An ERP-based method for monitoring the response to antidepressant medication
Through the ERP-based method, patients' sociodemographic, medical history, psychological assessment and EEG data were analyzed, functional connection values were calculated, case database was established, feature recognition and two-line evaluation were carried out, and the accuracy and efficiency of traditional efficacy evaluation were solved. Problem, improving the effectiveness of antidepressant treatment.
Patent Information
- Application Number
- CN202410564482.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-09
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The efficacy evaluation of traditional antidepressant treatment depends on the clinical experience of doctors, with low accuracy and low efficiency, making it difficult to accurately predict drug responses early in the treatment, resulting in an extended time for ineffective treatment.
Using an ERP-based method, by collecting and analyzing sociodemographic data, past medical history, psychological assessment data, resting EEG and event-related potential data before and after treatment, calculating functional connection values, establishing a case library, performing feature identification and Two-line evaluation to determine the effectiveness of antidepressant treatment.
It is achieved to more accurately evaluate the efficacy of antidepressants in the early stages of treatment, reduce trial and error time, improve treatment effect, and reduce the uncertainty of doctors' experience and judgment.
Smart Images

Figure CN118155834B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of antidepressant disorder treatment efficacy identification, and specifically is an antidepressant drug treatment response monitoring method based on ERP. Background Art
[0002] Depression is a serious mental health problem that imposes a huge burden on individuals and society. The efficacy and response time of antidepressant drugs vary depending on the type of drug and individual differences. Traditional efficacy evaluation usually takes 4-6 weeks. By monitoring the early drug treatment response in the early stage of treatment, stratification and individualized treatment of patients can be achieved. For those with a good prognosis, the original plan should be adhered to, and for those with a poor prognosis, the drug or dosage should be adjusted in time, which can improve the overall efficacy and reduce ineffective treatment.
[0003] In addition, traditional methods for identifying the efficacy of antidepressant treatment often rely on the doctor's clinical experience and subjective judgment, and have problems such as low accuracy and low efficiency. ERP and EEG functional connectivity can dynamically reflect changes in the brain's neural activity and network pathways. However, a single measurement before taking the medicine is difficult to accurately predict changes after drug intervention. By comparing the changes in data before and after treatment, the patient's neurophysiological response to the drug can be better evaluated, thereby more accurately predicting the efficacy.
[0004] Based on this, the present invention provides an antidepressant drug treatment response monitoring method based on ERP. Summary of the invention
[0005] In order to solve the problems existing in the above scheme, the present invention provides an antidepressant drug treatment response monitoring method based on ERP.
[0006] The purpose of the present invention can be achieved by the following technical solutions:
[0007] An ERP-based antidepressant drug treatment response monitoring method, the method comprising:
[0008] Step 1: Obtaining initial material data before treatment and one week after drug treatment, respectively, the initial material data includes sociodemographic information, medical history, psychological assessment data, resting-state electroencephalogram, and event-related potential data of each tester; processing the initial material data to obtain corresponding material analysis data, the material analysis data is composed of each material unit data, and the material unit data includes resting-state electroencephalogram, functional connectivity data, and event-related potential data;
[0009] Furthermore, the method for obtaining functional connectivity data includes:
[0010] Step SA1: Establishing a region of interest and calculating the current source density of each region of interest;
[0011] Step SA2: Divide the EEG signal wave into four sub-bands, namely δ, θ, α, and β in descending order of frequency;
[0012] Step SA3: Assuming two EEG signals x and y, calculate the functional connectivity value between every two ROIs in each sub-band according to the EEG signals x and y;
[0013] Step SA4: Integrate the data obtained from steps SA1 to SA3 into functional connectivity data.
[0014] Furthermore, the calculation method of the functional connectivity value includes:
[0015] According to the formula Calculate the corresponding functional connectivity value;
[0016] Where: PLI xy is the functional connection value, and 0≤PLI xy ≤1, where PLI xy =0 means no coupling, PLI xy =1 means full coupling; θ x (tk) is the instantaneous phase of the EEG signal x, θ y (tk) is the instantaneous phase of the EEG signal y.
[0017] Further, ; .
[0018] Step 2: respectively setting corresponding depression characteristic items and efficacy characteristic items for the material analysis data before and after treatment; performing feature recognition on the material analysis data before and after treatment according to each depression characteristic item and efficacy characteristic item, and obtaining each depression characteristic item data and efficacy characteristic item data; establishing a case library according to each depression characteristic item data and efficacy characteristic item data, wherein the case library is used to store each preset case and the antidepressant effectiveness probability corresponding to each preset case;
[0019] Furthermore, the method for establishing the case library includes:
[0020] According to the data of each depression characteristic item, several preset cases are formed, and the corresponding anti-depression effectiveness probability is determined according to the data of each efficacy characteristic item corresponding to the preset case, and the obtained anti-depression effectiveness probability is marked on the corresponding preset case; after integrating the preset cases, a case library is established.
[0021] Step 3: Collect target collection data of the target person, process the target collection data, and obtain target analysis data corresponding to the target person;
[0022] Step 4: Extract features from the target analysis data to obtain feature recognition data, and determine matching feature data and various screening items based on the feature recognition data;
[0023] Furthermore, the method of determining matching feature data and each screening item according to feature recognition data includes:
[0024] Analyze the feature recognition data through the preset judgment model to obtain the abnormal value corresponding to each recognition item;
[0025] The expression of the judgment model is ; In the formula: i represents the corresponding identification item, i=1, 2, ..., n, n is a positive integer; si is the corresponding identification item data; the output data is an abnormal value 1 or 0;
[0026] The identification items with an outlier value of 0 are marked as matching items, and the identification items with an outlier value of 1 are marked as filtering items, and matching feature data are set according to the matching item data corresponding to each matching item.
[0027] Step 5: Filter the preset cases in the case library according to the filter items to obtain the reference cases; calculate the similarity between the matching feature data and the reference cases, remove the reference cases whose similarity is lower than the threshold value X1, and mark the remaining reference cases as filter cases;
[0028] Step six: Identify the similarity and antidepressant effectiveness probability of each screening case; establish a statistical coordinate graph, wherein the horizontal axis of the statistical coordinate graph is the similarity and the vertical axis is the antidepressant effectiveness probability, and integrate the similarity and antidepressant effectiveness probability of each screening case into corresponding coordinate points and input them into the statistical coordinate graph; perform a double-line evaluation on the statistical coordinate graph to obtain corresponding double-line evaluation values, and judge whether the antidepressant treatment is effective based on the double-line evaluation values.
[0029] Furthermore, the method of performing a double-line evaluation on the statistical coordinate graph includes:
[0030] Generate similar positioning lines and depression positioning lines in the statistical coordinate graph; form a double-line evaluation area based on the similar positioning lines and depression positioning lines;
[0031] Mark each coordinate point in the double-line evaluation area as an evaluation point; mark each evaluation point as G(x, y);
[0032] Calculate the single evaluation value of each evaluation point according to the formula YPj=(x-x0)×(y-y0);
[0033] Where: YPj is a single evaluation value, j represents the corresponding evaluation point, j=1, 2, ..., m, m is a positive integer; x0 is the similarity corresponding to the similar positioning line; y0 is the antidepressant effectiveness probability corresponding to the depression positioning line;
[0034] Match the corresponding adjustment coefficient according to the single assessment value corresponding to each assessment point;
[0035] According to the formula Calculate the corresponding double-line evaluation value;
[0036] Where: WU is the double-line assessment value; δj is the adjustment coefficient; j=1, 2, ..., m, m is a positive integer.
[0037] Furthermore, the method for judging whether the antidepressant treatment is effective based on the dual-line evaluation value includes:
[0038] When the double-line evaluation value is greater than the threshold value X2, it is determined that the antidepressant treatment is effective;
[0039] When the double-line evaluation value is not greater than the threshold value X2, it is determined that the antidepressant treatment is ineffective.
[0040] Compared with the prior art, the present invention has the following beneficial effects:
[0041] The present invention realizes continuous and systematic collection and analysis of symptom data of patients, which can more accurately evaluate the efficacy of antidepressant drugs in the early stages of treatment, reduce trial and error time, and help doctors adjust treatment plans in a timely manner, thereby improving treatment effects; avoiding the uncertainty caused by relying on doctors' experience to make judgments; and improving the efficacy of antidepressant treatment. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0043] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0044] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0045] like Figure 1 As shown, a method for monitoring antidepressant drug treatment response based on ERP, the method comprising:
[0046] Step 1: Obtain the initial data of the material before treatment and one week after drug treatment, specifically ERP-based drug treatment for depressive disorders; the initial data of the material includes the socio-demographic information, medical history, psychological assessment data, resting-state electroencephalogram, event-related potential (EPR) and other relevant data of each tester; and process the initial data of the material accordingly to obtain the material analysis data, which is composed of the data of each material unit, including resting-state electroencephalogram, resting-state functional connectivity data, event-related potential data and task-state functional connectivity data;
[0047] Sociodemographic data: including age, gender, education level and marital status.
[0048] Past medical history: including smoking history, drinking history, chronic disease history, and cardiovascular and cerebrovascular disease history.
[0049] Psychological assessment data: including Hamilton Depression Rating Scale-21 items, Hamilton Anxiety Rating Scale, Social Support Scale, Patient Health Questionnaire-15, Pittsburgh Sleep Quality Index, Chinese version of Perceived Stress Scale, Time Experience Pleasure Scale and Digital Symbol Conversion Test, etc.
[0050] Exemplary, resting-state EEG acquisition: An electrode cap with 128 electrode points was used, and the electrodes were placed according to the 10 / 5 system. TP9 and TP10 were placed on the left and right mastoids, respectively, and the ECG channel was placed above the right eye to monitor electrooculographic activity. The sampling rate was 1000 Hz, and the impedance was maintained below 20 kΩ. Participants remained still during the recording process, reducing head and body movements and eye movements to reduce electromyographic and electrooculographic artifacts, respectively, and continuously recorded 6 minutes of EEG (electroencephalogram) data in the open and closed eyes state.
[0051] The steps of EEG data preprocessing are as follows:
[0052] 1) Re-reference: Use TP9 and TP10 leads for re-reference, and return the original reference electrode to the FCz channel;
[0053] 2) Filtering: Use infinite impulse response (IIR) filtering with a filtering range of 0.5Hz to 45Hz and a notch filter at 50Hz;
[0054] 3) Reduce the sampling rate: the sampling rate is reduced to 500Hz;
[0055] 4) Artifact Correction: A semi-automatic method based on independent component analysis (ICA) is used to identify and remove vertical and horizontal eye movement artifacts and blink signals;
[0056] 5) Bad segment removal: A semi-automatic procedure is used to remove bad segments according to the following criteria: the voltage difference between two sampling points is greater than 50μV, the voltage difference within a segment is greater than 100μV, and the maximum voltage difference within a 100-ms interval is less than 0.50μV. Finally, each segment is visually inspected and bad segments are manually removed.
[0057] EEG functional connectivity analysis:
[0058] Functional connectivity: The functional connectivity based on the scalp surface channels cannot completely avoid the influence of volume conduction effect, which may lead to the existence of false connections. However, directly calculating the functional connectivity in the source space after source tracing analysis will significantly improve its accuracy. First, use the Low Resolution Brain Electromagnetic Tomography (LORETA) function of Brain Vision Analyzer 2.2 to perform source tracing analysis on the pre-processed clean EEG data. The steps of source tracing analysis are as follows: 1) Establish regions of interest (ROI): The establishment of ROI depends on the source tracing analysis file provided by Brain Vision Analyzer 2.2. 2) Calculate the current source density of each ROI: Select current density in the LORETA interface and calculate the current density value in each ROI; 3) Output: Select the general data export module on the LORETA node to export the current density value of each ROI at each time point; 4) Import into MATLAB: The source tracing data after the above steps is imported into MATLAB using the EEGLAB toolbox based on MATLAB R2023a for subsequent functional connectivity analysis.
[0059] PLI is applied to calculate the functional connectivity between different ROIs. As a phase-based connectivity analysis, PLI relies on the distribution of the phase angle difference between two electrodes. The idea is that when the neural clusters are functionally coupled, the oscillation sequence at this time becomes synchronized in phase. For two given EEG signals x and y, the calculation formula of PLI value is as follows, where k=1, 2,…, N; N is a positive integer;
[0060] ;
[0061] Where: 0≤PLI xy ≤1, where PLI xy =0 means no coupling, PLI xy =1 means full coupling; θ x (tk) is the instantaneous phase of the EEG signal x, θ y (tk) is the instantaneous phase of the EEG signal y;
[0062] ;
[0063] ;
[0064] Through the above steps, the functional connectivity data between every two ROIs were generated.
[0065] In order to better understand brain activity, the EEG signal waves can also be divided into four main sub-bands, which are divided into 4 frequency bands according to the frequency. From low frequency to high frequency, they are delta (δ, 1-4Hz), theta (θ, 4-8Hz), alpha (α, 8-12Hz) and beta (β, 13-30Hz) bands, and finally the functional connectivity values between every two ROIs in the above frequency bands are generated.
[0066] ERP data collection: After completing the clinical data collection, the participants were asked to complete the classic visual oddball experiment in a dimly lit, soundproof, and electrically shielded room. The specific experimental process is as follows:
[0067] The visual oddball experiment consists of two visual stimuli, a red car picture as a deviant stimulus (S2) and a blue car picture as a standard stimulus (S1). The duration of both stimuli is 200 milliseconds (including a 10-millisecond rise / fall amplitude). The specific process is as follows: the computer screen is adjusted to an appropriate brightness, and the participant first receives a text prompt: "An experiment will be conducted later. When a blue car appears on the computer screen, no reaction is required. When a red car appears on the computer screen, the computer space bar needs to be pressed as quickly as possible." The experiment begins when the patient is ready to press any key on the computer. The experiment begins with 10 training exercises. When the accuracy of the exercise reaches more than 80%, the subsequent formal experiment is entered, and repeated training exercises are denied. After the formal experiment begins, red or blue cars appear randomly on the computer screen, with a random stimulus interval of 900-1100 milliseconds, including a total of 310 trials (20% deviant stimuli and 80% standard stimuli), and the deviant stimuli appear in the standard stimuli in a pseudo-random order, so that at least two standard stimuli appear before each deviant stimulus.
[0068] ① Preprocessing: Use existing software to preprocess ERP data. The preprocessing steps are roughly the same as EEG, but slightly different. The specific steps are as follows: 1) Re-reference: TP9 and TP10 were used for re-reference, and the original reference electrode was regressed to the FCz channel; 2) Filtering: Infinite impulse response (IIR) filtering was used with a filtering range of 1 Hz to 30 Hz and a 50 Hz notch filter; 3) Artifact correction: A semi-automatic method based on independent component analysis (ICA) was used to identify and remove vertical and horizontal eye movement artifacts and blinks; 4) Removal of bad segments and interpolation of bad leads: A semi-automatic procedure was used to remove bad segments according to the following criteria: (a) the voltage difference between two sampling points was greater than 50 μV, (b) the voltage difference within a segment was greater than 300 μV, and (c) the maximum voltage difference within a 100 ms interval was less than 0.50 μV. In addition, each segment was visually inspected and bad segments were manually excluded; 5) Difference bad leads: For channels that were eliminated by more than 20%, we will perform interpolation processing to ensure the accuracy of the final result. Finally, all segments were visually inspected for artifact removal, and channels with more than 20% artifact removal were interpolated. 5) Segmentation: ERP data from 200 ms before stimulation to 1000 ms after stimulation were selected, and baseline correction was performed using the 200 ms before stimulation data as the baseline.
[0069] ②ERP component analysis: ERP component analysis mainly focuses on the N1, P2, N2, and P3 components of standard and deviant stimuli. The traceable data after the above steps are pre-processed and cleaned. The data is imported into MATLAB using the EEGLAB toolbox based on MATLAB R2023a, and the amplitude and latency of each ERP component of each subject are calculated respectively.
[0070] Use complex network analysis methods, such as graph theory analysis, to reveal network characteristics among different groups.
[0071] Step 2: According to the identification needs before and after treatment, various depression feature items are set for the material analysis data before treatment; the depression feature items are data items related to the patient's depression status; various efficacy feature items are set for the material analysis data after treatment; the efficacy feature items are data items related to the efficacy of the patient after treatment;
[0072] Perform feature recognition on the corresponding material analysis data according to each depression feature item and efficacy feature item to obtain each depression feature item data and efficacy feature item data;
[0073] According to the material analysis data, various combinations of depression characteristic items are obtained and marked as preset cases; and the corresponding anti-depression effectiveness probability is determined according to the data of each efficacy characteristic item corresponding to the preset case, and the corresponding average value is taken; the obtained anti-depression effectiveness probability is marked on the corresponding preset case; and the preset cases are integrated to establish a case library.
[0074] Step three: Mark the patients who need to be monitored for treatment response as target persons, and collect data on the target persons according to the content of the initial data of the material before treatment; obtain the target collection data corresponding to the target persons, process the target collection data, and obtain the target analysis data corresponding to the target persons, that is, process the target collection data according to the method of obtaining material analysis data.
[0075] Step 4: Extract features from the target analysis data to obtain feature recognition data, i.e., data related to each depression feature item, which is equivalent to depression feature item data. Each depression feature item is used as an identification item for recognition extraction to obtain identification item data corresponding to each identification item, and integrate them into feature recognition data corresponding to the depression feature item data.
[0076] A judgment model is established based on the material analysis data. The judgment model is used to evaluate whether the identification item data of each identification item meets the data range of the corresponding depression feature item, that is, it may be a feature of depression. If the data that is completely impossible to appear in the depression feature item under depression conditions is considered to not meet the feature requirements, that is, it is abnormal data; it is specifically established based on the isolation forest algorithm, and the expression is ; In the formula: i represents the corresponding identification item, i=1, 2, ..., n, n is a positive integer; si is the identification item data of the corresponding identification item; the output data is an abnormal value 1 or 0;
[0077] The judgment model is used to analyze the identification item data in the feature recognition data to obtain the outlier value corresponding to each identification item, mark the identification item with an outlier value of 0 as a matching item, and mark the identification item with an outlier value of 1 as a screening item, and set the matching feature data according to the matching item data corresponding to each matching item.
[0078] Step 5: According to each filtering item, the preset cases corresponding to the filtering items in the case library are marked to indicate elimination, and the unmarked preset cases are marked as reference cases; the similarity between the matching feature data and each reference case is calculated, and the reference cases with similarity lower than the threshold X1 are eliminated, and the remaining reference cases are marked as filtering cases.
[0079] Step 6: Identify the similarity and antidepressant efficacy probability of each screening case; generate a statistical coordinate graph based on the similarity and antidepressant efficacy probability of each screening case, the horizontal axis of the statistical coordinate graph is the similarity, and the vertical axis is the antidepressant efficacy probability, and substitute the similarity and antidepressant efficacy probability of each screening case into the corresponding coordinate points; perform a double-line evaluation on the statistical coordinate graph to obtain the corresponding double-line evaluation value, and judge whether the antidepressant treatment is effective based on the double-line evaluation value;
[0080] Methods for performing dual-line evaluation of statistical coordinate graphs include:
[0081] Set the minimum similarity and the minimum antidepressant effectiveness probability. Both the minimum similarity and the minimum antidepressant effectiveness probability are set by professionals, mainly referring to whether the screening case can be applied and the critical state of whether the antidepressant is effective. Set the corresponding similarity positioning line and depression positioning line according to the minimum similarity and the minimum antidepressant effectiveness probability, that is, the straight lines perpendicular to the minimum similarity and the minimum antidepressant effectiveness probability, divide the coordinate area into four areas by the depression positioning line and the similarity positioning line, and mark the upper right area corresponding to the right side of the similarity positioning line and the upper side of the depression positioning line as the double-line evaluation area.
[0082] Mark each coordinate point in the double-line evaluation area as an evaluation point; mark each evaluation point as G(x, y), and calculate the single evaluation value of each evaluation point according to the formula YPj=(x-x0)×(y-y0); where: YPj is a single evaluation value, j represents the corresponding evaluation point, j=1, 2, ..., m, m is a positive integer; x0 is the similarity corresponding to the similar positioning line, that is, the minimum similarity; y0 is the antidepressant effective probability corresponding to the depression positioning line, that is, the minimum antidepressant effective probability;
[0083] According to the single evaluation value, the corresponding adjustment coefficient is matched for each evaluation point. The adjustment coefficient is set according to the size of the single evaluation value. The larger the single evaluation value, the larger the adjustment coefficient. The value range of the single evaluation value is determined to be 0 to (100-x0)×(100-y0); the value range of the adjustment coefficient is [0, 2], 0 corresponds to 0, (100-x0)×(100-y0) corresponds to 2, and the adjustment coefficient corresponding to each single evaluation value is determined according to the corresponding proportional relationship; however, this method has a large error, so a corresponding adjustment coefficient analysis model can be established based on a CNN network or a DNN network, and a corresponding training set can be established manually for training. The training set includes input data and output data, the input number of households includes a single evaluation value, and the output data includes the adjustment coefficient; training is performed through the established training set, and verification is performed in combination with the initial data of the material until the accuracy meets the requirements; analysis is performed through the adjustment coefficient analysis model after successful training to obtain the corresponding adjustment coefficient;
[0084] According to the formula Calculate the corresponding double-line evaluation value;
[0085] Where: WU is the double-line assessment value; δj is the adjustment coefficient; j=1, 2, ..., m, m is a positive integer;
[0086] When the double-line evaluation value is greater than the threshold value X2, it is determined that the antidepressant treatment is effective;
[0087] When the double-line evaluation value is not greater than the threshold value X2, it is determined that the antidepressant treatment is ineffective.
[0088] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.
[0089] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. An ERP-based antidepressant drug treatment response monitoring method, characterized in that: Methods include: Step 1: Obtaining initial data before treatment and one week after drug treatment, including sociodemographic information, medical history, psychological assessment data, resting electroencephalogram, and event-related potential data of each tester; Processing the initial material data to obtain corresponding material analysis data, wherein the material analysis data is composed of various material unit data, and the material unit data includes resting state electroencephalogram, functional connectivity data and event-related potential data; Step 2: respectively setting corresponding depression feature items and efficacy feature items for the material analysis data before and after treatment; performing feature recognition on the material analysis data before and after treatment according to each depression feature item and efficacy feature item to obtain each depression feature item data and efficacy feature item data; Establishing a case library according to each depression characteristic item data and efficacy characteristic item data, wherein the case library is used to store each preset case and the antidepressant effectiveness probability corresponding to each preset case; Step 3: Collect target collection data of the target person, process the target collection data, and obtain target analysis data corresponding to the target person; Step 4: Extract features from the target analysis data to obtain feature recognition data, and determine matching feature data and various screening items based on the feature recognition data; Step 5: Filter the preset cases in the case library according to the filter items to obtain the reference cases; calculate the similarity between the matching feature data and the reference cases, remove the reference cases whose similarity is lower than the threshold value X1, and mark the remaining reference cases as filter cases; Step 6: Identify the similarity and antidepressant effectiveness probability of each screening case; establish a statistical coordinate graph, wherein the horizontal axis of the statistical coordinate graph is the similarity and the vertical axis is the antidepressant effectiveness probability, integrate the similarity and antidepressant effectiveness probability of each screening case into corresponding coordinate points and input them into the statistical coordinate graph; perform a double-line evaluation on the statistical coordinate graph to obtain a corresponding double-line evaluation value, and judge whether the antidepressant treatment is effective according to the double-line evaluation value; The method for determining matching feature data and each filter item according to feature recognition data includes: Analyze the feature recognition data through the preset judgment model to obtain the abnormal value corresponding to each recognition item; The expression of the judgment model is ; In the formula: i represents the corresponding identification item, i=1, 2, ..., n, n is a positive integer; si is the corresponding identification item data; the output is an abnormal value 1 or 0; Mark the identification items with an abnormal value of 0 as matching items, mark the identification items with an abnormal value of 1 as filtering items, and set matching feature data according to the matching item data corresponding to each matching item; Methods for performing dual-line evaluation of statistical coordinate graphs include: Set the minimum similarity and the minimum antidepressant effective probability, set the corresponding similarity positioning line and depression positioning line according to the minimum similarity and the minimum antidepressant effective probability, divide the coordinate area into four areas through the depression positioning line and the similarity positioning line, and mark the area on the right side of the similarity positioning line and the upper side of the depression positioning line as the double-line evaluation area; Mark each coordinate point in the double-line evaluation area as an evaluation point; mark each evaluation point as G(x, y); Calculate the single evaluation value of each evaluation point according to the formula YPj=(x-x0)×(y-y0); Where: YPj is a single evaluation value, j represents the corresponding evaluation point, j=1, 2, ..., m, m is a positive integer; x0 is the similarity corresponding to the similar positioning line; y0 is the antidepressant effectiveness probability corresponding to the depression positioning line; Match the corresponding adjustment coefficient according to the single assessment value corresponding to each assessment point; According to the formula Calculate the corresponding double-line evaluation value; Where: WU is the double-line assessment value; δj is the adjustment coefficient; j=1, 2, ..., m, m is a positive integer; Methods for judging whether antidepressant treatment is effective based on the two-line assessment value include: When the double-line evaluation value is greater than the threshold value X2, it is determined that the antidepressant treatment is effective; When the double-line evaluation value is not greater than the threshold value X2, it is determined that the antidepressant treatment is ineffective.
2. The method for monitoring antidepressant drug treatment response based on ERP according to claim 1, characterized in that: Methods for obtaining functional connectivity data include: Step SA1: Establishing a region of interest and calculating the current source density of each region of interest; Step SA2: Divide the EEG signal wave into four sub-bands, namely δ, θ, α, and β in descending order of frequency; Step SA3: Assuming two EEG signals x and y, calculate the functional connectivity value between every two ROIs in each sub-band according to the EEG signals x and y; Step SA4: Integrate the data obtained from steps SA1 to SA3 into functional connectivity data.
3. The method for monitoring antidepressant drug treatment response based on ERP according to claim 2, characterized in that: The calculation method of functional connectivity value includes: According to the formula Calculate the corresponding functional connectivity value; Where: PLI xy is the functional connection value, and 0≤PLI xy ≤1, where PLI xy =0 means no coupling, PLI xy =1 means full coupling; θ x (tk) is the instantaneous phase of the EEG signal x, θ y (tk) is the instantaneous phase of the EEG signal y.
4. The method for monitoring antidepressant drug treatment response based on ERP according to claim 3, characterized in that: ; 。 5. The method for monitoring antidepressant drug treatment response based on ERP according to claim 1, characterized in that: The method of establishing a case library includes: According to the data of each depression characteristic item, several preset cases are formed, and the corresponding anti-depression effectiveness probability is determined according to the data of each efficacy characteristic item corresponding to the preset case, and the obtained anti-depression effectiveness probability is marked on the corresponding preset case; after integrating the preset cases, a case library is established.
