阅读说明：本技术 抑郁症症状预测变量的选择方法、计算机设备及存储介质 (Method for selecting depression symptom predictive variable, computer device and storage medium ) 是由 马小红 杨潇 赵连生 王敏 杜玥 于 2021-08-24 设计创作，主要内容包括：本发明属于抑郁症诊断技术领域,具体涉及一种抑郁症症状预测变量的选择方法、计算机设备及存储介质。本发明的方法包括如下步骤：对抑郁症患者和正常对照进行考察,采集一般资料、汉密尔顿抑郁量表评分和静息态功能磁共振成像扫描数据；在抑郁症患者接受治疗且个体临床症状缓解后,再次采集上述资料；对静息态功能磁共振成像扫描数据进行处理得到DC大脑图；根据得到的DC大脑图,分析抑郁症患者治疗前后大脑功能活动的变化,找到能够表征抑郁症症状缓解的变量。本发明进一步提供了实现上述方法的计算机设备。本发明能够为抑郁症的早期临床疗效预测提供客观支持依据,降低疾病的社会负担,具有很好的应用前景。(The invention belongs to the technical field of depression diagnosis, and particularly relates to a method for selecting a depression symptom predictive variable, computer equipment and a storage medium. The method of the invention comprises the following steps: inspecting depression patients and normal controls, and collecting general data, Hamilton depression scale scores and resting state functional magnetic resonance imaging scanning data; collecting the data again after the depression patients are treated and the individual clinical symptoms are relieved; processing resting state functional magnetic resonance imaging scanning data to obtain a DC brain map; and analyzing the change of the brain function activity of the depression patient before and after treatment according to the obtained DC cerebral graph, and finding out a variable capable of representing the relief of the depression symptom. The invention further provides computer equipment for implementing the method. The invention can provide objective support basis for early clinical curative effect prediction of depression, reduces social burden of diseases and has good application prospect.)

1. A method for selecting a predictive variable for a symptom of depression comprising the steps of:

step 1, baseline assessment: inspecting depression patients and normal controls, and collecting general data, Hamilton depression scale scores and resting state functional magnetic resonance imaging scanning data; the general data includes age, gender, and educational age;

step 2, follow-up assessment: collecting Hamilton depression scale scores and resting state functional magnetic resonance imaging scanning data for the depression patients and normal controls again after the depression patients receive treatment and the individual clinical symptoms are relieved;

step 3, processing the rest state functional magnetic resonance imaging scanning data acquired in the step 1 and the step 2 to obtain a DC brain image, acquiring a baseline DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 1, and acquiring a follow-up DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 2;

and 4, analyzing the change of the brain function activity of the depression patient before and after treatment according to the DC cerebral graph obtained in the step 3, and finding out a variable capable of representing the relief of the depression symptom.

2. The method of selecting a predictive variable for symptoms of depression according to claim 1, wherein: in step 2, the method for confirming the individual clinical symptom relief of the depression patients is to evaluate the depression patients by using a Hamilton depression scale after the depression patients receive treatment, and judge that the individual clinical symptom relief is achieved if the score of the Hamilton depression scale is less than 7.

3. The method of selecting a predictive variable for symptoms of depression according to claim 2, wherein: in step 2, patients with depression were evaluated using the hamilton depression scale at 8 weeks, 24 weeks, and 48 weeks after receiving treatment.

4. The method of selecting a predictive variable for symptoms of depression according to claim 1, wherein: in step 3, the step of obtaining the DC brain map by using the resting state functional magnetic resonance imaging scanning data comprises the following steps:

step 3A, preprocessing the resting state functional magnetic resonance imaging scanning data;

step 3B, degree center index analysis: calculating Pearson correlation r between the blood oxygen dependent signal time sequences of each pair of voxels by utilizing the preprocessed resting state functional magnetic resonance imaging scanning data to obtain a functional connection matrix covering the whole brain;

step 3C, measuring the weight of the functional connection in the functional connection matrix by adopting a threshold value method, and converting the functional connection matrix into a binary matrix according to the threshold value r being greater than 0.25;

step 3D, calculating the connectivity D of each voxel according to the binary matrix obtained in the step 3C;

and 3E, performing Z conversion on the connectivity D of each voxel obtained in the step 3D to obtain a DC cerebral graph.

5. The method of selecting a predictive variable for symptoms of depression according to claim 4, wherein: in step 3A, the pretreatment comprises: removing at least one of the first 10 time point data, the layer time correction, the head movement correction estimation, the linear trend in the removed signal and the low-pass filtering of 0.01-0.08 Hz;

and/or, in step 3E, the obtained DC brain map is also normalized by performing 6mm full-width half-height gaussian smoothing on the DC brain map.

6. The method of selecting a predictive variable for symptoms of depression according to claim 1, wherein: the step 4 specifically comprises the following steps:

step 4A, counting the DC brain graph obtained in the step 3, and taking a brain area with changed time point specificity function of the depression patient as a central node;

step 4B, analyzing the functional connection indexes of the central node and each area of the brain, and analyzing the change of the functional connection indexes in a baseline DC cerebral graph and a follow-up DC cerebral graph of the depression patient;

step 4C, calculating the HAMD reduction rate of the depression patient before and after treatment, wherein the calculation formula is as follows:

HAMD score ═ baseline score-follow up score ]/baseline score × 100%;

wherein the baseline score is the HAMD score assessed using the hamilton depression scale in step 1 and the follow-up score is the HAMD score assessed using the hamilton depression scale in step 2;

and 4D, performing linear regression model analysis on the change of the functional connection index obtained in the step 4B and the HAMD reduction rate obtained in the step 4C to obtain a functional connection index related to the HAMD reduction rate, namely a variable capable of representing depression symptom relief.

7. The method of selecting a predictive variable for symptoms of depression according to claim 6, wherein: the method for determining the brain area with the time-point specific function change of the depression patient comprises the following steps:

step 4Aa, counting the DC brain image obtained in the step 3 by adopting a linear model, wherein the linear model takes diagnosis multiplied by time point as an independent variable and takes general data acquired in the step 1 as a covariate; in independent variables, the diagnosis refers to a depressed patient or a normal control, and the time point refers to a baseline DC brain map or a follow-up DC brain map;

and step 4Ab, performing simple effect analysis on the areas where the significant interaction is found in the statistical result of the step 4Aa, wherein the simple effect analysis comprises the following steps: baseline DC profile for depression patients vs follow-up DC profile for depression patients, baseline DC profile for normal controls vs follow-up DC profile for normal controls, baseline DC profile for depression patients vs normal controls, follow-up DC profile for depression patients vs normal controls; identifying as a region specific for the time point change a region where the baseline DC profile of the normal control vs the follow-up DC profile of the normal control shows a significant functional change; after excluding the areas specific for the time point changes, the areas of the baseline DC profile of the depressed patients versus the follow-up DC profile of the depressed patients showing significant functional changes were identified as brain areas of time-specific functional changes in the depressed patients.

8. A computer device for selection of a depressive symptom predictive variable, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements a method of selecting a depressive symptom predictive variable according to any of claims 1-7.

9. A computer-readable storage medium characterized by: stored thereon a computer program for implementing a method for selecting a predictive variable for symptoms of depression according to any one of claims 1 to 7.

Technical Field

The invention belongs to the technical field of depression diagnosis, and particularly relates to a method for selecting a depression symptom predictive variable, computer equipment and a storage medium.

Background

Depression is the most common depressive disorder, with significant and persistent mood swings as the primary clinical feature, the major type of mood disorder. Each episode lasts at least 2 weeks, more than long, or even years, and most cases have a tendency to have recurrent episodes, most of which can be alleviated, and some of which can have residual symptoms or become chronic.

The treatment of depression mainly comprises methods such as drug therapy, psychological therapy, physical therapy and the like. The early prediction of the treatment effect is carried out in advance after the treatment, which is helpful for the selection and adjustment of treatment strategies by doctors, improves the treatment effect and reduces the social burden of diseases.

Resting functional magnetic resonance imaging (rs-fMRI) is a method for studying functional connections or networks within the brain. In the past decade, studies of patients with depression using resting functional magnetic resonance imaging have found that abnormalities in some brain regions are important in the mood management and regulation of depression, and that abnormal functional connections in the brain may be regulated by antidepressant therapy.

In the prior art, the study of the relationship between depression and brain region abnormality by resting state functional magnetic resonance imaging is mostly carried out by selecting several regions or networks of interest in advance for further study by using seed-based analysis (JAMA pathology 70, 373-382; neuropsychology: the official publication of the American College of neuropsychology 30, 1334-1344). However, for depression, the neural targets at which abnormalities occur are not known. Also, due to the complex etiology of depression, these aberrant neural targets may vary from individual to individual or over the course of treatment. Therefore, it is difficult to accurately correlate depression with abnormalities in the brain region with the above-mentioned methods in the prior art, and further, it is impossible to predict the development of the depression in the latter stage by the resting-state functional magnetic resonance imaging method.

Disclosure of Invention

Aiming at the difficulties in the prior art, the invention provides a method for selecting a depression symptom predictive variable, a computer device and a storage medium, aiming at early predicting the treatment effect and the disease development of depression patients.

A method of selecting a predictive variable for a symptom of depression comprising the steps of:

step 1, inspecting depression patients and normal controls, and collecting general data, Hamilton depression scale scores and resting state functional magnetic resonance imaging scanning data; the general data includes age, gender, and educational age;

step 2, collecting Hamilton depression scale scores and resting state function magnetic resonance imaging scanning data for depression patients and normal controls again after the depression patients receive treatment and the individual clinical symptoms are relieved;

step 3, processing the rest state functional magnetic resonance imaging scanning data acquired in the step 1 and the step 2 to obtain a DC brain image, acquiring a baseline DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 1, and acquiring a follow-up DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 2;

and 4, analyzing the change of the brain function activity of the depression patient before and after treatment according to the DC cerebral graph obtained in the step 3, and finding out a variable capable of representing the relief of the depression symptom.

Preferably, in step 2, the method for confirming the individual clinical symptom relief of the depression patients is to evaluate the depression patients after receiving treatment by using a Hamilton depression scale, and the individual clinical symptom relief is judged if the Hamilton depression scale score is less than 7.

Preferably, in step 2, patients with depression are evaluated on the hamilton depression scale at 8 weeks, 24 weeks and 48 weeks after receiving treatment.

Preferably, in step 3, the step of obtaining the DC brain map using the resting-state functional magnetic resonance imaging scan data includes:

step 3A, preprocessing the resting state functional magnetic resonance imaging scanning data;

step 3B, degree center index analysis: calculating Pearson correlation r between the blood oxygen dependent signal time sequences of each pair of voxels by utilizing the preprocessed resting state functional magnetic resonance imaging scanning data to obtain a functional connection matrix covering the whole brain;

step 3C, measuring the weight of the functional connection in the functional connection matrix by adopting a threshold value method, and converting the functional connection matrix into a binary matrix according to the threshold value r being greater than 0.25;

step 3D, calculating the connectivity D of each voxel according to the binary matrix obtained in the step 3C;

and 3E, performing Z conversion on the connectivity D of each voxel obtained in the step 3D to obtain a DC cerebral graph.

Preferably, in step 3A, the pretreatment comprises: removing at least one of the first 10 time point data, the layer time correction, the head movement correction estimation, the linear trend in the removed signal and the low-pass filtering of 0.01-0.08 Hz;

and/or, in step 3E, the obtained DC brain map is also normalized by performing 6mm full-width half-height gaussian smoothing on the DC brain map.

Preferably, the step 4 specifically comprises the following steps:

step 4A, counting the DC brain graph obtained in the step 3, and taking a brain area with changed time point specificity function of the depression patient as a central node;

step 4B, analyzing the functional connection indexes of the central node and each area of the brain, and analyzing the change of the functional connection indexes in a baseline DC cerebral graph and a follow-up DC cerebral graph of the depression patient;

step 4C, calculating the HAMD reduction rate of the depression patient before and after treatment, wherein the calculation formula is as follows:

HAMD score ═ baseline score-follow up score ]/baseline score × 100%;

wherein the baseline score is the HAMD score assessed using the hamilton depression scale in step 1 and the follow-up score is the HAMD score assessed using the hamilton depression scale in step 2;

and 4D, performing linear regression model analysis on the change of the functional connection index obtained in the step 4B and the HAMD reduction rate obtained in the step 4C to obtain a functional connection index related to the HAMD reduction rate, namely a variable capable of representing depression symptom relief.

Preferably, the method for determining the brain region with time-point-specific function change of the depression patient comprises the following steps:

step 4Aa, counting the DC brain image obtained in the step 3 by adopting a linear model, wherein the linear model takes diagnosis multiplied by time point as an independent variable and takes general data acquired in the step 1 as a covariate; in independent variables, the diagnosis refers to a depressed patient or a normal control, and the time point refers to a baseline DC brain map or a follow-up DC brain map;

and step 4Ab, performing simple effect analysis on the areas where the significant interaction is found in the statistical result of the step 4Aa, wherein the simple effect analysis comprises the following steps: baseline DC profile for depression patients vs follow-up DC profile for depression patients, baseline DC profile for normal controls vs follow-up DC profile for normal controls, baseline DC profile for depression patients vs normal controls, follow-up DC profile for depression patients vs normal controls; identifying as a region specific for the time point change a region where the baseline DC profile of the normal control vs the follow-up DC profile of the normal control shows a significant functional change; after excluding the areas specific for the time point changes, the areas of the baseline DC profile of the depressed patients versus the follow-up DC profile of the depressed patients showing significant functional changes were identified as brain areas of time-specific functional changes in the depressed patients.

The invention also provides a computer device for selecting the depression symptom predictive variable, which comprises a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor executes the program to realize the method for selecting the depression symptom predictive variable.

The present invention also provides a computer-readable storage medium having stored thereon a computer program for implementing the above-described method for selecting a predictive variable for symptoms of depression.

In the present invention, the "depression symptom predictive variable" or "variable" refers to at least one index selected from indexes of functional connection between all central nodes and respective regions of the brain in the DC brain map, the selected index having a correlation with the condition of the depression patient. After obtaining the 'depression symptom predictive variable' or 'variable', a researcher (or doctor) can further analyze the variable in a subsequent resting state functional magnetic resonance imaging test, so as to predict and analyze the subsequent treatment effect and the disease development of depression patients. In the present invention, the term "significant" means that the p-value is less than 0.05 by statistical analysis, for example: "significant interaction" refers to an interaction having a p-value of less than 0.05.

By adopting the technical scheme of the invention, the brain function connection index of the use centrality (DC) can be analyzed, the brain is regarded as a huge and complete network, and a researcher (or doctor) is allowed to obtain variables capable of predicting the development condition of depression symptoms of individual patients with depression without selecting a priori interested area. In the subsequent diagnosis and treatment process of the depression patient, corresponding variables in other detection data (such as later-stage resting-state functional magnetic resonance imaging scanning data) of the depression patient are further analyzed, and the early prediction of the treatment effect and the disease development in the remission stage can be realized. The invention has good application prospect in the diagnosis and treatment of depression patients.

Obviously, many modifications, substitutions, and variations are possible in light of the above teachings of the invention, without departing from the basic technical spirit of the invention, as defined by the following claims.

The present invention will be described in further detail with reference to the following examples. This should not be understood as limiting the scope of the above-described subject matter of the present invention to the following examples. All the technologies realized based on the above contents of the present invention belong to the scope of the present invention.

Drawings

FIG. 1 is a schematic flow chart of example 1 of the present invention;

fig. 2 shows the results of the centrality analysis and the interaction analysis of the time points in embodiment 1 of the present invention.

Detailed Description

It should be noted that, in the embodiment, the algorithm of the steps of data acquisition, transmission, storage, processing, etc. which are not specifically described, as well as the hardware structure, circuit connection, etc. which are not specifically described, can be implemented by the contents disclosed in the prior art.

Example 1

The embodiment provides a method for selecting a depression symptom predictive variable and a computer device, wherein the computer device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, and the processor executes the program to realize the method for selecting the depression symptom predictive variable.

The flow of the method for selecting the predictive variable of the symptoms of the depression is shown in figure 1, and specifically comprises the following steps:

step 1, inspecting depression patients and normal controls, and collecting general data, Hamilton depression scale scores and resting state functional magnetic resonance imaging scanning data; the general data includes age, gender, and educational age;

step 2, collecting Hamilton depression scale scores and resting state function magnetic resonance imaging scanning data for depression patients and normal controls again after the depression patients receive treatment and the individual clinical symptoms are relieved; the method for confirming the individual clinical symptom relief of depression patients is to evaluate the individual clinical symptom relief by using Hamilton depression scale at 8 weeks, 24 weeks and 48 weeks after the depression patients receive treatment, and when the HAMD score is less than 7, the individual clinical symptom relief is judged.

Step 3, processing the rest state functional magnetic resonance imaging scanning data acquired in the step 1 and the step 2 to obtain a DC brain image, acquiring a baseline DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 1, and acquiring a follow-up DC brain image through the rest state functional magnetic resonance imaging scanning data acquired in the step 2;

the method for obtaining the DC brain map by using the resting state functional magnetic resonance imaging scanning data comprises the following steps:

step 3A, preprocessing the resting state functional magnetic resonance imaging scanning data; the pretreatment comprises the following steps: removing at least one of the first 10 time point data, the layer time correction, the head movement correction estimation, the linear trend in the removed signal and the low-pass filtering of 0.01-0.08 Hz;

step 3B, degree center index analysis: calculating Pearson correlation r between blood oxygen dependent (BOLD) signal time sequences of each pair of voxels by utilizing preprocessed resting state functional magnetic resonance imaging scanning data to obtain a functional connection matrix covering the whole brain;

the element in the functional connection matrix is r (i, j), and r (i, j) >0.25 indicates that a functional connection exists between the voxel i and the voxel j.

Step 3C, measuring the weight of the functional connection in the functional connection matrix by adopting a threshold value method, and converting the functional connection matrix into a binary matrix according to the threshold value r being greater than 0.25;

the element in the binary matrix is d_ijWhen r (i, j)>At 0.25, d_ij1 is ═ 1; when r (i, j) is less than or equal to 0.25, d_ij＝0。

Step 3D, calculating the connectivity D of each voxel according to the binary matrix obtained in the step 3C;

for any voxel i, the connectivity is the total number of voxels (j) with which there is a contiguous functional connection, and the calculation formula of the connectivity D is as follows:

D_i＝Σd_ij

where j is 1,2, … N, i ≠ j, and N is the total number of voxels.

And 3E, performing Z conversion on the connectivity D of each voxel obtained in the step 3D to obtain a DC brain map, and normalizing the obtained DC brain map by performing 6mm full-width half-height Gaussian smoothing on the DC brain map.

Step 4, analyzing the change of brain function activities before and after treatment according to the DC cerebral graph before and after treatment, and finding out a variable capable of representing depression symptom relief, wherein the specific steps are as follows:

step 4A, counting the DC brain graph obtained in the step 3, and taking a brain area with changed time point specificity function of the depression patient as a central node;

the method for determining the brain area with the time-point specific function change of the depression patient comprises the following steps:

step 4Aa, counting the DC brain map obtained in the step 3 by adopting a linear model in SPM8 software, wherein the linear model takes diagnosis multiplied by time point as an independent variable and general data acquired in the step 1 as a covariate; in independent variables, the diagnosis refers to a depressed patient or a normal control, and the time point refers to a baseline DC brain map or a follow-up DC brain map;

and step 4Ab, performing simple effect analysis on the areas where the significant interaction is found in the statistical result of the step 4Aa, wherein the simple effect analysis comprises the following steps: baseline DC profile for depression patients vs follow-up DC profile for depression patients, baseline DC profile for normal controls vs follow-up DC profile for normal controls, baseline DC profile for depression patients vs normal controls, follow-up DC profile for depression patients vs normal controls; identifying as a region specific for the time point change a region where the baseline DC profile of the normal control vs the follow-up DC profile of the normal control shows a significant functional change; after excluding the areas specific for the time point changes, the areas of the baseline DC profile of the depressed patients versus the follow-up DC profile of the depressed patients showing significant functional changes were identified as brain areas of time-specific functional changes in the depressed patients.

Step 4B, analyzing the functional connection indexes of the central node and each area of the brain, and analyzing the change of the functional connection indexes in a baseline DC cerebral graph and a follow-up DC cerebral graph of the depression patient;

step 4C, calculating the HAMD reduction rate of the depression patient before and after treatment, wherein the calculation formula is as follows:

HAMD score ═ baseline score-follow up score ]/baseline score × 100%;

wherein the baseline score is the HAMD score assessed using the hamilton depression scale in step 1 and the follow-up score is the HAMD score assessed using the hamilton depression scale in step 2;

and 4D, performing linear regression model analysis on the change of the functional connection index obtained in the step 4B and the HAMD reduction rate obtained in the step 4C to obtain a functional connection index related to the HAMD reduction rate, namely a variable capable of representing depression symptom relief. The term "correlated with the HAMD reduction ratio" means that the correlation with the change in the dependent variable (HAMD reduction ratio) in the regression analysis has a statistical significance.

An example of the variable selection of a depression patient by the above method is shown in fig. 2, and fig. 2 is an interaction analysis graph of diagnosis x time point in step 4Aa, in which (a) is the left temporomandibular gyrus, (B) is the right cerebellar gyrus, (C) is the left lingual gyrus, and (D) is the left dorsal medial prefrontal gyrus. L represents the left hemisphere, and R represents the right hemisphere.

Further regression analysis found that DC changes in the D brain region shown in fig. 2, i.e., the left dorsal medial prefrontal gyrus, correlated with the extent of clinical remission in depression patients (B3.404, p < 0.001).

The embodiment shows that the variable which can most accurately predict the treatment effect of the depression and the disease development in the remission stage can be obtained for the individual with depression without selecting a priori interested area. After obtaining the variable, the researcher (or doctor) further analyzes the corresponding variable in other detection data (such as later resting state functional magnetic resonance imaging scanning data) of the depression patient, and can predict the later treatment effect of the depression patient and whether the depression can be relieved. The invention can provide objective support basis for early clinical curative effect prediction of depression, reduces social burden of diseases and has good application prospect.