Drinking degree judgment method and device based on electroencephalogram analysis

Abstract

The invention provides a drinking degree judgment method and device based on electroencephalogram analysis. The method comprises the steps that electroencephalogram data before drinking and electroencephalogram data after drinking are collected; using a neural network algorithm in a machine learning algorithm to construct a head-to-head degree determination model; and according to an output result of the judgment model and a corresponding scale subjective score, formulating a judgment standard of the drinking degree. According to the method, electroencephalogram data are collected, fine processing and feature extraction are carried out on electroencephalogram signals through an advanced data analysis technology, and the correlation between brain neural activity and the head-up degree can be accurately captured. A judgment model constructed by applying a neural network algorithm is trained and optimized through a large amount of data, electroencephalogram features can be more accurately converted into a quantitative judgment result of the head-to-head degree, and accurate grading of the head-to-head degree is achieved.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary technical field of electronic monitoring, data analysis, and wine tasting. Specifically, it relates to a method and device for determining the degree of intoxication caused by drinking based on electroencephalography (EEG) analysis. Background Technology

[0002] In current alcohol tasting and research, the assessment of the degree of intoxication after drinking primarily relies on subjective experience rating scales. Common rating scales include alcohol hangover severity scales, acute hangover scales, hangover symptom scales, and post-drinking experience rating scales. For example, the alcohol hangover severity scale asks participants to rate their own experiences across multiple dimensions, such as headache, fatigue, thirst, and dizziness, to assess the severity of the hangover and reflect the degree of intoxication. Similarly, the acute hangover scale sets corresponding scoring standards around various discomfort symptoms after drinking, allowing participants to quantify their own experiences. The application of these scales provides a data foundation for research and evaluation of alcohol-related intoxication to a certain extent.

[0003] Evaluation methods based on subjective feelings have significant limitations. Individuals vary greatly in their ability to express and understand discomfort symptoms, making evaluation results susceptible to subjective bias. For example, outgoing and articulate individuals may describe their discomfort in more detail, while introverted or less sensitive individuals may underestimate their level of intoxication. Furthermore, the lack of objective quantitative indicators makes the assessment of intoxication levels imprecise. In practical applications, it is difficult to accurately compare the intoxication levels of different individuals and to deeply explore the underlying mechanisms of alcohol-induced intoxication. Moreover, the evaluation process of these subjective scales is cumbersome, requiring a certain level of comprehension and cooperation from participants, which is not conducive to large-scale rapid testing and analysis. Therefore, there is an urgent need for an objective, quantitative, accurate, and efficient method for assessing the degree of alcohol-induced intoxication. Summary of the Invention

[0004] This application provides a method and device for determining the degree of intoxication after drinking based on electroencephalogram (EEG) analysis. It acquires EEG data through electronic monitoring technology and processes and analyzes the data using data analysis technology, thereby realizing the technology for determining the degree of intoxication after drinking.

[0005] In a first aspect, embodiments of this application provide a method for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis, including: S1, collect EEG data before and after drinking; S2 uses neural network algorithms in machine learning to build a model for determining the degree of intoxication; S3. Based on the output of the judgment model and the corresponding subjective scores of the scale, formulate the judgment criteria for the degree of intoxication caused by drinking.

[0006] In step S1, an intelligent EEG testing device is used to collect EEG data. This device includes an electrode cap, a signal acquisition unit, and a data transmission module. The signal acquisition unit amplifies and performs preliminary processing on the microvolt-level EEG signals transmitted from the electrodes. The data transmission module uses wireless Bluetooth transmission technology to transmit the collected and processed EEG signals to a data analysis terminal in real time. The subject wears the adjusted electrode cap of the intelligent EEG testing device and collects EEG data over a period of time in a quiet environment. This data is used as the baseline EEG data before drinking and is stored in a dedicated database on a computer. Each subject's baseline EEG data corresponds to a unique identifier for easy data comparison and analysis later.

[0007] The process in step S1 further includes continuously collecting EEG data at certain time points after the subject finishes drinking alcohol, extracting EEG characteristic parameters related to the degree of intoxication caused by alcohol consumption, and standardizing the extracted EEG characteristic parameters. The EEG characteristic parameters include EEG rhythm characteristics, inter-brain synchrony characteristics, and event-related potential characteristics. The EEG rhythm characteristics include the average amplitude and relative power of delta waves in each brain region, the average amplitude and relative power of theta waves in each brain region, the average amplitude and relative power of alpha waves in each brain region, the peak value of alpha wave frequency, and the average amplitude and relative power of beta waves in each brain region. The inter-brain synchrony characteristics include calculating the coherence coefficient of EEG signals in different brain regions, such as the frontal lobe and parietal lobe, and the temporal lobe and occipital lobe. The event-related potential characteristics include presenting simple visual stimuli to the subject via computer during the EEG data collection process at various time points after drinking alcohol consumption, collecting the subject's event-related potentials in response to the visual stimuli, and extracting the latency and amplitude of the P300 component in the potentials.

[0008] In step S1, the subject is also assessed using an acute alcohol effect scale to evaluate the degree of intoxication. Subjects are asked to rate their subjective feelings on a scale of 0-10. The subjective ratings are stored together with the corresponding EEG data for subsequent validation of the model.

[0009] Step S2 includes collecting a large number of EEG data from different subjects after drinking alcohol and their corresponding subjective evaluation results of intoxication as training samples. The effective experimental data after preprocessing and feature extraction are randomly divided into training set and test set in a 7:3 ratio. A backpropagation (BP) neural network is selected as the model for judging the degree of intoxication after drinking alcohol. The number of nodes in the input layer of the BP neural network is the number of extracted EEG feature parameters. Two hidden layers are set, with 16 nodes in the first hidden layer and 8 nodes in the second hidden layer. The number of nodes in the output layer is 1, and the output value is a value between 0 and 10. The mean squared error is selected as the loss function of the model. The Adam optimization algorithm is used to optimize the model parameters and continuously adjust the weights and thresholds of the neural network so that the model can accurately judge the degree of intoxication based on the extracted EEG features. During the training process, the loss value of the model on the training set and validation set is monitored regularly. When the loss value on the validation set no longer decreases after 100 consecutive rounds, the training is stopped.

[0010] Step S2 further includes inputting the test set data into the trained BP neural network judgment model to obtain the model output of the intoxication level prediction score, comparing and analyzing the prediction score with the corresponding subject AHSS subjective score, and calculating the correlation coefficient and mean absolute error between the two.

[0011] In step S3, the degree of headache is divided into four levels: no headache, mild, moderate and severe.

[0012] Secondly, embodiments of this application provide a device for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis, comprising: The data acquisition unit is used to collect EEG data before and after drinking alcohol. The building block is used to construct a model for determining the degree of intoxication using neural network algorithms in machine learning. The scoring unit is used to determine the degree of intoxication caused by drinking based on the output of the judgment model and the corresponding subjective rating scale.

[0013] Thirdly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0014] Fourthly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0015] The method and device for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis in this application have the following beneficial effects: This application utilizes electroencephalography (EEG) devices to monitor the degree of intoxication caused by alcohol consumption. These devices can objectively and accurately collect EEG data from consumers after drinking. Advanced data analysis techniques are used to process and analyze the data in depth, thereby achieving a quantitative assessment of the degree of intoxication. This addresses the shortcomings of existing subjective evaluation methods and provides a scientific and reliable basis for alcohol quality evaluation, research and development, and healthy drinking practices for consumers. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the method for determining the degree of intoxication caused by drinking based on electroencephalogram (EEG) analysis, as described in an embodiment of this application. Figure 2 This is a schematic diagram of the structure of the alcohol intoxication determination device based on electroencephalogram (EEG) analysis according to an embodiment of this application. Detailed Implementation

[0017] The present application will be further described below with reference to the accompanying drawings and embodiments.

[0018] In the following description, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. The following description provides multiple embodiments of the invention, which can be substituted or combined with each other. Therefore, this application can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then this application should also be considered to include embodiments containing one or more other possible combinations of features A, B, C, and D, even if such embodiments are not explicitly described in the following text.

[0019] Example 1 like Figure 1 As shown in the embodiment of this application, a method for determining the degree of intoxication caused by drinking based on electroencephalography (EEG) analysis is provided, including: S1, collecting EEG data before drinking and EEG data after drinking; S2, constructing an intoxication degree determination model using a neural network algorithm in machine learning; S3, formulating a determination standard for the degree of intoxication caused by drinking based on the output results of the determination model and the corresponding subjective rating scale.

[0020] Compared to existing methods for assessing the degree of intoxication from alcohol consumption based on subjective feeling evaluation scales, which rely heavily on the subject's self-perception and expression and are easily influenced by individual subjective factors, this application utilizes intelligent consumer EEG testing equipment to directly collect brain electrical activity data. This data objectively reflects brain neural activity and is unaffected by subjective factors such as individual expressive ability and cognitive differences, thus ensuring the objectivity and reliability of the evaluation results. The judgment model constructed using neural network algorithms, trained and optimized with extensive data, can more accurately convert EEG characteristics into quantitative judgments of the degree of intoxication.

[0021] Example 2 Existing evaluation scales lack objective quantitative indicators, making it difficult to accurately assess the degree of "overbearing" or "overbearing."

[0022] This application provides a method for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis. In some embodiments, the specific operation is as follows: 1 device The Fermentation Institute's intelligent consumer EEG testing device used in this application is based on the principles of electroencephalography (EEG) technology. It records the spontaneous electrical activity of the brain over a period of time by placing multiple electrodes on the scalp. Its main components include an electrode cap, a signal acquisition unit, and a data transmission module.

[0023] The signal acquisition unit features high sensitivity and high resolution, enabling amplification and preliminary processing of microvolt-level EEG signals transmitted from the electrodes, ensuring signal accuracy and stability. The data transmission module utilizes wireless Bluetooth technology to transmit the acquired and processed EEG signals to the data analysis terminal in real time, achieving rapid and stable data transmission for convenient subsequent analysis and processing.

[0024] 2 Monitoring Process Experimental Preparation: The experiment should be conducted in a quiet, comfortable environment with soft lighting, maintaining an ambient temperature of approximately 25℃ and a relative humidity of 40%-60% to minimize interference from external factors on the subjects' EEG signals. Participants must undergo rigorous screening, requiring them to be between 25 and 45 years old, in good health, with no history of brain disease or mental illness, and with similar alcohol consumption history to ensure the consistency and reliability of the experimental data.

[0025] Pre-drinking EEG data collection: Subjects were seated comfortably in a chair and fitted with the electrode caps of the intelligent consumer EEG testing device. They were kept calm and relaxed, avoiding complex thinking and large physical movements. 5-10 minutes of baseline EEG data were collected; this data will serve as a benchmark for subsequent comparative analysis to determine changes in brain electrical activity after drinking.

[0026] Drinking Procedure: Subjects were instructed to drink a fixed-proof spirit in a prescribed amount. They were to sip slowly, at a rate of 10 ml every 5 minutes, to ensure stable absorption and metabolism of the alcohol. Throughout the drinking process, the subjects' basic physiological indicators, such as heart rate and blood pressure, were continuously monitored to ensure safety.

[0027] Continuous EEG data collection after alcohol consumption: EEG data was continuously collected from subjects at multiple time points, including 15 minutes, 30 minutes, 1 hour, 2 hours, and 3 hours after drinking, with each collection session lasting 5 minutes. By analyzing the EEG data at different time points, a comprehensive understanding of the changes in brain electrical activity over time after drinking can be obtained, thereby more accurately determining the degree of intoxication.

[0028] 3. Data Analysis and Processing Feature Extraction: This involves extracting EEG parameters related to intoxication, focusing on amplitude and frequency changes in brain regions such as the prefrontal and temporal lobes. For example, analyzing the amplitude decrease of alpha waves (8-13Hz) in different brain regions and the amplitude increase of beta waves (1-30Hz). The study also examines changes in energy distribution across different frequency bands, calculating the proportions of delta waves (0-4Hz), theta waves (4-8Hz), alpha waves, and beta waves in the total energy. These features effectively reflect changes in brain neural activity after alcohol consumption and are closely related to the degree of intoxication.

[0029] Establish a judgment model: A neural network algorithm from machine learning was used to construct a model for judging the degree of intoxication. A large amount of post-drinking EEG data from different subjects and their corresponding subjective evaluations of intoxication (e.g., obtained through existing evaluation scales) were collected as training samples. By continuously adjusting the weights and thresholds of the neural network, the model was able to accurately determine the degree of intoxication based on the extracted EEG features, classifying the degree of intoxication into four levels: no intoxication, mild, moderate, and severe. During training, cross-validation was used to continuously optimize the model's performance and improve the accuracy and reliability of the judgment.

[0030] This application utilizes advanced data analysis technology to finely process and extract features from electroencephalogram (EEG) signals, accurately capturing the correlation between brain neural activity and the degree of mental exertion. The judgment model constructed using neural network algorithms, trained and optimized with extensive data, can more accurately convert EEG features into quantitative judgments of mental exertion levels, achieving precise grading of mental exertion severity.

[0031] Example 3 Existing subjective evaluation methods mainly rely on the subjects' self-perception and expression, which are easily affected by individual subjective factors. They are usually evaluated by the subjects recalling and filling out scales some time after drinking, and cannot reflect the dynamic changes in the brain state during drinking in real time.

[0032] This application provides a method for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis, as follows: 1. Experimental Preparation 1.1 Experimental Environment Setup The laboratory must meet the following conditions: temperature controlled between 22-25℃, humidity maintained between 40%-60%, and the interior decorated with soundproofing materials to ensure a noise level below 30 decibels, avoiding interference from external environmental factors on the subjects' EEG signals. The laboratory interior is simply furnished, containing only necessary experimental equipment (such as intelligent EEG assessment equipment and computers) and subject chairs. The chairs are ergonomically designed to ensure the subjects remain comfortable during the experiment, reducing fluctuations in EEG signals caused by physical discomfort.

[0033] 1.2 Experimental Equipment Debugging This embodiment utilizes the intelligent consumer EEG testing device from the Fermentation Research Institute. This device consists of an EEG electrode cap, a signal amplifier, a data transmission module, and data analysis software. Before the experiment, the device underwent comprehensive testing and debugging. Electrode cap inspection and calibration: Check that the electrodes on the electrode cap are intact and that the conductive paste at the contact point between the electrodes and the scalp is sufficient and even. Select a staff member as a simulated subject. After wearing the electrode cap, use the device's built-in calibration program to ensure that the impedance value of each electrode is below 5kΩ. If the impedance value of any electrode is too high, the electrode position needs to be readjusted or conductive paste needs to be added until the impedance value meets the requirements, ensuring stable and accurate acquisition of EEG signals.

[0034] Signal amplifier calibration: Connect the signal amplifier to the electrode cap and computer, turn on the amplifier power, and set the amplifier's sampling frequency to 250Hz, gain to 10000, and filtering range to 0.5-30Hz to filter out low-frequency drift and high-frequency noise, ensuring the quality of the acquired EEG signals. Verify the accuracy of the amplifier's output signal by inputting a standard signal. If deviations are found, adjust the amplifier parameters until they meet the standard.

[0035] Data transmission and software testing: Start the data transmission module and check the stability of its wireless communication connection with the computer to ensure that EEG data can be transmitted to the computer in real time without loss. Open the data analysis software and test its data reception, storage, and preliminary processing functions, such as whether the data waveform display is normal and whether the data storage format is correct, to ensure that the software can run normally.

[0036] 1.3 Subject Screening and Grouping According to the experimental requirements, 100 healthy volunteers aged 25-45 years were selected as subjects. The selection criteria are as follows: No history of brain diseases (such as epilepsy, cerebral infarction, etc.), mental illnesses (such as depression, anxiety, etc.), or serious heart, lung, liver, or kidney diseases.

[0037] No medications affecting the central nervous system (such as sedatives, hypnotics, or antipsychotics) have been taken in the past three months.

[0038] Similar drinking histories, no more than 3 times a week, and each drink (in pure alcohol) no more than 25g, to avoid large differences in alcohol tolerance caused by long-term heavy drinking or no drinking.

[0039] Non-smokers or those who smoke no more than 5 cigarettes per day and have not smoked within 12 hours before the experiment were selected to reduce the impact of nicotine on EEG signals.

[0040] No coffee, strong tea, or other beverages containing stimulating ingredients were consumed within 24 hours prior to the experiment.

[0041] One hundred qualified participants were randomly divided into five groups of 20 each, labeled as experimental groups 1-5. There were no significant differences among the groups in terms of age, gender, weight, drinking history, and other basic information (P>0.05), ensuring the comparability of the experimental results.

[0042] 2. Experimental Procedure 2.1 EEG data collection before drinking alcohol Each group of participants was asked to sit quietly in the laboratory for 10 minutes to acclimatize to the environment and alleviate any anxiety caused by the change in environment. During this time, the experimental procedures, precautions, and methods for wearing the EEG device were explained to the participants in detail, and their questions were answered to ensure their active cooperation in the experiment.

[0043] Assist the subject in wearing the adjusted electrode cap of the intelligent EEG testing device, and check the electrode impedance values ​​again to confirm that all electrode impedance values ​​meet the requirements.

[0044] Subjects were instructed to remain relaxed with their eyes closed, and their EEG data was continuously collected for 10 minutes in a quiet environment. This data was used as the baseline EEG data before drinking and stored in a dedicated computer database. Each subject's baseline EEG data corresponded to a unique identifier, facilitating subsequent data comparison and analysis.

[0045] 2.2 Control of the drinking process To ensure the consistency of experimental variables, all subjects in each group drank 100 ml of the same brand and batch of 52-degree strong-aroma baijiu.

[0046] To avoid adverse reactions such as acute alcohol poisoning caused by excessive drinking in a short period of time, and to ensure that the alcohol is absorbed evenly in the body, the subjects were required to drink the allocated liquor at a steady pace within 10 minutes. The entire drinking process was supervised by the experimenters, who recorded the drinking time and amount of alcohol consumed to ensure that each subject completed the drinking as required.

[0047] 3. Continuous collection of EEG data after alcohol consumption After the subjects finished drinking alcohol, EEG data were continuously collected at the following time points: 15 minutes after drinking: Subjects were kept in the same relaxed state as before drinking (eyes closed, sitting quietly) and EEG data were collected for 5 minutes.

[0048] 30 minutes after drinking: Subjects were kept relaxed and EEG data were collected for 5 minutes.

[0049] One hour after drinking: Collect 5 minutes of EEG data, with the same collection conditions as the previous two times.

[0050] Two hours after drinking: Collect 5 minutes of EEG data.

[0051] Three hours after drinking: Collect 5 minutes of EEG data.

[0052] Before each EEG data collection, the electrode cap fitting and electrode impedance values ​​must be checked. If loose electrodes or abnormal impedance values ​​are found, adjustments and calibrations are made promptly to ensure the accuracy and reliability of the collected EEG data. Each EEG data collection is stored along with the subject's identification and collection time, establishing a complete subject EEG data file.

[0053] Simultaneously, after each EEG data acquisition point, the Acute Alcohol Effect Scale was used to assess the subjective degree of intoxication in the subjects. Subjects rated their own sensations (e.g., dizziness, headache, nausea, fatigue) on a scale of 0-10, with 0 indicating no intoxication and 10 indicating extremely severe and unbearable intoxication. The subjective ratings, along with the corresponding EEG data, were stored for subsequent validation of the assessment model.

[0054] 4. Feature Extraction For valid EEG data, EEG feature parameters related to the degree of intoxication caused by alcohol consumption were extracted, specifically including the following categories: EEG rhythm characteristics: Delta waves (0.5-4Hz): The average amplitude and relative power of delta waves in each brain region (the proportion of delta wave power to total brain electrical power) are extracted. After drinking alcohol, as alcohol is metabolized in the body and its inhibitory effect on the central nervous system increases, the amplitude and relative power of delta waves may change, and the degree of change is related to the degree of intoxication.

[0055] Theta waves (4-8Hz): The average amplitude and relative power of theta waves in each brain region were also extracted. Theta waves are usually associated with the brain's relaxation state and mood regulation. When intoxicated after drinking alcohol, the characteristic parameters of theta waves may show abnormal fluctuations.

[0056] Alpha waves (8-13Hz): The average amplitude, relative power, and peak frequency of alpha waves in each brain region were extracted. Alpha waves are the main electroencephalographic rhythm of the brain in a relaxed state. After drinking alcohol, alcohol inhibits alpha wave activity, and changes in their amplitude, power, and peak frequency can serve as important indicators for determining the degree of intoxication.

[0057] Beta waves (13-30Hz): The average amplitude and relative power of beta waves in each brain region were extracted. Beta waves are related to cognitive activity and concentration. After drinking alcohol, the intoxication can lead to a decline in cognitive function, and the characteristic parameters of beta waves will also change accordingly.

[0058] Synchronicity characteristics between brain regions: The coherence coefficient of EEG signals from different brain regions (such as the frontal and parietal lobes, temporal and occipital lobes, etc.) is calculated. The coherence coefficient reflects the degree of synchronization of EEG activity between two brain regions. After drinking alcohol, due to the effects of alcohol on the central nervous system, information transmission between brain regions may be hindered, leading to a decrease in inter-brain synchronicity. Changes in the coherence coefficient can serve as an auxiliary indicator for determining the degree of intoxication.

[0059] Event-Related Potential Characteristics: During the EEG data collection at various time points after alcohol consumption, simple visual stimuli were presented to the subjects via computer, and event-related potentials (ERPs) in response to these stimuli were collected. The latency and amplitude of the P300 component (a positive potential with a latency of approximately 300 ms) within the ERPs were extracted. The P300 component is closely related to cognitive processing in the brain (such as attention, memory, and decision-making). After drinking, the onset of intoxication slows down cognitive processing, prolongs the P300 latency, and reduces its amplitude. The degree of change in these changes directly reflects the degree of intoxication.

[0060] The extracted EEG feature parameters were standardized to eliminate differences in dimensions and orders of magnitude between different feature parameters, laying the foundation for the subsequent establishment of a judgment model.

[0061] 5. Judgment Model Establishment and Validation 5.1 Dataset Splitting: The effective experimental data after preprocessing and feature extraction (containing complete data from 78 subjects, with data from 22 subjects removed due to substandard quality) were randomly divided into training and test sets in a 7:3 ratio. The training set contained data from 55 subjects (55 × 6 time points = 330 data sets) and was used for model training; the test set contained data from 23 subjects (23 × 6 time points = 138 data sets) and was used for model validation.

[0062] 5.2 Model Selection and Training: In this embodiment, a BP (Back Propagation) neural network is selected as the model for determining the degree of intoxication caused by drinking. This network has strong nonlinear mapping ability and self-learning ability, and can handle the complex relationship between EEG feature parameters and the degree of intoxication well.

[0063] Network structure design: The number of nodes in the input layer of the BP neural network is the number of extracted EEG feature parameters (in this embodiment, a total of 28 feature parameters are extracted, so the number of nodes in the input layer is 28); there are 2 hidden layers, with 16 nodes in the first hidden layer and 8 nodes in the second hidden layer; the number of nodes in the output layer is 1, and the output value is a value between 0 and 10, representing the subject's alcohol-induced intoxication score.

[0064] Model training parameter settings: Mean squared error (MSE) was selected as the model's loss function. The Adam optimization algorithm was used to optimize the model parameters. The learning rate was set to 0.001, the number of training epochs was set to 1000, and the batch size for each training iteration was set to 32. During training, the model's loss value on the training set and the validation set (20% of the data in the training set was used as the validation set) was monitored periodically. Training was stopped when the validation set loss value no longer decreased after 100 consecutive epochs to avoid overfitting.

[0065] 5.3 Model Validation: The test set data is input into the trained BP neural network judgment model to obtain the predicted score of the degree of intoxication output by the model. The predicted score is compared with the corresponding subject AHSS subjective score, and the correlation coefficient (Pearson correlation coefficient) and mean absolute error (MAE) of the two are calculated.

[0066] Correlation coefficient calculation: The Pearson correlation coefficient between the predicted score and the subjective score was calculated to be 0.89, indicating that the two have a strong positive correlation, which shows that the judgment model can reflect the actual degree of intoxication caused by drinking in the subjects.

[0067] Mean absolute error calculation: The mean absolute error between the predicted score and the subjective score was calculated to be 0.62 points, indicating that the model has high prediction accuracy and can meet the actual needs of judging the degree of intoxication caused by drinking.

[0068] 6. Establishment of criteria for judging the degree of intoxication Based on the output of the trained neural network judgment model and the corresponding subjective ratings on the scale, combined with statistical analysis of a large amount of experimental data, the criteria for judging the degree of intoxication caused by drinking were formulated as follows: No headache (0-2 points): The model output score is in the range of 0-2 points, and the subject's subjective score is also in the range of 0-2 points. At this time, the subject has no discomfort symptoms such as dizziness, headache, or nausea. The subject is clear-headed, has normal cognitive function, and the EEG characteristics are low relative power of delta waves and theta waves, high relative power of alpha waves and stable frequency, normal relative power of beta waves, high brain region coherence coefficient, normal P300 latency and high amplitude.

[0069] Mild dizziness (2-4 points): The model output score is in the range of 2-4 points, and the subject's subjective score is also in the range of 2-4 points. The subject occasionally experiences mild dizziness, but without symptoms such as headache or nausea. Consciousness remains clear, and cognitive function is basically normal. EEG characteristics show a slightly increased relative power of delta waves and theta waves, a slightly decreased relative power of alpha waves, a slightly decreased relative power of beta waves, a slightly decreased brain region coherence coefficient, a slightly prolonged P300 latency, and a slightly decreased amplitude.

[0070] Moderate dizziness (4-7 points): The model output score is in the range of 4-7 points, and the subject's subjective score is also in the range of 4-7 points. The subject experiences obvious dizziness and headache, possibly accompanied by mild nausea and fatigue. The clarity of consciousness is somewhat reduced, and cognitive functions (such as attention and memory) are somewhat impaired. EEG characteristics show significantly increased relative power of delta waves and theta waves, significantly decreased relative power of alpha waves with large frequency fluctuations, significantly decreased relative power of beta waves, significantly reduced brain region coherence coefficient, and significantly prolonged P300 latency and significantly reduced amplitude.

[0071] Severe dizziness (7-10 points): The model output score is in the range of 7-10 points, and the subject's subjective score is also in the range of 7-10 points. The subject experiences severe dizziness, headache, nausea, vomiting, and other symptoms, confusion, severe cognitive impairment, and even instability and slurred speech. EEG characteristics show a significant increase in the relative power of delta and theta waves, a significant decrease in the relative power of alpha waves (or even near disappearance), a significant decrease in the relative power of beta waves, a significant decrease in the brain region coherence coefficient, a significant prolongation of P300 latency, a significant decrease in amplitude, and even unrecognizable P300 waves.

[0072] When applying this method to determine the degree of intoxication caused by alcohol consumption, it is only necessary to collect, preprocess, and extract features from the subject's EEG data according to the experimental procedure described above, and then input it into the trained neural network judgment model. Based on the score output by the model and compared with the judgment criteria described above, the degree of intoxication caused by alcohol consumption can be determined quickly and accurately.

[0073] Compared with existing methods for assessing alcohol-induced intoxication based on subjective feeling rating scales, this application has several significant advantages: Regarding objectivity, existing subjective evaluation methods mainly rely on the subjects' self-perception and expression, which are easily influenced by individual subjective factors. In contrast, this invention utilizes an intelligent consumer EEG evaluation device to directly collect brain electrical activity data. This data is an objective reflection of brain neural activity, unaffected by subjective factors such as individual expressive ability and cognitive differences, thus ensuring the objectivity and reliability of the evaluation results.

[0074] In terms of accuracy, existing evaluation scales lack objective quantitative indicators, resulting in a vague assessment of the degree of "hysteria" and making precise evaluation difficult. This invention utilizes advanced data analysis technology to finely process and extract features from electroencephalogram (EEG) signals, accurately capturing the correlation between brain neural activity and the degree of "hysteria." The judgment model, constructed using a neural network algorithm and trained and optimized with extensive data, can more accurately convert EEG features into quantitative judgments of "hysteria" degree, achieving precise grading of "hysteria" severity.

[0075] From a real-time perspective, existing methods typically assess alcohol consumption some time after drinking by having subjects recall and complete scales, failing to reflect the dynamic changes in brain state during the drinking process. This invention, leveraging the real-time data acquisition and transmission capabilities of an intelligent consumer EEG testing device, can continuously monitor changes in EEG signals at various time points after drinking. Timely analysis of this real-time data provides strong support for studying the evolution of post-drinking intoxication levels over time and allows researchers to grasp changes in subjects' brain state immediately.

[0076] This invention provides a more reliable basis for alcoholic beverage production and research. For alcoholic beverage producers, this invention enables a more accurate understanding of the effects of different alcoholic beverages on consumers' brains, allowing for targeted optimization of production processes, adjustment of alcoholic beverage composition and proportions, reduction of the likelihood of products causing intoxication in consumers, and improvement of product quality and market competitiveness. In the field of alcoholic beverage research, the objective and accurate method for determining the degree of intoxication provided by this invention helps to delve deeper into the physiological mechanisms of alcohol-induced intoxication, promoting the development of alcoholic beverage scientific research.

[0077] Example 4 like Figure 2As shown in the figure, this application provides a device for determining the degree of intoxication caused by drinking based on electroencephalogram (EEG) analysis, including: a data acquisition unit 201 for acquiring EEG data before drinking and EEG data after drinking; a data construction unit 202 for constructing a model for determining the degree of intoxication using a neural network algorithm in machine learning; and a scoring unit 203 for formulating a standard for determining the degree of intoxication caused by drinking based on the output results of the determination model and the corresponding subjective scores on a scale.

[0078] In this application, the embodiment of the alcohol intoxication determination device based on EEG analysis is basically similar to the embodiment of the alcohol intoxication determination method based on EEG analysis. For relevant details, please refer to the description of the alcohol intoxication determination method based on EEG analysis.

[0079] This invention also provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, this program implements the steps of the above-described method for determining the degree of intoxication caused by alcohol consumption based on electroencephalogram (EEG) analysis. The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs, CD-ROMs, microdrives, magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.

[0080] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for determining the degree of alcoholism based on electroencephalographic analysis, characterized by, The method comprises the following steps: S1, collecting pre-drinking and post-drinking brain electrical data; S2, using a neural network algorithm in a machine learning algorithm to build a degree of drunkenness determination model; S3, formulating a determination standard for the degree of drunkenness according to the output result of the determination model and the corresponding subjective scale score.

2. The method of claim 1, wherein the degree of alcoholism is determined by analyzing the electroencephalogram. In step S1, the electroencephalogram data is collected by using an intelligent electroencephalogram evaluation device, which comprises an electrode cap, a signal collector and a data transmission module; the signal collector amplifies and preliminarily processes the microvolt-level electroencephalogram signals from the electrodes; The data transmission module uses wireless Bluetooth transmission technology to transmit the collected and processed electroencephalogram signals in real time to a data analysis terminal; the subject wears the adjusted intelligent electroencephalogram evaluation device electrode cap, and the electroencephalogram data in a period of time in a quiet environment is collected as the pre-drinking basic electroencephalogram data, which is stored in a special database of a computer; the basic electroencephalogram data of each subject corresponds to a unique identity mark, which is convenient for subsequent data comparison and analysis.

3. The method according to claim 1, wherein the method is characterized by: In step S1, after the subject finishes drinking, the electroencephalogram data is continuously collected according to certain time nodes, the electroencephalogram feature parameters related to the degree of drunkenness are extracted, and the extracted electroencephalogram feature parameters are standardized; the electroencephalogram feature parameters include electroencephalogram rhythm characteristics, brain region synchronization characteristics and event-related potential characteristics; the electroencephalogram rhythm characteristics include the average amplitude and relative power of the delta wave in each brain region, the average amplitude and relative power of the theta wave in each brain region, the average amplitude, relative power and alpha wave frequency peak value of the alpha wave in each brain region, and the average amplitude and relative power of the beta wave in each brain region; the brain region synchronization characteristics include the coherence coefficients of the electroencephalogram signals of different brain regions, the frontal lobe and the parietal lobe, the temporal lobe and the occipital lobe; the event-related potential characteristics include the latency and amplitude of the P300 component in the event-related potential of the subject to the visual stimulus collected by the computer to the subject during the electroencephalogram data collection process at each time node after drinking.

4. The method according to any one of claims 1 to 3, wherein the method is characterized by, In step S1, the subject is scored for the subjective degree of drunkenness by using an acute alcohol effect scale, the subject scores in the range of 0-10 according to his own feelings, and the subjective score result is stored together with the corresponding electroencephalogram data for subsequent verification of the determination model.

5. The method of claim 1 to 3, wherein the method is characterized by, In step S2, the brain electrical data after drinking and the corresponding subjective evaluation results of the effects of the alcohol are collected from a large number of different subjects as training samples, and the effective experimental data after preprocessing and feature extraction are randomly divided into a training set and a test set in a ratio of 7:3; a BP neural network is selected as the drinking effects determination model, the number of input layer nodes of the BP neural network is the number of extracted brain electrical feature parameters, two hidden layers are set, the number of nodes of the first hidden layer is 16, the number of nodes of the second hidden layer is 8; the number of output layer nodes is 1, and the output value is a value between 0 and 10; the mean square error is selected as the loss function of the model, the Adam optimization algorithm is used to optimize the model parameters, and the weights and thresholds of the neural network are adjusted to make the model accurately determine the effects according to the extracted brain electrical features; during the training process, the loss values of the model on the training set and the validation set are monitored regularly, and when the loss value of the validation set does not decrease continuously for 100 rounds, the training is stopped.

6. The method according to any one of claims 1 to 3, wherein the method is characterized by, In step S2, the test set data is input into the trained BP neural network determination model to obtain the effects prediction score output by the model, and the prediction score is compared with the corresponding subject AHSS subjective score for analysis, and the correlation coefficient and the mean absolute error are calculated.

7. The method according to any one of claims 1 to 3, wherein the method is a method for determining the degree of alcoholism based on electroencephalogram analysis. In step S3, the effects are divided into four levels: no effects, mild effects, moderate effects, and severe effects.

8. A device for determining the degree of alcoholism based on electroencephalographic analysis, characterized by comprising: The application comprises: a collection unit for collecting pre-drinking brain electrical data and post-drinking brain electrical data; a construction unit for constructing an effects determination model using a neural network algorithm in machine learning; a scoring unit for formulating a determination standard for the effects of drinking according to the output results of the determination model and the corresponding scale subjective scores.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the steps of the method of any one of claims 1-7.

10. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the method of any one of claims 1-7.

Images