Typing method of schizophrenia neurotransmitter subtype

Abstract

A typing method for schizophrenia neurotransmitter subtypes relates to the technical field of medical treatment, and mainly comprises the following steps: constructing a CTh brain structure development norm by using public available and previously acquired healthy control brain structure MRI data; extracting the cortex thickness CTh value of each brain region of the patient; calculating the CTh deviation of each brain region of the patient; collecting a PET molecular image map of a neurotransmitter receptor or transporter, calculating a correlation coefficient rho and forming a CTh anomaly correlation neurotransmitter mass spectrum; dimensionality reduction is carried out on the CTh anomaly related neurotransmitter mass spectrum; establishing a final schizophrenia neurotransmitter subtype typing model by using a consistency clustering method; and inputting patient data to be typed into the typing model to obtain subtype types. According to the method, the PET neurotransmitter receptor or transporter map and individualized brain structure deviation can be obtained through association disclosure, possible neurotransmitter system abnormality of an SCZ patient individual can be indirectly reflected, and the neurotransmitter related SCZ biological subtype can be identified.

Description

Technical Field

[0001] The invention relates to the field of medical technology, and in particular to a method for typing neurotransmitter subtypes of schizophrenia. Background Art

[0002] Schizophrenia (SCZ) is a severe mental illness that is prone to occur in adolescence and early adulthood and seriously affects an individual's thinking, emotions, cognition, and behavior. Previous case-control studies used magnetic resonance imaging (MRI) technology to find that SCZ patients have extensive structural abnormalities in the brain. However, the group-level comparison method used in the above studies assumed that SCZ patients were a homogeneous group, and the results obtained were inconsistent, suggesting that the SCZ patient group based on the same symptom diagnostic criteria may be composed of multiple subtypes with different neuropathological mechanisms. This biological heterogeneity can be traced back to the heterogeneity of individual neuropathology and genetic risks in SCZ patients, and its existence hinders the development of reliable and effective objective biomarkers related to the diagnosis and treatment of SCZ. Separating SCZ biological subtypes can effectively overcome heterogeneity, promote accurate exploration of the genetic and neuropathological mechanisms of SCZ, and help develop stable and reliable objective biomarkers.

[0003] The heterogeneity of SCZ is also reflected in the treatment response: less than half of SCZ patients can achieve clinical cure after the first antipsychotic treatment, and about 30% of SCZ patients develop refractory SCZ after multiple antipsychotic treatments fail. Antipsychotics targeting the neurotransmitter system are currently the main treatment for SCZ. Existing drugs mainly act on dopamine receptors D2 and 5-hydroxytryptamine receptors 5-HT2a. Differences in drug treatment responses indicate that abnormalities in the neurotransmitter system in SCZ also have heterogeneity. There is no doubt that the detection of abnormalities in the brain of the above neurotransmitter systems can guide drug selection. However, the current molecular imaging detection methods of neurotransmitter systems based on positron emission tomography (PET) have the disadvantages of invasive operation, risk of drug allergy, and high economic burden. In addition, a PET imaging can only detect one neurotransmitter receptor\transporter, but existing evidence shows that in addition to the dopamine and serotonin neurotransmitter systems, there are also abnormalities in multiple neurotransmitter systems in SCZ, including choline, norepinephrine, histamine and γ-aminobutyric acid, and antipsychotic drugs targeting the above neurotransmitter systems are currently under development or in clinical trials. Therefore, we urgently need to develop a feasible and effective method to achieve simultaneous detection of multiple neurotransmitter systems, so as to analyze the heterogeneity related to the SCZ neurotransmitter system and support the development of personalized precision drug treatment plans in the future.

[0004] Previous studies have demonstrated that the neurotransmitter system is associated with brain structure: Zhen et al. found that the covariation network of cortical thickness (CTh) was correlated with the covariation network of the neurotransmitter system; Liao et al. found that the decrease in CTh during neurodevelopment reflects the overall organizational characteristics of the neurotransmitter system in the brain; Liang et al. found that the decrease in CTh in SCZ was associated with the increase in glutamate concentration in the dorsal anterior cingulate region. These research results suggest that in SCZ, abnormal brain structure, especially CTh abnormality, can indirectly reflect abnormal neurotransmitter system. On this basis, we propose the following methodological pipeline for analyzing the heterogeneity of neurotransmitters related to SCZ, that is, firstly to achieve individualized measurement of CTh abnormality in SCZ patients, then to associate the neurotransmitter system with individual CTh abnormality, so as to indirectly measure the abnormality of individual neurotransmitter system in SCZ patients, and on this basis to separate biological subtypes, so as to explore the genetic and neuropathological mechanisms of different subtypes of SCZ from the perspective of multiple neurotransmitter systems, and provide scientific basis and reference for the selection of clinical individualized precision antipsychotic drugs.

[0005] In recent years, the development of imaging transcriptomics technology has brought more possibilities to brain imaging research. By correlating disease imaging findings with publicly available multimodal data, such as brain gene expression, PET neurotransmitter receptor\transporter maps or brain microstructure maps, we can infer the genetic, molecular and cellular tissue characteristics corresponding to abnormal brain structure and function in diseases. Using this method, Park et al. found that brain structural abnormalities shared by multiple mental illnesses are related to the dopamine and 5-hydroxytryptamine neurotransmitter systems. In previous studies, the applicant also found that the brain function damage pattern of SCZ is associated with the brain distribution of D2 and 5-HT2a receptors. Combining the above methods with the technology of constructing brain structure development norms, we can associate brain structural abnormalities with multiple neurotransmitter systems at the individual patient level, identify neurotransmitter-related SCZ biological subtypes based on individualized brain structural deviations of neurotransmitter markers, and analyze SCZ heterogeneity from the perspective of neurotransmitter systems, the core target of SCZ treatment. Based on a similar research idea, Biase et al. successfully separated SCZ biological subtypes from the perspective of brain cell types related to CTh abnormalities.

[0006] In summary, current SCZ research has problems such as large interference from disease heterogeneity, difficulty in detecting multiple neurotransmitter systems in individual patients, and overlap between clinical manifestations and corresponding neuropathologies and genetic mechanisms. Existing research technologies have not solved the problem of measuring multiple neurotransmitters simultaneously, and cannot obtain the abnormal spectrum of multiple neurotransmitter systems of a single individual, making it difficult to use individualized neurotransmitter abnormality information to classify schizophrenia subtypes.

[0007] Previous studies have mainly analyzed the heterogeneity of SCZ disease from two perspectives: symptom subtypes and biological subtypes: Zhang et al. and Nenadic et al. classified SCZ into three symptom subtypes based on clinical symptoms: positive (paranoid) symptom subtype, negative symptom subtype and disintegrative symptom subtype, and found differences in brain structural abnormalities among different subtypes, but the results were not consistent; Xiao et al. classified SCZ into three biological subtypes with different brain structural abnormalities based on brain surface area, volume and CTh, and Pan et al. also classified SCZ into three biological subtypes based on CTh, but did not find differences in clinical symptom manifestations among different biological subtypes. Chai et al. systematically compared the symptom and biological subtype classification methods of SCZ and found that the subtypes obtained by the two methods did not match, such as Figure 1 The above studies show that the same clinical symptoms in SCZ may have very different biological bases. Starting from the biological subtypes, the genetic and neuropathological basis of SCZ heterogeneity can be more accurately analyzed.

[0008] like Figure 2 As shown in the figure: In previous studies, the identification of biological subtypes often directly uses biological indicators of SCZ patients (such as CTh) for classification. Varol et al. first pointed out that this method may introduce factors other than disease effects, such as individual differences, age and gender effects, etc. that also exist in the normal population. Based on this, they proposed the "Heterogeneity through Discriminative Analysis (Hydra)", which introduces normal control information in the subtype identification process and identifies subtypes with pure disease effects. Chand et al. used the Hydra method to divide SCZ into two biological subtypes. This subtype classification method was repeated in multiple independent studies, reflecting the superiority of this research idea. However, it should be noted that the separation process of Hydra for disease effects occurs in the high-dimensional space of the machine learning model. At the individual patient level, the overall disease effect can only be represented by the abstract "signature presence index", which cannot reflect the impact of the disease effect on specific brain regions, which is not conducive to further exploring the neuropathological mechanism of specific subtypes. The introduction of the normative method solves the above problems well. Similar to the common height or weight development norms, by using neuroimaging big data to construct brain structure development norms, we can obtain the normal range of brain structure indicators corresponding to specific age, gender and other variables. Then, by comparing the brain structure indicators of SCZ patients and their corresponding norms, we can individually measure the deviation between the brain structure of each brain region and the normal value, obtain the brain structure deviation map of a single patient, separate the disease effects at the individual level, and use it for subsequent subtype identification.

[0009] Disadvantages of existing technology: (I) Molecular imaging detection methods of neurotransmitter systems based on positron emission tomography (PET) can only detect one neurotransmitter receptor / transporter at a time, but existing evidence shows that in addition to dopamine and 5-hydroxytryptamine neurotransmitter systems, SCZ also has abnormalities in multiple neurotransmitter systems including choline, norepinephrine, histamine and γ-aminobutyric acid, and each neurotransmitter system contains multiple neurotransmitter transporter\receptor subtypes. In addition, PET operation has the problems of invasive operation, risk of drug allergy and high economic burden. These defects make it extremely difficult and unfeasible to use PET technology to measure abnormalities of multiple neurotransmitter systems on the same individual at the same time.

[0010] (II) The process of separating disease effects by the heterogeneity analysis method based on discriminant analysis (Hydra) occurs in the high-dimensional space of the machine learning model. At the individual patient level, it can only characterize the overall disease effect through the abstract "signature presence", which cannot reflect the impact of the disease effect on specific brain regions. This is not conducive to further exploring the neuropathological mechanism of specific subtypes and locating abnormal brain regions at the individual level.

[0011] (III) Traditional clustering methods such as K-means clustering or hierarchical clustering have problems such as high randomness, poor stability and poor repeatability between different data sets. The models obtained using the above clustering methods have the problem of poor transferability. The use of this type of model is not conducive to identifying neurotransmitter-related SCZ subtypes that are stable and robust at different sites or different magnetic resonance imaging instruments. Summary of the invention

[0012] Based on the above problems, the purpose of the present invention is to provide a method for classifying neurotransmitter subtypes of schizophrenia, which can indirectly reflect the possible abnormalities of the neurotransmitter system in individual SCZ patients by correlating publicly available PET neurotransmitter receptor or transporter maps with individualized brain structure deviations. This can be used as a starting point to analyze the heterogeneity of SCZ disease from the perspective of multiple neurotransmitter systems and identify neurotransmitter-related SCZ biological subtypes.

[0013] The technical solution adopted by the present invention to achieve its invention object is a method for typing neurotransmitter subtypes of schizophrenia, comprising the following steps: S1. Using brain structural MRI data of multiple healthy controls, the hierarchical Bayesian regression method (HBR) was used to train the CTh brain structural development norm, which is an HBR model related to the predicted value of CTh cortical thickness in each brain region. In specific implementation, in addition to the HBR method, general linear models, Gaussian process regression, additive models and other techniques can also be used to construct the norms of CTh brain structure development; S2, collecting magnetic resonance data of multiple schizophrenia patients, and then extracting the measured value of cortical thickness CTh of each brain region of each patient; S3. Using the HBR model trained in step S1, calculate the CTh deviation Z value of each brain region of each patient by the following formula:

[0014] Where: y is the actual measured value of CTh of the patient, is the CTh value predicted by the HBR model, σ is the prediction variance, σ n is the internal variance of the HBR model; S4. Collect A kinds of PET molecular imaging spectra of neurotransmitter receptors or transporters and calculate the mean of each brain region, where A is a positive integer, and then use Spearman correlation to calculate the correlation coefficient ρ between each PET molecular imaging spectra and the individualized CTh deviation of the patient, and then the Spearman correlation coefficient ρ of all A kinds of PET molecular imaging spectra and the individualized CTh deviation of the patient is combined into a CTh abnormality-related neurotransmitter spectrum; In specific implementations, Pearson correlation, cosine correlation and other techniques may be used instead of Spearman correlation to correlate individual brain structure deviation maps with neurotransmitter maps, and construct individual CTh abnormality-related neurotransmitter spectra; S5, reducing the dimension of the CTh abnormality-related neurotransmitter spectrum obtained in step S4 from A to 3; S6. Randomly sample the healthy control brain structure MRI data set B times at a set sampling rate, where B is a positive integer, and perform K-means clustering on each randomly selected data set, where the number of clusters ranges from 2 to C, where C is a positive integer greater than 2, and perform consistency clustering on each number of clusters, and then evaluate and determine the optimal number of clusters, i.e., the number of subtypes, and then select the clustering model corresponding to the number of subtypes as the final schizophrenia neurotransmitter subtype classification model; In specific implementations, other clustering methods besides consensus clustering can also be used to achieve subtype identification and model construction; S7. Collect magnetic resonance data of schizophrenia patients to be classified, extract the CTh measured values ​​of each brain region of the patient, and then use the method of step S4 to construct the patient's CTh abnormality-related neurotransmitter spectrum, and then input the final schizophrenia neurotransmitter subtype classification model obtained in step S6 to obtain the patient's neurotransmitter-related subtypes.

[0015] Further, the HBR model in step S1 is as follows:

[0016] Where: Y is the predicted CTh value of each brain region, f represents the HBR model function, X is the input data, including age, gender, MRI scanner type and average cortical thickness of the cerebral hemisphere, θ is the model parameter obtained after training the model, is the model residual.

[0017] Furthermore, in step S1, before formal training, the hierarchical Bayesian regression method is used for pre-training, and then the HBR model is trained using the five-fold cross-validation method. After the performance of the HBR model is determined by cross-validation, formal training is performed.

[0018] Further, the Spearman correlation calculation in step S4 is specifically as follows: First, the CTh deviation value Z of each brain region of each patient and the PET molecular imaging atlas value of the neurotransmitter receptor or transporter are converted into rank values ​​respectively, and then the Spearman correlation coefficient ρ is calculated using the following formula:

[0019] Where: n is the number of samples of each type of data, d is the interpolation of a pair of data ranks; The specific method of converting the two types of data into rank values ​​is to assign values ​​1, 2, 3, ... in ascending order.

[0020] Furthermore, the specific method of dimensionality reduction in step S5 is: using the uniform manifold approximation and projection method, namely the UMAP method, to perform dimensionality reduction.

[0021] Furthermore, the specific method for evaluating and determining the optimal number of clusters, i.e., the number of subtypes, described in step S6 is: evaluating the increase in the fuzzy clustering ratio score, the average silhouette index, and the area under the cumulative distribution function of the consensus matrix, and taking the number of clusters that the three indicators point to the most as the optimal number of clusters, i.e., the number of subtypes.

[0022] Furthermore, the number of brain regions is 68, and the brain regions are divided using the DK (Desikan-Killiany) brain region segmentation template of the prior art.

[0023] Furthermore, A is equal to 21, and the PET molecular imaging atlas of the A neurotransmitter receptors or transporters includes PET molecular imaging atlases of 21 neurotransmitter receptors or transporters of 9 types of neurotransmitter systems, specifically including: Dopamine system: receptor D1, receptor D2, transporter DAT, neuronal dopamine uptake capacity marker FDOPA; Serotonin system: receptor 5-HT1a, receptor 5-HT1b, receptor 5-HT2a, receptor 5-HT4, receptor 5-HT6, transporter 5-HTT; Cannabinoid system: receptor CB1; Opioid system: receptor MOR, receptor KOR; Glutamate system: receptor mGluR5, receptor NMDA; Gamma-aminobutyric acid system: receptor GABAa; Histamine system: receptor H3; Adrenaline system: transporter NET; Cholinergic system: receptor M1, receptor α4β2, transporter VAChT.

[0024] At this time, in step S5, the dimension of the CTh abnormality-related neurotransmitter spectrum is reduced from 21 to 3.

[0025] Further, in step S6, B is equal to 100, that is, the previously collected healthy control brain structure MRI data set is randomly sampled 100 times at the set sampling rate.

[0026] Further, in step S6, C is equal to 8, that is, the number of clusters is set to range from 2 to 8.

[0027] The beneficial effects of the present invention are: (I) Multi-neurotransmitter labeling of brain structural abnormalities in SCZ patients indirectly reflects abnormalities in the neurotransmitter system, and innovatively achieves simultaneous measurement of multiple neurotransmitter systems at the individual level. Abnormalities in the neurotransmitter system are the core neuropathological mechanism and therapeutic target of SCZ, but since PET testing is expensive, invasive, and there is a risk of drug allergy, it is difficult to use PET technology to simultaneously detect multiple neurotransmitter systems in a single project. The present invention discloses the available PET neurotransmitter receptor / transporter maps and individualized brain structure deviations through association, indirectly reflecting possible abnormalities in the neurotransmitter system of individual SCZ patients, and using this as a starting point to analyze the heterogeneity of SCZ disease from the perspective of multiple neurotransmitter systems and identify neurotransmitter-related SCZ biological subtypes.

[0028] (ii) By utilizing the currently available human brain imaging data and previously collected data, a relatively large sample size can be achieved. The HBR method can be used to handle the possible multi-center bias in the data while performing nonlinear modeling of the brain development process. Using the above data and modeling scheme, a good CTh brain structure development norm can be obtained.

[0029] (III) Using the consistency clustering method, repeated sampling and clustering are performed to obtain the most consistent clustering solution, which can effectively improve the stability and repeatability of clustering analysis.

[0030] (IV) By performing consistent clustering of neurotransmitter spectra associated with structural abnormalities at the individual level of subjects and validating them in multiple independent samples, we obtained stable and reliable SCZ-related neurotransmitter-related subtypes, achieving a subtype classification that is closer to the biological basis of SCZ. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of the mismatch between SCZ symptom subtypes and biological subtypes; Figure 2 Schematic diagram of two traditional subtype typing methods and the subtype typing method of the present invention; Figure 3 PET molecular imaging atlas of neurotransmitter receptors or transporters collected for the embodiments of the present invention; Figure 4 The figure is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0033] Figure 4 A specific embodiment of the method for typing neurotransmitter subtypes of schizophrenia according to the present invention is shown, and the steps include: S1. Construct the CTh brain structure development norm, specifically: use at least 2000 publicly available and previously collected healthy control brain structure MRI data, use the "estimate" command in the mature toolkit PCNtoolkit for building brain development norms, choose to use the hierarchical Bayesian regression method (i.e., Hierarchical Bayesian Regression, HBR), and use the five-fold cross-validation method to train the HBR model. After cross-validation determines the performance of the HBR model, use at least 2000 publicly available and previously collected healthy control brain structure MRI data again, use the hierarchical Bayesian regression method to train the HBR model, and obtain the model parameter set θ; the HBR model formula is as follows:

[0034] Where: Y is the predicted CTh value of each brain region, f represents the HBR model function, X is the input data, including age, gender, MRI scanner type and average cortical thickness of the cerebral hemisphere, θ is the model parameter obtained after training the model, is the model residual; CTh represents the cortical thickness; The number of brain regions is 68, and the division of brain regions adopts the DK (Desikan-Killiany) brain region segmentation template of the prior art; S2. Use a 3.0 T MRI scanner to collect high-resolution T1-weighted structural phase MRI data of at least 100 schizophrenia patients. The spatial resolution must be less than 1 mm × 1 mm × 1 mm per voxel. Use the existing dcm2niix toolkit to convert the DICOM format raw image data exported from the MRI scanner into NIFTI format data for analysis. Then use the "recon_all" command in the existing Freesurfer software to process it, and use the "aparc2table" command to extract the cortical thickness (CTh) data of 68 brain regions of each patient, that is, the actual CTh value. S3. Using the HBR model trained in step S1, calculate the CTh deviation, i.e., Z value, of each brain region of each patient using the following formula:

[0035] Where: y is the actual measured value of CTh of the patient, is the CTh value predicted by the HBR model, σ is the prediction variance, σ n is the internal variance of the HBR model, reflecting the internal variation of CTh values ​​in the healthy control group; S4. Collect the existing PET molecular imaging maps of 21 neurotransmitter receptors or transporters in 9 types of neurotransmitter systems, such as Figure 3 As shown, specifically including: Dopamine system: receptor D1, receptor D2, transporter DAT, neuronal dopamine uptake capacity marker FDOPA; Serotonin system: receptor 5-HT1a, receptor 5-HT1b, receptor 5-HT2a, receptor 5-HT4, receptor 5-HT6, transporter 5-HTT; Cannabinoid system: receptor CB1; Opioid system: receptor MOR, receptor KOR; Glutamate system: receptor mGluR5, receptor NMDA; Gamma-aminobutyric acid system: receptor GABAa; Histamine system: receptor H3; Adrenaline system: transporter NET; Cholinergic system: receptor M1, receptor α4β2, transporter VAChT; Each PET molecular imaging atlas was aligned to the standard space and the mean of each brain region was calculated. Then, the correlation coefficient ρ between each PET molecular imaging atlas and the individualized CTh deviation of the patient was calculated using Spearman correlation. In the calculation of the Spearman correlation coefficient, the CTh deviation values ​​of the 68 brain regions of each patient and the PET molecular imaging atlas values ​​of the neurotransmitter receptors or transporters were first converted into rank values, and then the Spearman correlation coefficient was calculated using the following formula:

[0036] Where: ρ is the Spearman correlation coefficient, n is the number of samples in each category, and d is the interpolation of a pair of data ranks; The specific method of converting the two types of data into rank values ​​is to assign 1, 2, 3, ... in ascending order; Then, the PET molecular imaging maps of all 21 neurotransmitter receptors or transporters and the Spearman correlation coefficient ρ of the individual CTh deviation of the patient were combined to form a CTh abnormality-related neurotransmitter spectrum, so as to label the individual CTh deviation map of the patient with neurotransmitters. This vector was used for subsequent cluster analysis to identify neurotransmitter-related SCZ biological subtypes. S5. Using the Uniform Manifold Approximation and Projection (UMAP) method to reduce the dimension of the CTh abnormality-related neurotransmitter spectrum obtained in step S4 from 21 to 3. Dimensional data reduction can improve clustering performance, and the UMAP method is used for dimensionality reduction because UMAP performs better than commonly used component analysis or independent component analysis methods in maintaining the local and global structure of the data. The parameters of UMAP are set to 30 and the minimum distance is set to 0 according to its guidelines. S6. Consensus clustering was performed after dimensionality reduction using the "run_all_consensus_partition_method" command in the Cola package in R. The previously collected healthy control brain structural MRI data set was randomly sampled 100 times at a sampling rate of 80%, and K-means clustering was performed on each randomly selected data set. The number of clusters ranged from 2 to 8, and the consistency clustering method was performed for each number of clusters, that is, after creating multiple clustering schemes, a consistency matrix was constructed to represent the frequency of each pair of data points being assigned to the same group in the 100 clustering schemes, and the clustering scheme under each number of clusters was determined by selecting the maximum frequency; After completing the consistency clustering with the number of clusters ranging from 2 to 8, three indicators are evaluated, including: The higher the Proportion of Ambiguous Clustering (PAC) score, the better; The higher the Mean Silhouette Index, the better; The increase in the Area Under Cumulative Distribution Function of the Consensus Matrix increases the number of clusters before the sharp drop in value. The number of clusters with the most three indicators is determined as the optimal number of clusters, i.e., the number of subtypes. Then, the cluster model with the corresponding number of subtypes is selected as the final neurotransmitter subtype classification model for schizophrenia. S7. Collect magnetic resonance data of schizophrenia patients to be classified, extract the cortical thickness data of each brain region of the patient, i.e., the actual CTh value, and then use the method of step S4 to construct the patient's CTh abnormality-related neurotransmitter spectrum, and then input the final schizophrenia neurotransmitter subtype classification model obtained in step S6 to obtain the neurotransmitter-related subtype type of the schizophrenia patient.

[0037] The above embodiments of the present invention are merely examples for illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes and modifications can be made based on the above description. It is impossible to list all the embodiments here. Any obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for classifying neurotransmitter subtypes of schizophrenia, characterized in that: The steps include: S1. Using brain structural MRI data of multiple healthy controls, the hierarchical Bayesian regression method was used to train the CTh brain structural development norm, i.e., the HBR model related to the predicted value of cortical thickness CTh in each brain region; S2, collecting magnetic resonance data of multiple schizophrenia patients, and then extracting the measured value of cortical thickness CTh of each brain region of each patient; S3. Using the HBR model trained in step S1, calculate the CTh deviation Z value of each brain region of each patient by the following formula: ； Where: y is the patient's actual CTh value, is the CTh value predicted by the HBR model, σ is the prediction variance, σ n is the internal variance of the HBR model; S4. Collect A kinds of PET molecular imaging spectra of neurotransmitter receptors or transporters and calculate the mean of each brain region, where A is a positive integer, and then use Spearman correlation to calculate the correlation coefficient ρ between each PET molecular imaging spectra and the individualized CTh deviation of the patient, and then the Spearman correlation coefficient ρ of all A kinds of PET molecular imaging spectra and the individualized CTh deviation of the patient is combined into a CTh abnormality-related neurotransmitter spectrum; S5, reducing the dimension of the CTh abnormality-related neurotransmitter spectrum obtained in step S4 from A to 3; S6. Randomly sample the healthy control brain structure MRI data set B times at a set sampling rate, where B is a positive integer, and perform K-means clustering on each randomly selected data set, where the number of clusters ranges from 2 to C, where C is a positive integer greater than 2, and perform consistency clustering on each number of clusters, and then evaluate and determine the optimal number of clusters, i.e., the number of subtypes, and then select the clustering model corresponding to the number of subtypes as the final schizophrenia neurotransmitter subtype classification model; S7. Collect magnetic resonance data of schizophrenia patients to be classified, extract the CTh measured values ​​of each brain region of the patient, and then use the method of step S4 to construct the patient's CTh abnormality-related neurotransmitter spectrum, and then input the final schizophrenia neurotransmitter subtype classification model obtained in step S6 to obtain the patient's neurotransmitter-related subtypes.

2. A method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: The HBR model in step S1 is as follows: ； Where: Y is the predicted CTh value of each brain region, f represents the HBR model function, X is the input data, including age, gender, MRI scanner type and average cortical thickness of the cerebral hemisphere, θ is the model parameter obtained after training the model, is the model residual.

3. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: In step S1, before formal training, the hierarchical Bayesian regression method is used for pre-training, and then the HBR model is trained using the five-fold cross-validation method. After the cross-validation determines the performance of the HBR model, formal training is performed.

4. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: The Spearman correlation calculation in step S4 is specifically as follows: First, the CTh deviation value Z of each brain region of each patient and the PET molecular imaging atlas value of the neurotransmitter receptor or transporter are converted into rank values ​​respectively, and then the Spearman correlation coefficient ρ is calculated using the following formula: ； Where: n is the number of samples of each type of data, d is the interpolation of a pair of data ranks; The specific method of converting the two types of data into rank values ​​is to assign values ​​1, 2, 3, ... in ascending order.

5. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: The specific method of dimensionality reduction in step S5 is: using the uniform manifold approximation and projection method, namely the UMAP method, to perform dimensionality reduction.

6. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: The specific method for evaluating and determining the optimal number of clusters, i.e., the number of subtypes, described in step S6 is: evaluating the increase in the fuzzy clustering ratio score, the average silhouette index, and the area under the cumulative distribution function of the consensus matrix, and taking the number of clusters that the three indicators point to the most as the optimal number of clusters, i.e., the number of subtypes.

7. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: The number of brain regions is 68, and the brain regions are divided using the DK brain region segmentation template.

8. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: A is equal to 21, and the PET molecular imaging atlas of the A neurotransmitter receptors or transporters includes PET molecular imaging atlases of 21 neurotransmitter receptors or transporters in 9 neurotransmitter systems, specifically including: Dopamine system: receptor D1, receptor D2, transporter DAT, neuronal dopamine uptake capacity marker FDOPA; Serotonin system: receptor 5-HT1a, receptor 5-HT1b, receptor 5-HT2a, receptor 5-HT4, receptor 5-HT6, transporter 5-HTT; Cannabinoid system: receptor CB1; Opioid system: receptor MOR, receptor KOR; Glutamate system: receptor mGluR5, receptor NMDA; Gamma-aminobutyric acid system: receptor GABAa; Histamine system: receptor H3; Adrenaline system: transporter NET; Cholinergic system: receptor M1, receptor α4β2, transporter VAChT.

9. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: In step S6, B is equal to 100.

10. The method for classifying neurotransmitter subtypes of schizophrenia according to claim 1, characterized in that: In step S6, C is equal to 8.

Images