System for evaluating treatment effect of patient with alcohol use disorder by utilizing EEG (electroencephalogram) index

Abstract

The invention provides a system for evaluating the treatment effect of a patient suffering from alcohol use disorder by utilizing EEG indexes. The system utilizes EEG indexes to monitor and analyze physiological and neural responses of the patient. The system measures electroencephalogram activity through an electroencephalogram monitoring sensor, then calculates a nonlinear kinetic index of the EEG through a data processing unit, and compares the nonlinear kinetic index of the EEG with a preset reference value or index mode to identify the psychological state and physiological response characteristics of the patient. Finally, the system judges the physiological state, the brain activity state and the treatment effect of the alcohol use disorder patient according to the analysis result of the EEG indexes, and the analysis and judgment result is displayed through an output unit. The system can accurately evaluate the treatment effect of the patient with the alcohol use disorder, so that accurate treatment and prevention strategies are provided for the patient with the alcohol use disorder.

Description

Technical Field

[0001] The present invention relates to a system for evaluating the treatment effect of patients with alcohol use disorder. Background Art

[0002] Alcohol use disorders (AUD) are characterized by compulsive heavy drinking and loss of control over alcohol intake. It is a common chronic disease that is extremely harmful to society. At present, drug treatment for AUD is mainly symptomatic treatment to relieve withdrawal symptoms, but not to reduce psychological craving. Therefore, the psychological craving and relapse rate of AUD patients after discharge from hospital are still very high. Obviously, alcohol use disorders require a multidisciplinary approach that combines drug treatment, social support and psychological intervention. In addition, to achieve abstinence and prevent relapse, detection methods that can accurately assess psychological craving are also needed.

[0003] How to accurately evaluate the treatment effect of alcohol addiction is currently facing a great challenge. The neural activity of the human brain reflected in EEG is a complex activity with nonlinear dynamic characteristics. The nervous system has nonlinear and chaotic characteristics. Therefore, methods related to nonlinear theory are helpful to help us understand and explain the nonlinear dynamic characteristics of EEG and the corresponding brain neural activity processes. At present, among the various nonlinear dynamic algorithms that have been applied to EEG analysis, "complexity" and "entropy" are the most widely used. These algorithms show the nonlinear characteristics of EEG signals from different levels. EEG complexity measures the information capacity of signal segments, thereby reflecting the potential activity characteristics of neurons.

[0004] It should be noted that the information disclosed in the above background technology section is only used for understanding the background of the present application, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the invention

[0005] The main purpose of the present invention is to overcome the defects of the above-mentioned background technology and provide a system for evaluating the treatment effect of patients with alcohol use disorder AUD using EEG indicators. The present invention uses nonlinear dynamic indicators to evaluate the treatment effect of AUD patients. By performing nonlinear characteristic analysis through the system of the present invention, the treatment effects of different interventions (VBE group, VBN group and CVE group) on AUD patients can be accurately evaluated, and more accurate treatment and detection and prevention strategies can be provided for AUD patients, which has a good guiding significance.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A system for evaluating the treatment effect of patients with alcohol use disorder (AUD) using EEG indicators, comprising:

[0008] EEG monitoring sensors, which measure the user’s electrical brain activity;

[0009] A data processing unit, connected to the EEG monitoring sensor, is used to perform the following operations:

[0010] a. Obtain the user's brain electrical activity through the EEG signal measured in real time by the EEG monitoring sensor, and calculate the nonlinear dynamic indicators reflecting the brain state, wherein the nonlinear dynamic indicators include: one or more of the EEG complexity index, the entropy index and the recursive quantitative analysis RQA index;

[0011] b. Comparing the calculated nonlinear dynamic index with a preset baseline value or an index pattern under different psychological states to identify the brain activity characteristics of the patient under a specific psychological state; the preset baseline value or index pattern is based on the classification of AUD patients under different interventions and is analyzed and verified using statistical methods;

[0012] c. analyzing and determining the characteristics of the neural mechanism of the patient's brain when processing alcohol stimulation under different conditions based on the comparison results of the nonlinear dynamic index with the preset reference value or index pattern;

[0013] d. Comprehensively judge the brain activity status and treatment effect of AUD patients based on the analysis results of EEG indicators;

[0014] The output unit is used to output the EEG index analysis and judgment results.

[0015] The present invention has the following beneficial effects:

[0016] The system of the present invention provides a new and effective method for evaluating the treatment effect of patients with alcohol use disorder (AUD) by monitoring and analyzing EEG indicators. The system can accurately capture the physiological and neural responses of patients during treatment, thereby providing clinicians with accurate treatment effect evaluation. Using statistical methods, the system of the present invention can verify the significance and relevance of preset benchmark values ​​or indicator patterns, thereby enhancing the reliability of the evaluation results. The system uses nonlinear dynamic indicators of EEG to provide a comprehensive treatment effect evaluation tool that can more comprehensively reflect the physiological and psychological state of patients. The system can also detect abnormal changes in the patient's state in a timely manner by continuously monitoring the trend of indicator changes, so as to achieve early intervention and prevent deterioration. In addition, the real-time feedback provided by the system can improve the patient's treatment compliance and participation. Using the personalized monitoring and analysis results generated by the system, the treatment plan can be adjusted according to the patient's specific response to achieve personalized treatment and improve the effectiveness of treatment, thereby providing strong technical support for the rehabilitation of AUD patients. The system implements a non-invasive, low-cost treatment effect evaluation, and the design of the system allows it to be integrated with existing medical equipment and workflows, which is convenient for implementation and use in different medical environments, while helping to reduce medical costs and improve the efficiency of medical resource use.

[0017] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 FIG. 4 is a power spectrum density diagram of a subject according to an embodiment of the present invention.

[0019] Figure 2 These are pictures related to the nonlinear HRV analysis results of a subject according to an embodiment of the present invention, (a) Poincare map, (b) detrended fluctuation analysis double logarithmic map.

[0020] Figure 3 Schematic diagram of the position of the EEG cap electrodes according to an embodiment of the present invention.

[0021] Figure 4 Schematic diagram of a recursive graph matrix according to an embodiment of the present invention.

[0022] Figure 5 Schematic diagram of the system structure for evaluating the treatment effect of patients with alcohol use disorder using EEG indicators according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope and application of the present invention.

[0024] Key terms and concepts:

[0025] EEG related concepts

[0026] Electroencephalogram (EEG): EEG signals contain neural information from the cerebral cortex. They are spontaneous, rhythmic electrical activities of brain cell groups, and contain a wealth of thinking, emotional, spiritual, and psychological activities. The biopotential of the brain is amplified from the scalp through sophisticated electronic instruments and recorded through electrodes. This is the EEG. The neural activity of the human brain reflected in the EEG is a complex activity with nonlinear dynamic characteristics. The nervous system has nonlinear and chaotic characteristics. Therefore, methods related to nonlinear theory are helpful in helping to understand and explain the nonlinear dynamic characteristics of EEG and the corresponding brain neural activity processes. For example, in anesthesia, nonlinear dynamics has become an important reference indicator for evaluating the EEG state of anesthesia.

[0027] At present, among the various nonlinear dynamics algorithms that have been applied to EEG analysis, "complexity" and "entropy" are the most widely used. These algorithms show the nonlinear characteristics of EEG signals from different levels. EEG complexity measures the information capacity of signal segments, thereby reflecting the potential activity characteristics of neurons. Complexity and various algorithms derived from it have been applied to characterize the randomness of various biological signals. EEG complexity measurements have demonstrated their potential as markers of psychopathology (such as depression and schizophrenia). Entropy is a characteristic distribution of metric signals based on Shannon information theory. It can evaluate the possibility of predicting the amplitude of EEG signals in the future based on the probability distribution of the amplitude of EEG signals in the past, including EEG power, permutation entropy (PE), approximate entropy (ApEn), etc.

[0028] In addition to complexity and entropy, other nonlinear dynamics methods for analyzing EEG include recurrence quantification analysis (RQA). RQA does not require assumptions about the stationarity, noise, or length of the time series, so it has the advantage of analyzing EEG signals in complex environments. RQA is also used to analyze EEG signals of pathological brain activity. RQA can also quantify the activity of the system by processing linear and nonlinear time series without considering the quantity or dynamic nature of individual information sources.

[0029] like Figure 5 As shown, an embodiment of the present invention provides a system for evaluating the treatment effect of patients with alcohol use disorder (AUD) using EEG indicators, including:

[0030] EEG monitoring sensors, which measure the user’s electrical brain activity;

[0031] A data processing unit, connected to the EEG monitoring sensor, is used to perform the following operations:

[0032] a. Obtain the user's brain electrical activity through the EEG signal measured in real time by the EEG monitoring sensor, and calculate the nonlinear dynamic indicators reflecting the brain state, the nonlinear dynamic indicators include: EEG complexity index, entropy index and recursive quantitative analysis RQA index;

[0033] b. Comparing the calculated nonlinear dynamic index with a preset baseline value or an index pattern under different psychological states to identify the brain activity characteristics of the patient under a specific psychological state; the preset baseline value or index pattern is based on the classification of the subconscious and conscious reactions of AUD patients under different exposure conditions, and is analyzed and verified using statistical methods;

[0034] c. Analyze and determine the characteristics of the neural mechanism of the brain when patients process alcohol stimulation under different conditions based on the comparison results of EEG complexity index, entropy index and recursive quantitative analysis RQA index with the preset benchmark value or index pattern;

[0035] d. Comprehensively judge the brain activity status and treatment effect of AUD patients based on the analysis results of EEG indicators;

[0036] The output unit, such as a display screen, is used to output (eg, display) the EEG index analysis and judgment results.

[0037] In some embodiments, the complexity indicators include one or more of LZCmean, LZCmedian, LZCmid-point and LZCk-means; the entropy indicators include permutation entropy PE; the recursive quantitative analysis RQA indicators include one or more of recursion rate RR, determinism DET, ENTR and average diagonal length L.

[0038] In some embodiments, the baseline value or indicator pattern is established in the following manner: classifying the responses of AUD patients under different conditions; establishing EEG indicator patterns corresponding to subconscious processing of alcohol stimuli, subconscious processing of neutral stimuli, and conscious processing of alcohol stimuli based on the classification results; and using statistical methods, such as one-way analysis of variance and Pearson correlation analysis, to verify the significance and correlation of the indicator pattern.

[0039] Experimental Example 1

[0040] Processing and Extraction of EEG Nonlinear Dynamics Indexes

[0041] The nonlinear dynamics related indexes of EEG signals in the prefrontal, frontal, central, parietal and occipital regions were extracted using MATLAB-based programming codes. The five electrode points marked from front to back on the anterior-posterior sagittal line (a line from the root of the nose to the external occipital protuberance) were selected to correspond to the five regions selected for analysis, including the frontal pole midpoint (Fpz), frontal midpoint (Fz), central point (Cz), vertex (Pz) and occipital point (Oz). The EEG signals collected by these five electrode points located in the midline of the EEG cap are less disturbed and of better quality than those of other electrode points on the side. The nonlinear dynamics methods used include complexity, entropy and recursive quantitative analysis.

[0042] Complexity

[0043] Complexity-based indicators can be used to evaluate the randomness of signal time series, evaluate the "information content" in EEG signals, and the degree of neuronal activation. For example, the complexity of EEG under eyes-open resting conditions should be greater than that under eyes-closed resting conditions, and evaluate the rate at which new patterns appear in time series signals. The first step of complexity analysis is to coarse-grain the sequence, that is, to convert the signal into a sequence of 0s and 1s by comparing the signal amplitude and threshold. There are four complexity indicators: (1) LZC mean The average value of the signal amplitude is used as the threshold; (2) LZC median The median of the signal amplitude is used as the threshold; (3) LZC mid-point The mean of the maximum and minimum signal amplitudes is used as the threshold; (4) LZC k-means Two thresholds are set: z1=s m +ε*s m and z2 = s m -ε*s m , ε can be set to 0.05 according to experience, s m Represents the average value of the signal amplitude. The first three complexity indicators take 0 if the signal amplitude is less than the threshold and take 1 if it is greater than the threshold. The fourth one compares the distance between the data point and the two thresholds. If the distance to z1 is less than the distance to z2, it takes 1, otherwise it takes 0.

[0044] The sorting complexity PLZC considers not the absolute signal amplitude value but the size order between the values. The sorting complexity PLZC is proposed based on the complexity LZC combined with the sorting method, and its robustness is improved by this method. The overall steps of PLZC are: (1) Convert the EEG signal into a character sequence by sorting; (2) Set the initialization array S, Q, c(n); (3) Combine S and Q into SQ, and remove the last character of SQ into SQv; (4) Determine whether Q is the last character of the character sequence; (5) Determine whether Q is a combination in SQv; (6) Extract the next character from the character sequence to Q; (7) Assign SQ to S, and increase the complexity c(n) by 1; (8) Normalize the complexity c(n). In the process, two important parameters, embedding dimension m and time delay τ, are determined by mutual information and pseudo minimum neighbor methods (each group selects the subjects with the best signal segment quality as examples) and previous research experience. The embedding dimension m=5 and the value of time delay τ are as follows: Fpz is 13, Fz is 13, Cz is 11, Pz is 13, and Oz is 13 in the VBE group; Fpz is 5, Fz is 10, Cz is 10, Pz is 12, and Oz is 6 in the VBN group; Fpz is 11, Fz is 12, Cz is 14, Pz is 11, and Oz is 13 in the CVE group.

[0045] entropy

[0046] Entropy methods have been applied to information processing of neural activities. The entropy-based index, ranking entropy (PE), is a quantitative complexity algorithm used to study the local structure of dynamic time series. It converts a given time series into a series of ordered patterns, each of which describes the order relationship between the current and equidistant past values ​​within a given time. It can evaluate the probability distribution of EEG signal amplitudes in the past and predict the possibility of EEG signal amplitudes in the future, reflecting the ability of information interaction in the brain. Compared with other algorithms, PE has the characteristics of simplicity and strong noise resistance. The overall steps of ranking entropy are: (1) ranking; (2) calculating the ratio of the occurrence of ranking patterns; (3) calculating entropy according to the ratio of ranking patterns; (4) normalizing to ranking entropy. The two important parameters in the process are the embedding dimension m and the value of time delay τ, which are the same as the complexity part.

[0047] Recursive quantitative analysis

[0048] In the framework of nonlinear methods, recurrence graph (RP)-based analysis (RQA) describes the dynamic characteristics of nonlinear systems. It proposes a graphical representation of recurrence in complex time series. Overall steps:

[0049] (1) For a time series U t (u1, u2, u3, ...u T ), select the embedding dimension m and time delay τ (the same value as the complexity part) and then reconstruct the phase space xi =(u i ,u i+τ ,……,u i+(m-1)τ ).

[0050] (2) Calculate the distance (such as Euclidean distance) of the above T-(m-1)τ vectors. Suppose the i-th vector x i and the jth vector x j If the distance is less than a predetermined threshold r, then R ij =1; otherwise R ij = 0 After all vectors are compared pairwise, the recursive graph matrix RP is obtained. RP is a 0 / 1 binary matrix with T-(m-1)τ in both rows and columns; (3) Calculate the relevant indicators of RQA analysis from the recursive graph matrix. See the recursive graph diagram for details. Figure 4 In the figure, the binary recursive graph matrix RP, black and white, 0 represents black, and 1 represents white.

[0051] Commonly used indicators of RQA:

[0052] (1) Recurrence rate (RR) is a measure of the density of recurrence points.

[0053]

[0054] N is the number of rows / columns of the recursive graph RP, that is, T-(m-1)τ.

[0055] (2) Determinism (DET) is a measure of the certainty of a system.

[0056]

[0057] P(l) is the frequency of the line in RP that is parallel to the diagonal and has a length of l, l min is the minimum value of l given in advance (usually 2).

[0058] (3)ENTR is the Shannon entropy of the frequency distribution of the line with diagonal length l.

[0059]

[0060] ENTR is considered to be a complexity calculation of deterministic structures in dynamic systems. The more complex the deterministic structure, the larger the ENTR value.

[0061] (4) Average diagonal length L.

[0062] Extraction of whole-brain spatial complexity index

[0063] Traditional functional connectivity is affected by the reference electrode and volume conduction effects. The number of calculated indicators is large, and there is a lot of redundant information that makes the analysis complicated. Even indicators based on phase synchronization cannot eliminate the influence of data amplitude. The basic idea of ​​spatial complexity is that if the EEG data of each channel are completely correlated (that is, the correlation coefficients are all 1 or -1), then the data of all electrodes can be considered to constitute a spatial pattern; if the EEG data of each channel are completely uncorrelated (that is, the correlation coefficients are all 0), then the number of spatial patterns formed by all electrode data is the largest (the spatial complexity is the largest).

[0064] The overall steps of the spatial complexity indicator Omega complexity are: (1) perform reference electrode conversion on the EEG data and convert it into a whole-brain average re-reference; (2) perform principal component analysis on the EEG signal with the channel as the independent variable. Assuming the number of channels is m, m mutually orthogonal principal components will be obtained. Each principal component has an eigenvalue, which indicates the proportion of the total variance explained by the principal component. The eigenvalues ​​are standardized (each eigenvalue is divided by the sum of all eigenvalues) to obtain the standardized eigenvalues.

[0065]

[0066] If the standardized eigenvalue of one principal component is equal to 1 and the standardized eigenvalues ​​of the other principal components are equal to 0, then all channel EEG data can be represented by one principal component, there is only one spatial pattern, and the spatial complexity is the lowest; if the standardized eigenvalues ​​of all principal components are 1 / m, then all channel EEG data need to be represented by m principal components, there are m spatial patterns, and the spatial complexity is the highest; (3) Quantify spatial complexity with the help of Shannon entropy

[0067]

[0068] The first two steps of the overall steps of standardized spatial complexity NSC are the same as the first two steps of spatial complexity indicator Omega complexity. The difference is that the third step quantifies the spatial complexity with the help of standard entropy.

[0069]

[0070] The above two types of spatial complexity are extracted for the whole brain.

[0071] Results Analysis

[0072] The obtained EEG complexity, entropy, RQA, and spatial complexity related indicators were analyzed for differences.

[0073] Difference analysis of EEG nonlinear dynamics indices among AUD patients in each group

[0074] Complexity Index

[0075] Table 1 Multiple comparisons of complexity indicators

[0076]

[0077]

[0078] Entropy index

[0079] Table 2 Multiple comparisons of entropy indicators

[0080]

[0081] RQA indicators

[0082] Table 3 Multiple comparisons of RQA indicators

[0083]

[0084]

[0085] There was no significant difference in the difference analysis of the whole brain spatial complexity indexes among the AUD patients in each group

[0086] The three groups of whole-brain spatial complexity indices were compared multiple times. mega Compared with NSC in the VBE group, VBN group and CVE group, they all showed a consistent decreasing trend, but there was no significant difference (P>0.05), as shown in Table 4.

[0087] Table 4 Multiple comparisons of whole-brain spatial complexity indices

[0088]

[0089] The present invention is based on a nonlinear dynamic analysis method, and its EEG mechanism for AUD patients is experimentally studied. According to the experimental results, the complexity, entropy, and RQA analysis between the VBE group, the VBN group, and the CVE group all showed significant difference indicators. Furthermore, in order to accurately evaluate the treatment effect of patients with alcohol use disorder (AUD), a benchmark or reference point is set, and the calculated EEG index is compared with the preset benchmark or index pattern to measure the changes before and after treatment, confirm the effectiveness of the treatment and the improvement of the patient. Through statistical verification and classification of the responses of AUD patients under different exposure conditions, a benchmark value or index pattern can be established to ensure the accuracy and reliability of the treatment effect evaluation.

[0090] The embodiment of the present invention provides an innovative technical solution, which can effectively evaluate the treatment effect of patients with alcohol use disorder AUD by setting a benchmark or reference point, performing statistical verification, and performing classification analysis. Specifically, the benchmark or reference point can be set according to the means of classifying the subconscious and conscious reactions of AUD patients under different exposure conditions, and combined with statistical verification and classification analysis, an accurate evaluation of the treatment effect of patients with alcohol use disorder AUD can be achieved. This process can use known statistical methods such as one-way analysis of variance and Pearson correlation analysis to determine the baseline value or indicator pattern of EEG indicators, and use this as a reference to confirm the effectiveness of the treatment and the improvement of the patient by comparing the changes before and after treatment. That is, the changes before and after treatment are measured by comparing the detected heart rate variability HRV indicators or electroencephalogram EEG indicators with the preset baseline value or indicator pattern. In addition, by applying statistical methods, such as one-way analysis of variance and Pearson correlation analysis, the significance and correlation of the indicator pattern can be verified, thereby ensuring the accuracy and reliability of the treatment effect evaluation. Through this systematic solution, the present invention can effectively evaluate and judge the physiological state and treatment effect of AUD patients, so as to provide more accurate treatment and prevention strategies for clinicians.

[0091] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art of the present invention, several equivalent substitutions or obvious variations can be made without departing from the concept of the present invention, and the performance or use is the same, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. A system for evaluating the treatment effect of patients with alcohol use disorder (AUD) using EEG indicators, characterized in that: include: EEG monitoring sensors, which measure the user’s electrical brain activity; A data processing unit, connected to the EEG monitoring sensor, is used to perform the following operations: a. Obtain the user's brain electrical activity through the EEG signal measured in real time by the EEG monitoring sensor, and calculate the nonlinear dynamic indicators reflecting the brain state, wherein the nonlinear dynamic indicators include: one or more of the EEG complexity index, the entropy index and the recursive quantitative analysis RQA index; b. Comparing the calculated nonlinear dynamic index with a preset baseline value or an index pattern under different psychological states to identify the brain activity characteristics of the patient under a specific psychological state; the preset baseline value or index pattern is based on the classification of the subconscious and conscious reactions of AUD patients under different exposure conditions, and is analyzed and verified using statistical methods; c. analyzing and determining the characteristics of the neural mechanism of the patient's brain when subconsciously processing alcohol stimulation based on the comparison results of the nonlinear dynamic index with the preset reference value or index pattern; d. Comprehensively judge the brain activity status and treatment effect of AUD patients based on the analysis results of EEG indicators; The output unit is used to output the EEG index analysis and judgment results.

2. The system according to claim 1, characterized in that The complexity index includes one or more of LZCmean, LZCmedian, LZCmid-point and LZCk-means; The entropy index includes permutation entropy PE; The recursive quantitative analysis RQA index includes one or more of the recursive rate RR, the certainty DET, the ENTR and the average diagonal length L.

3. The system according to claim 1 or 2, characterized in that: The benchmark value or indicator model is established based on the following method: Classify the subconscious and conscious responses of AUD patients under different exposure conditions; based on the classification results, establish EEG indicator patterns corresponding to subconscious processing of alcohol stimuli, subconscious processing of neutral stimuli, and conscious processing of alcohol stimuli; Statistical methods, such as one-way analysis of variance and Pearson correlation analysis, were used to verify the significance and relevance of the indicator patterns.

4. The system according to any one of claims 1 to 3, characterized in that: The processing of the data processing unit includes: Filter and baseline correct the EEG signal; Segment and denoise the EEG signal; Independent component analysis (ICA) was performed on the EEG signal to remove artifacts; Perform complexity analysis, entropy analysis, and recursive quantitative analysis (RQA) on EEG signals.

5. The system according to any one of claims 1 to 4, characterized in that: The output unit comprises: Display screen, used to display EEG index analysis and judgment results.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, data processing of the system according to any one of claims 1 to 5 is realized.

7. A computer program product, characterized in that When the computer program product is executed by a processor, the computer program product implements data processing of the system according to any one of claims 1 to 5.

Images