the reaction solution B, stirring and ultrasonicating them fully to obtain a mixed reaction solution;

[0024] Step 1.2: transferring the mixed reaction solution obtained in step 1.1 to a reaction vessel for solvothermal reaction, cooling to room temperature after the reaction is completed, collecting the solid product, washing it, and drying it to obtain an iron-cobalt bimetallic organic framework material;

[0025] Step 1.3: The iron-cobalt bimetallic organic framework material obtained in step 1.2 is calcined in air. After the reaction is completed, it is cooled to room temperature and the product is collected to obtain an iron-cobalt bimetallic oxide.

[0026] In some embodiments of the present invention, the application performs LDI-MS analysis based on batch serum samples to obtain a serum metabolic fingerprint, and performs a machine learning algorithm based on the serum metabolic fingerprint to obtain a probability score.

[0027] In some embodiments of the present invention, the specific conditions for obtaining the serum metabolic fingerprint are: preprocessing the mass spectra obtained by LDI-MS analysis of batch serum samples, wherein the preprocessing includes peak intensity normalization, peak smoothing, peak averaging, baseline correction and peak extraction, so as to obtain the serum metabolic fingerprint; the batch serum samples include a training set and a test set.

[0028] In some embodiments of the present invention, the machine learning algorithm steps include: inputting training set sample data, training the model, evaluating the model, inputting test set sample data for blind testing, and evaluating the test set blind test results.

[0029] In some embodiments of the present invention, the selected machine learning algorithm is orthogonal partial least squares discriminant analysis.

[0030] The present invention also provides a chronic obstructive pulmonary disease diagnosis system, comprising a laser desorption / ionization mass spectrometer for analyzing serum metabolic information and an inorganic nanomatrix for assisting serum desorption / ionization, wherein the inorganic nanomatrix is ​​an iron-cobalt bimetallic oxide;

[0031] The laser desorption / ionization mass spectrometer is used to obtain the metabolic fingerprint of the subject's serum sample, and analyze the metabolic fingerprint in combination with a machine learning algorithm to discriminate and classify the sample to determine whether the subject suffers from chronic obstructive pulmonary disease. The machine learning algorithm is orthogonal partial least squares discriminant analysis.

[0032] In some embodiments of the present invention, a diagnostic device is further included, which is provided with a machine learning module and a diagnostic module; the machine learning module is used to train a model with training set sample data, and input the sample data to be tested into the trained model to obtain the probability score of the sample to be tested; the diagnostic module is used to output a diagnosis result of whether the patient suffers from chronic obstructive pulmonary disease based on the probability score obtained by the machine learning module.

[0033] Preferably, the machine learning algorithm used by the machine learning module is orthogonal partial least squares discriminant analysis.

[0034] The iron-cobalt bimetallic oxide material prepared by the present invention has strong ultraviolet absorption and high desorption / ionization efficiency, can efficiently absorb mass spectrometry laser energy and transfer it to metabolites for ionization, successfully realizes the extraction of serum metabolic fingerprints, and combines with machine learning algorithms (orthogonal partial least squares discriminant analysis) to successfully realize rapid in vitro diagnosis of chronic obstructive pulmonary disease with high sensitivity and high accuracy. Therefore, the method of the present invention has great potential in large-scale disease diagnosis and screening.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] (1) The serum metabolic fingerprint extraction method provided by the present invention uses the iron and cobalt bimetallic oxide materials derived from metal organic framework materials as LDI-MS matrix for LDI-MS analysis, which can realize efficient high-throughput metabolic detection and analysis of trace samples;

[0037] (2) The iron-cobalt bimetallic oxide material prepared by the present invention has strong ultraviolet absorption and high desorption / ionization efficiency, and can efficiently absorb mass spectrometry laser energy and transfer it to metabolites to ionize them, thereby generating high-intensity mass spectrometry signals;

[0038] (3) The present invention uses a machine learning algorithm to analyze the differences in serum metabolism between COPD patients and healthy controls, successfully distinguishes the serum samples of the two, and realizes the molecular diagnosis of COPD. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a transmission electron microscope photo of the iron-cobalt bimetallic oxide material prepared in Example 1;

[0040] Figure 2 This is a scanning electron microscope photograph of the iron-cobalt bimetallic oxide material prepared in Example 1;

[0041] Figure 3 This is a UV-visible absorption spectrum of the iron-cobalt bimetallic oxide material prepared in Example 1;

[0042] Figure 4 This is a sample of the mass spectrum obtained by detection in Example 2;

[0043] Figure 5 is the receiver characteristic curve based on the test set in Example 3;

[0044] Figure 6 The metabolic levels of potential biomarkers screened using the machine learning algorithm in Example 4, wherein COPD represents chronic obstructive pulmonary disease samples, and HC represents healthy samples;

[0045] Figure 7 is the receiver characteristic curve based on the training set in Example 4;

[0046] Figure 8 This is the receiver characteristic curve based on the test set in Example 4. DETAILED DESCRIPTION

[0047] In order to make the present invention more clearly understood, preferred embodiments are described in detail below with reference to the accompanying drawings.

[0048] In the following examples, the materials, reagents, etc. used are conventional commercially available products unless otherwise specified.

[0049] Example 1: Preparation of a suspension of iron-cobalt bimetallic oxide material

[0050] Step (1): 0.48 g of cobalt chloride hexahydrate and 0.49 g of ferric chloride hexahydrate were dissolved in 20 mL of N,N-dimethylformamide, and the mixture was stirred and ultrasonicated to obtain a reaction solution A; 0.54 g of 2-aminoterephthalic acid was dissolved in 20 mL of N,N-dimethylformamide to obtain a reaction solution B. The reaction solution A was added dropwise to the reaction solution B, and the mixture was stirred and ultrasonicated to obtain a reaction solution B.

[0051] Step (2): The mixed reaction solution obtained in step (1) is transferred to a reactor, and the reaction is carried out at 120° C. for 8 hours. After the reaction is completed, the reactor is cooled to room temperature, and the solid product is collected and washed with deionized water, anhydrous ethanol and N,N-dimethylformamide respectively, and vacuum dried at 50° C. to obtain an iron-cobalt bimetallic organic framework material.

[0052] Step (3): placing the iron-cobalt bimetallic organic framework material obtained in step (2) in a crucible, calcining it in air at 300° C. for 2 hours, with a heating rate of 2° C. per minute. After the reaction is completed, the crucible is cooled to room temperature, and the product is collected to obtain an iron-cobalt bimetallic oxide material.

[0053] Step (4): Evenly disperse the iron-cobalt bimetallic oxide material obtained in step (3) in deionized water to obtain a suspension of the iron-cobalt bimetallic oxide material.

[0054] The transmission electron microscope photo of the iron-cobalt bimetallic oxide material prepared as above is as follows: Figure 1 As shown; the scanning electron microscope photo of the iron-cobalt bimetallic oxide material is shown Figure 2 As shown; the UV-visible absorption spectrum of the iron-cobalt bimetallic oxide material is as shown Figure 3 shown.

[0055] Analysis results: From Figure 1 and Figure 2 It can be seen that the iron-cobalt bimetallic oxide material presents a clear spindle-like structure, about 1.5 microns long; Figure 3 It can be seen that the iron-cobalt bimetallic oxide material has a strong absorption intensity at the wavelength of the laser light source.

[0056] Example 2: The iron-cobalt bimetallic oxide material suspension obtained in Example 1 was used as an LDI-MS matrix material for the extraction and analysis of metabolic fingerprints of 102 COPD sera and 100 healthy human sera

[0057] (1) 102 COPD sera and 100 healthy subjects sera were randomly divided into training and test sets in a ratio of 4:1, corresponding to 162 and 40 samples, respectively.

[0058] (2) Dilute the sample 20 times with physiological saline, and drip 1.5 μL of each sample onto the stainless steel target plate array point of the LDI-MS instrument and dry naturally. Drip 1.5 μL of the iron-cobalt bimetallic oxide material suspension obtained in Example 1 onto the dried sample point and dry naturally.

[0059] (3) Mass spectrometry analysis: Place the stainless steel target plate loaded with dried sample points in step (2) into the LDI-MS instrument for analysis. The mass spectrometer used is Bruker ultrafleXtreme, which adopts a 355nm Nd:YAG laser light source, a laser frequency of 2000Hz, an acceleration voltage of 20kV, wherein the voltage at ion source 1 is 20kV, and the voltage at ion source 2 is 17.6kV; the acquisition mode is reflective cation mode, and the acquisition range m / z is 100-1000Da; each sample is tested 5 times; the mass spectrometry data is obtained from flexControl 3.4, and the data is exported in flexAnalysis 3.4. The obtained mass spectrum pattern is, for example, Figure 4 shown.

[0060] Analysis results: From Figure 4 It can be seen that the serum metabolic fingerprint obtained with the assistance of the iron-cobalt bimetallic oxide material matrix contains rich information and can be used for further metabolic analysis.

[0061] Example 3: The serum metabolic fingerprint obtained in Example 2 was analyzed using a machine learning algorithm (orthogonal partial least squares discriminant analysis) and used for the diagnosis of chronic obstructive pulmonary disease

[0062] (1) Use the MALDIquant package in R software to process the raw mass spectrometry data and export the mass spectrometry peak intensity table.

[0063] (2) The mass spectrometry peak intensity table data is used to train a machine learning model (orthogonal partial least squares discriminant analysis).

[0064] (3) Use the trained model to perform discriminant analysis on the samples, and the model outputs the discriminant classification results of the samples.

[0065] (4) The receiver characteristic curve was plotted based on the discriminant analysis results, and the area under the curve (AUC) was calculated. Figure 5 shown.

[0066] Analysis results: From Figure 5 It can be seen that the classification model obtained by the machine learning algorithm can better distinguish the metabolic profiles of COPD samples and healthy samples, and is feasible for molecular diagnosis and range screening of COPD in clinical practice.

[0067] Example 4: The serum metabolic profile obtained in Example 2 was used to extract a potential disease biomarker panel using a feature selection algorithm, and the biomarker panel was combined with a machine learning algorithm (orthogonal partial least squares discriminant analysis) for the diagnosis of chronic obstructive pulmonary disease

[0068] (1) Use the MALDIquant package in R software to process the raw mass spectrometry data and export the mass spectrometry peak intensity table.

[0069] (2) The mass spectrum peak intensity table is divided into a training set and a test set according to the grouping method in Example 2, and the training set data is used for feature selection.

[0070] (3) Use the relevant indicators given by the machine learning algorithm (orthogonal partial least squares discriminant analysis) to screen important features as potential disease biomarkers. The metabolic levels of potential markers are as follows Figure 6 As shown, COPD represents chronic obstructive pulmonary disease samples and HC represents healthy samples.

[0071] (4) An orthogonal partial least squares discriminant analysis model is trained based on the important features of the training set. The trained orthogonal partial least squares discriminant analysis model is used to predict the test set data. The model gives the discriminant classification results for all samples.

[0072] (5) Use the classification results of the training set and test set samples to draw the receiver characteristic curve and calculate the area under the curve (AUC). The receiver characteristic curves of the training set and test set samples are as follows: Figure 7 and Figure 8 shown.

[0073] Analysis results: From Figure 6 It can be seen that the expression levels of the selected markers in the disease and healthy groups are significantly different and have the potential to be used in the diagnosis and analysis of diseases. Figure 7 It can be seen that the machine learning model established based on the screened markers can better classify disease and healthy samples. Figure 8 It can be seen that the established machine learning model can also classify unknown new sample sets more accurately, has strong diagnostic capabilities, and has the potential to be used as a new method for COPD in clinical practice.

[0074] The above description is only a preferred embodiment of the present invention and is not any formal or substantial limitation of the present invention. It should be pointed out that a person skilled in the art can make several improvements and supplements without departing from the present invention, and these improvements and supplements should also be regarded as the protection scope of the present invention.

Claims

1. A method for extracting serum metabolic profiles, characterized in that: The steps include: Step 1: preparing a suspension of iron-cobalt bimetallic oxide; Step 2: The serum sample is co-crystallized with a suspension of iron-cobalt bimetallic oxides, and then naturally dried and directly subjected to LDI-MS analysis to obtain a serum metabolic profile.

2. The extraction method according to claim 1, characterized in that The preparation method of the iron-cobalt bimetallic oxide used in step 1 comprises the following steps: Step 1.1: dissolving a divalent cobalt salt and a trivalent iron salt in a reaction solvent, stirring and ultrasonicating them fully to obtain a reaction solution A; dissolving an organic ligand in a reaction solvent to obtain a reaction solution B; adding the reaction solution A to the reaction solution B, stirring and ultrasonicating them fully to obtain a mixed reaction solution; Step 1.2: transferring the mixed reaction solution obtained in step 1.1 to a reaction vessel for solvothermal reaction, cooling to room temperature after the reaction is completed, collecting the solid product, washing it, and drying it to obtain an iron-cobalt bimetallic organic framework material; Step 1.3: calcining the iron-cobalt bimetallic organic framework material obtained in step 1.2 in air, cooling to room temperature after the reaction is completed, and collecting the product to obtain an iron-cobalt bimetallic oxide; And / or, the preparation method of the suspension in step 1 includes: uniformly dispersing the iron-cobalt bimetallic oxide in deionized water to obtain a suspension of the iron-cobalt bimetallic oxide.

3. The extraction method according to claim 2, characterized in that The divalent cobalt salt in step 1.1 is cobalt chloride and / or cobalt chloride hydrate, the trivalent iron salt is ferric chloride and / or ferric chloride hydrate; the reaction solvent is an organic solvent; and the organic ligand is 2-aminoterephthalic acid.

4. The extraction method according to claim 3, characterized in that The molar ratio of the divalent cobalt salt, the trivalent iron salt and the 2-aminoterephthalic acid is 1:0.8-1.2:1-2.

5. The extraction method according to claim 2, characterized in that The reaction temperature in step 1.2 is 100-150° C. and the reaction time is 5-15 hours; And / or, the drying conditions in step 1.2 are: vacuum drying at 40-60°C; And / or, the calcination temperature in step 1.3 is 200-600° C. and the calcination time is 1-3 hours.

6. The extraction method according to claim 1, characterized in that The concentration of the suspension in step 1 is 0.5-1 mg / mL; And / or, step 2 specifically includes: diluting the serum sample with physiological saline, dripping it on the array point of the stainless steel target plate, and naturally drying it, then dripping the suspension of iron-cobalt bimetallic oxide on the serum sample point, and naturally drying it.

7. The extraction method according to claim 1, characterized in that The specific conditions of the LDI-MS analysis in step 2 are as follows: using a 355nm Nd:YAG laser light source, a laser frequency of 2000Hz, an acceleration voltage of 20kV, wherein the voltage at ion source 1 is 20kV, and the voltage at ion source 2 is 17.6kV; the acquisition mode is a reflection positive ion mode, and the acquisition range m / z is 100-1000Da; each sample is tested 5 times; mass spectrometry data is obtained from flexControl 3.4, and the data is exported in flexAnalysis 3.

4.

8. Use of an iron-cobalt bimetallic oxide in the preparation of a product for diagnosing chronic obstructive pulmonary disease, characterized in that: The preparation method of the iron-cobalt bimetallic oxide comprises the following steps: Step 1.1: dissolving a divalent cobalt salt and a trivalent iron salt in a reaction solvent, stirring and ultrasonicating them fully to obtain a reaction solution A; dissolving an organic ligand in a reaction solvent to obtain a reaction solution B; adding the reaction solution A to the reaction solution B, stirring and ultrasonicating them fully to obtain a mixed reaction solution; Step 1.2: transferring the mixed reaction solution obtained in step 1.1 to a reaction vessel for solvothermal reaction, cooling to room temperature after the reaction is completed, collecting the solid product, washing it, and drying it to obtain an iron-cobalt bimetallic organic framework material; Step 1.3: The iron-cobalt bimetallic organic framework material obtained in step 1.2 is calcined in air. After the reaction is completed, it is cooled to room temperature and the product is collected to obtain an iron-cobalt bimetallic oxide.

9. The use according to claim 8, characterized in that: The application performs LDI-MS analysis on batch serum samples to obtain a serum metabolic fingerprint, and performs a machine learning algorithm based on the serum metabolic fingerprint to obtain a probability score.

10. The use according to claim 9, characterized in that: The specific conditions for obtaining the serum metabolic fingerprint are: pre-processing the mass spectra obtained by LDI-MS analysis of the batch serum samples, wherein the pre-processing includes peak intensity normalization, peak smoothing, peak averaging, baseline correction and peak extraction, so as to obtain the serum metabolic fingerprint; the batch serum samples include a training set and a test set; And / or, the machine learning algorithm steps include: inputting training set sample data, training the model, evaluating the model, inputting test set sample data for blind testing, and evaluating the test set blind testing results.

11. A chronic obstructive pulmonary disease diagnosis system, characterized in that: It includes a laser desorption / ionization mass spectrometer for analyzing serum metabolic information and an inorganic nano-matrix for assisting serum desorption / ionization, wherein the inorganic nano-matrix is ​​an iron-cobalt bimetallic oxide; The laser desorption / ionization mass spectrometer is used to obtain the metabolic fingerprint of the serum sample of the subject, and analyze the metabolic fingerprint in combination with a machine learning algorithm to classify the sample and determine whether the subject suffers from chronic obstructive pulmonary disease.

12. The chronic obstructive pulmonary disease diagnostic system according to claim 11, characterized in that: It also includes a diagnostic device, which is provided with a machine learning module and a diagnostic module; the machine learning module is used to train a model with training set sample data, and input the sample data to be tested into the trained model to obtain the probability score of the sample to be tested; the diagnostic module is used to output a diagnostic result according to the probability score obtained by the machine learning module.

Images