Antidepressant drug compliance management system and method based on biological feature recognition

Abstract

The invention discloses an antidepressant drug compliance management system based on biological feature recognition and a method thereof, and belongs to the field of medical health systems. The system comprises a biological characteristic acquisition module for acquiring multi-modal biological characteristic data of a drug user in real time; the identity authentication module is used for verifying the identity of a medicine user to prevent illegal use or cheating behaviors; the medicine taking behavior management module is responsible for analyzing the behavior sequence of the medicine taking person in real time, identifying the appearance characteristics of the medicine and verifying the accuracy of the medicine taking action; the data analysis module performs cleaning, feature extraction and pattern analysis on the collected data, constructs a personalized medication behavior model and predicts a medication trend; the abnormity monitoring and early warning module is used for carrying out real-time monitoring and risk assessment on medication behaviors through a comprehensive risk assessment mechanism and generating early warning information and intervention suggestions of corresponding levels; and the interaction management module provides a self-adaptive user interface and is responsible for information display, medication reminding pushing and data visual presentation.

Description

Technical Field

[0001] This application relates to the field of healthcare systems, and more particularly, to an antidepressant drug compliance management system and method based on biometric recognition. Background Art

[0002] With the increasing psychological pressure in modern society, depression has become a common mental health problem, especially prominent among young and elderly populations. Antidepressant drugs are widely used in the treatment of depression, but the issue of drug compliance has always troubled patients and medical staff. Poor drug compliance not only leads to unsatisfactory treatment effects but also may cause serious health risks such as drug abuse and overdose. Therefore, how to improve patients' medication compliance and ensure the rational use of drugs has become an urgent problem to be solved in the field of mental illness treatment.

[0003] Most existing drug compliance management methods rely on patients' self-reporting, family member monitoring, or regular follow-up visits, suffering from problems such as incomplete compliance monitoring, inaccurate information, and inadequate intervention measures. Traditional management methods not only carry the risk of patients hiding their medication situations but also cannot monitor patients' actual medication behaviors and treatment effects in real time. At the same time, traditional identity authentication methods such as passwords, ID cards, and social security cards are easily misused or forged, unable to effectively ensure the authenticity and security of drug use.

[0004] In recent years, biometric recognition technology has been widely used. Through technologies such as facial recognition, fingerprint recognition, and voice recognition, precise identity verification can be achieved. These technologies can effectively enhance the security of identity authentication, prevent drug abuse or misuse, and ensure that patients can take medications rationally according to doctors' prescriptions. However, the application of existing biometric recognition technology in drug compliance management is still in the exploratory stage, lacking a systematic and integrated solution, and unable to achieve comprehensive and multi-dimensional real-time monitoring and personalized intervention.

[0005] In addition, existing systems lack intelligent recognition and analysis of drug-taking behaviors, unable to evaluate the normativity of medication behaviors in real time, resulting in the inability to provide timely and effective feedback on whether patients take medications as required. Therefore, how to use innovative technical means to combine biometric recognition with real-time behavior analysis to improve drug compliance and ensure patients' medication safety has become an urgent technical problem to be solved. Summary of the Invention

[0006] In order to overcome a series of defects existing in the prior art, the purpose of this application is to provide an antidepressant drug compliance management system based on biometric recognition for the above problems, including the following modules:

[0007] The biometric data collection module collects multi-modal biometric data of the medication user in real time through a high-definition camera, a voice sensor, and a fingerprint recognizer;

[0008] The identity authentication module reliably verifies the identity of the medication user based on the collected multi-modal biometric data to prevent impersonation or fraud;

[0009] The medication behavior management module is responsible for analyzing the behavior sequence of the medication user in real time, identifying the appearance features of the drug, and verifying the accuracy of the medication-taking action;

[0010] The data analysis module cleans, extracts features from, and analyzes the patterns of the collected data, constructs a personalized medication behavior model, and predicts the medication trend;

[0011] The anomaly monitoring and warning module, based on the data analysis results, conducts real-time monitoring and risk assessment of the medication behavior through a comprehensive risk assessment mechanism, and generates warning information and intervention suggestions at corresponding levels;

[0012] The interaction management module provides an adaptive user interface, is responsible for information display, medication reminder push, and data visualization presentation, and realizes the permission management for different user groups.

[0013] Furthermore, the biometric data collection module includes the following components:

[0014] The high-definition camera unit uses a high-resolution camera to collect the facial features of the medication user in real time;

[0015] The voice collection unit obtains the voice data of the medication user through a high-sensitivity voice sensor;

[0016] The fingerprint recognition unit uses a high-precision fingerprint recognizer to collect the fingerprint information of the medication user;

[0017] The infrared imaging unit uses infrared imaging technology to monitor the facial temperature distribution;

[0018] The heart rate detection unit collects heart rate data through a photoplethysmogram sensor;

[0019] The posture perception unit uses an inertial sensor to detect the head and body postures of the medication user.

[0020] Furthermore, the identity authentication module includes the following components:

[0021] The identity feature integration unit is responsible for generating a unique and highly credible identity feature template for identity authentication and preventing fraud;

[0022] The feature matching unit compares the identity feature template with the registered information in the database to determine the identity consistency;

[0023] The identity confirmation unit combines the matching result and the credibility score, comprehensively evaluates the identity authentication result, and gives the final authentication decision;

[0024] The security authentication optimization unit dynamically adjusts the authentication strategy based on historical authentication records, environmental characteristics, and behavior patterns to improve the flexibility and security of identity verification;

[0025] The data encryption unit encrypts the biometric data stored and transmitted to prevent data leakage and malicious tampering, and ensure the privacy and security of users.

[0026] Furthermore, the medication behavior management module includes the following components:

[0027] The drug appearance recognition unit uses computer vision algorithms to recognize the appearance characteristics of drugs and classify them to verify whether the drugs meet the predetermined standards;

[0028] The motion capture unit captures the motion trajectory of the user's hand in real time to determine whether an accurate medication-taking action has been completed;

[0029] The medication-taking action verification unit analyzes the user's medication-taking action and compares it with the standard action template to confirm whether the user has completed the correct medication behavior;

[0030] The behavior management unit generates a personalized health report based on the results of drug appearance recognition and action verification;

[0031] The data storage and synchronization unit stores the images and action data of each medication behavior and synchronizes them to cloud or local storage for subsequent viewing and analysis.

[0032] Furthermore, the data analysis module includes the following components:

[0033] The data cleaning unit performs data cleaning operations, including noise elimination, missing value filling, and outlier detection and handling, to ensure data quality and analysis reliability;

[0034] The data integration unit realizes the integration and unification of multi-source heterogeneous data, extracts key indicators through feature engineering, and constructs a standardized analysis data set;

[0035] The pattern mining unit performs pattern recognition and clustering analysis on the user's medication behavior to discover potential behavior patterns;

[0036] The user portrait modeling unit constructs a multi-dimensional user feature vector based on the user's historical medication records and lifestyle data to form a personalized medication behavior model;

[0037] The trend prediction unit uses time series analysis and prediction algorithms to quantitatively predict the user's future medication needs and provide support for intelligent early warning and decision-making.

[0038] Furthermore, the abnormal monitoring and early warning module includes the following components:

[0039] The abnormal recognition unit analyzes the medication behavior data based on machine learning algorithms to achieve automatic detection and recognition of abnormal patterns, including monitoring of abnormal dosages, abnormal medication times, and abnormal medication combinations;

[0040] The risk assessment unit conducts quantitative analysis on the detected abnormal behaviors, evaluates their potential risks to the user's health, and classifies them into different levels;

[0041] The hierarchical early warning mechanism unit adopts a hierarchical early warning mechanism according to the risk quantification results to generate early warning information at different levels;

[0042] The early warning message processing unit generates detailed early warning information based on the evaluation results and the early warning mechanism;

[0043] The intervention suggestion generation unit automatically generates targeted medication behavior correction suggestions and intervention measures based on the user profile and the risk assessment results.

[0044] Furthermore, the interaction management module includes the following components:

[0045] The user interface adaptation unit dynamically adjusts the interface layout according to different devices and screen sizes to ensure the fluency and consistency of user operations;

[0046] The information display unit is responsible for displaying key information and real-time data to ensure that users can obtain the required information clearly and accurately;

[0047] The medication reminder push unit automatically pushes medication reminders according to the user's medication use plan to ensure that users take their medications on time and improve compliance;

[0048] The data visualization presentation unit converts complex data into charts, images, or other easy-to-understand forms, enabling users to intuitively understand the trends and patterns behind the data;

[0049] The permission management unit dynamically assigns access permissions according to different user identities and roles to ensure security and information privacy protection;

[0050] The interaction feedback unit captures the feedback information after user operations and makes real-time adjustments and optimizations to improve the user experience;

[0051] The multi-user support unit manages and supports the independent operations and data access of different user groups, providing personalized interaction experiences.

[0052] The objective of this application is also to provide an anti - depressive drug compliance management method based on biometric recognition, including the following steps:

[0053] Generate a unique identity feature template through multimodal biometric collection and perform real - time comparison with the pre - stored user profile to ensure the authenticity of the user taking the medicine and prevent drug abuse or misuse;

[0054] Identify and classify drugs through computer vision technology to ensure the accuracy of drug types and dosages; meanwhile, analyze the user's medication - taking actions using motion capture technology to verify the standardization of medication - taking behaviors;

[0055] Clean, denoise, and process outliers for the collected raw data, extract key features through a data fusion unit, and establish a personalized medication - taking behavior model to provide data support for anomaly monitoring and trend prediction;

[0056] Based on the medication - taking behavior model, monitor the user's medication - taking behavior in real - time, and start a risk assessment mechanism when a deviation from the normal pattern is detected. Trigger an alarm according to the risk level and generate intervention suggestions;

[0057] Dynamically adjust the interface layout and interaction method according to the user's habits and device characteristics, and provide personalized function access permissions for different user groups through permission management to ensure data security;

[0058] Automatically push personalized medication reminders according to the medication plan and real - time monitoring data, and at the same time collect the user's feedback to continuously improve the effect of medication compliance management.

[0059] Compared with the prior art, this application has the following beneficial effects:

[0060] This application realizes real - time monitoring, identity verification, and risk warning of the medication compliance of depression patients through multimodal biometric recognition and data analysis, effectively preventing drug abuse, misuse, and providing personalized medication guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic structural diagram of an anti - depressive drug compliance management system based on biometric recognition disclosed in an embodiment of this application.

[0062] Figure 2 It is a schematic flow diagram of an anti - depressive drug compliance management method based on biometric recognition disclosed in an embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] To make the objectives, technical solutions, and advantages of the present invention more clear, the following will describe in more detail the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. In the drawings, the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The described embodiments are some, but not all, of the embodiments of the present invention.

[0064] All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0065] The embodiments described below with reference to the accompanying drawings and directional terms are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0066] As Figure 1 shown, an anti - depression drug compliance management system based on biometric recognition includes the following modules:

[0067] The biometric data acquisition module uses a high - definition camera, a voice sensor, and a fingerprint identifier to collect multi - modal biometric data of the drug user in real - time.

[0068] The identity authentication module, based on the collected multi - modal biometric data, reliably verifies the identity of the drug user to prevent impersonation or deception.

[0069] The medication behavior management module is responsible for analyzing the behavior sequence of the drug user in real - time, identifying the appearance characteristics of the drug, and verifying the accuracy of the medication action.

[0070] The data analysis module cleans, extracts features, and analyzes patterns of the collected data, constructs a personalized medication behavior model, and predicts the medication trend.

[0071] The anomaly monitoring and warning module, based on the data analysis results, monitors the medication behavior in real - time and conducts a risk assessment through a comprehensive risk assessment mechanism, generating warning information and intervention suggestions at corresponding levels.

[0072] The interaction management module provides an adaptive user interface, is responsible for information display, medication reminder push, and data visualization presentation, and realizes the permission management of different user groups.

[0073] In this embodiment, the biometric collection module is a core component of the entire antidepressant medication compliance management system, responsible for obtaining the physiological and behavioral data of the medication user in real time. By integrating multimodal sensing devices such as high-definition cameras, voice sensors, and fingerprint recognizers, this module can capture biometric features such as the appearance, voice, and fingerprint of the medication user through precise sensing technology. These data not only help to monitor the identity information of the medication user in real time, ensuring the uniqueness and accuracy of their identity, but also enable further analysis of their physiological state and behavioral patterns, providing a basis for subsequent behavior analysis and intervention. For example, the high-definition camera can identify the user's facial features, emotional changes, and behavior during medication, the voice sensor can monitor the user's voice features to infer possible emotional fluctuations or abnormal language expressions, and the fingerprint recognizer ensures the accuracy of the medication user's identity, avoiding the risk of identity fraud at the technical level. In addition, the design of this module also takes into account low power consumption and high efficiency. Through multi-channel data fusion, it can reduce the system's demand for hardware resources while ensuring high precision. In this way, based on real-time and stable data collection, it can quickly respond to the changes of the medication user, improving the response speed and reliability of the overall system.

[0074] In this embodiment, the core objective of the identity authentication module is to ensure that only appropriate users can access the system, preventing identity fraud or deception. By closely integrating with the biometric collection module, the identity authentication module can perform precise comparison and identity verification based on the collected facial, fingerprint, and voice data. In terms of technical implementation, advanced machine learning algorithms and deep learning models are adopted to complete the efficient matching of user identities by analyzing the high-dimensional feature vectors of biometric data. Compared with traditional single-biometric authentication technologies, this multimodal verification method greatly improves the accuracy and security of verification, reducing the risk of misidentification and false judgment. Especially during the identity verification process, this module can intelligently adapt to data changes in different environments, such as light changes, voice noise, or external interference in fingerprint recognition, thus ensuring effective identity verification even in complex situations. In addition, the system will regularly update the algorithm model and optimize the authentication process to adapt to changes in the user's biometric features, such as facial feature changes due to age growth.

[0075] In this embodiment, the main task of the medication behavior management module is to analyze the behavior sequence of the medication user in real time to verify the accuracy of their medication actions and the appearance characteristics of the drugs. This module determines whether the drugs are taken correctly by accurately identifying the user's medication process and confirms whether the appearance characteristics of the drugs are consistent with the preset standards. For example, by analyzing videos to identify the shapes and colors of the drug bottles and tablets, it determines whether the user has performed accurate medication-taking operations to ensure the safety and effectiveness of medication. In addition, this module can also detect whether the user takes the drugs at the prescribed time and dose through action recognition technology, further enhancing the control of medication compliance. To achieve this function, this module integrates a deep learning model and computer vision technology, which can perform precise temporal analysis on the medication-taking actions and conduct multi-dimensional analysis of the user's medication behavior in combination with the physiological data collected by sensors. These analysis results can not only provide real-time feedback but also serve as a basis for subsequent risk warning and behavior correction in the system.

[0076] In this embodiment, the data analysis module is responsible for cleaning, feature extraction, and pattern analysis of all the collected data, thereby constructing a personalized medication behavior model and predicting the medication trend. The core technologies of this module rely on big data analysis and machine learning algorithms. By deeply analyzing historical data, it extracts the key features related to medication compliance and models them. First, data cleaning and preprocessing ensure the data quality and eliminate the interference of noisy data and outliers. Then, through feature extraction, it can refine the patterns closely related to medication behavior, such as the medication-taking time, dose, and frequency, and establish a personalized medication prediction model in combination with the user's physiological characteristics and behavior data. These models can not only accurately reflect the user's medication habits but also predict the user's future medication behavior trends, providing decision support for the system. Through continuous iteration, update, and optimization, the data analysis module can adapt to the individual differences of the medication users and the changes in their medication patterns, thereby improving the intelligence level and prediction accuracy of the system and ensuring the provision of precise management solutions in a dynamically changing environment.

[0077] In this embodiment, the abnormal monitoring and warning module is a key link to ensure the safe medication of the user. Based on the data analysis results, this module is responsible for real-time monitoring of the medication behavior and conducting risk assessment. It continuously monitors the user's behavior by constructing a comprehensive risk assessment mechanism to identify potential drug use problems, such as missed doses, overdose, or improper medication. By analyzing the user's biometric data, medication behavior, and historical records, the module can promptly identify behaviors that do not conform to the normal medication pattern and issue warning messages. The core of the warning is risk assessment, which combines machine learning algorithms to evaluate the risk level of the current behavior based on factors such as the user's behavior deviation and medication compliance score, and generates targeted intervention suggestions. For example, when it is found that the user has not taken the medicine at a specific time period, it will immediately send a reminder message and give corresponding treatment suggestions, such as medicine replenishment or further medical intervention. The design of this module not only enhances the automation level but also improves the user's compliance, thus reducing the health risks caused by improper drug use.

[0078] In this embodiment, the interaction management module is the bridge between the system and the user, responsible for providing an adaptive user interface to realize functions such as information display, medication reminder push, and data visualization presentation. This module aims to improve the user experience through a simple and intuitive interface design to ensure that the user can easily interact with the system. Through the intelligent interface, the user can easily view their medication records, receive medication reminders, and obtain personalized medication suggestions in a timely manner. In addition, through data visualization technology, complex medication data is presented in the form of charts, curves, etc., enabling the user to intuitively understand their medication trends and health status. At the same time, this module also realizes the permission management for different user groups, such as users, doctors, family members, etc. According to different roles, the information and functions displayed by the system are also different, ensuring that all parties can effectively participate in and support the treatment process of the user.

[0079] In summary, the antidepressant medication compliance management system based on biometric recognition forms an intelligent and all-round medication management mode by integrating modules such as multi-modal biometric collection, identity authentication, medication behavior management, data analysis, abnormal monitoring and warning, and interaction management. Each module cooperates with each other technically to form a closed-loop feedback mechanism to ensure that the user can receive drug treatment under efficient and accurate management. It not only improves the medication compliance but also effectively reduces the potential safety hazards that may occur during the drug use process, providing more scientific and personalized support for the treatment of diseases such as depression.

[0080] Furthermore, the biometric collection module includes the following components:

[0081] A high-definition camera unit that uses a high-resolution camera to collect the facial features of the user in real time;

[0082] The voice acquisition unit acquires the voice data of the medication user through a high-sensitivity voice sensor;

[0083] The fingerprint recognition unit collects the fingerprint information of the medication user by using a high-precision fingerprint recognizer;

[0084] The infrared imaging unit monitors the facial temperature distribution by using infrared imaging technology;

[0085] The heart rate detection unit acquires the heart rate data through a photoplethysmogram sensor;

[0086] The attitude perception unit detects the head and body postures of the medication user by using an inertial sensor.

[0087] In summary, the biometric acquisition module comprehensively collects the multi-modal biometric information of the medication user through the cooperation of multiple components such as high-definition imaging, voice acquisition, fingerprint recognition, infrared imaging, heart rate detection, and attitude perception. Each component not only undertakes specific functions such as identity authentication, emotion monitoring, and behavior verification, but also forms a mutually supportive and collaborative working mechanism in the global system. Through these highly integrated technologies, it is possible to ensure the accuracy of the medication user's identity, the standardization of medication behavior, and the stability of physiological state in multiple aspects, thereby maximizing drug compliance and treatment effects. The implementation of this module provides a new perspective for precision medicine and intelligent management, demonstrating the powerful potential of biometric recognition in modern medicine.

[0088] Furthermore, the identity authentication module includes the following components:

[0089] The identity feature integration unit is responsible for generating a unique and highly credible identity feature template for identity authentication and preventing spoofing behaviors;

[0090] The feature matching unit compares the identity feature template with the registered information in the database to determine the identity consistency;

[0091] The identity confirmation unit comprehensively evaluates the identity authentication result based on the matching result and the credibility score, and gives the final authentication decision;

[0092] The security authentication optimization unit dynamically adjusts the authentication strategy based on historical authentication records, environmental features, and behavior patterns to improve the flexibility and security of identity verification;

[0093] The data encryption unit encrypts the stored and transmitted biometric data to prevent data leakage and malicious tampering, and ensure the privacy and security of users.

[0094] In summary, the identity authentication module provides a highly reliable and flexible identity authentication mechanism through multi-level and multi-dimensional security protection and authentication strategies. The identity feature integration unit, the high-precision feature matching unit, the comprehensively evaluated identity confirmation unit, the security authentication optimization unit, and the powerful data encryption unit work together to jointly ensure the accuracy, security, and protection of user privacy in identity authentication. Through the precise capture and dynamic optimization of user identity features, it can not only ensure the efficiency and accuracy of identity verification but also adjust the authentication strategy according to different scenarios and risk levels to adapt to various usage scenarios. The addition of the data encryption unit further enhances the security of the system and ensures the integrity and security of biometric data. Overall, the identity authentication module not only improves the reliability of identity verification in the medication compliance management system but also greatly enhances the ability to prevent malicious attacks and identity forgery, providing a solid technical foundation for applications in the fields of intelligent healthcare, health management, etc.

[0095] Furthermore, the medication behavior management module includes the following components:

[0096] The drug appearance recognition unit uses computer vision algorithms to recognize the appearance features of drugs and classify them to verify whether the drugs meet the predetermined standards;

[0097] The motion capture unit captures the motion trajectory of the user's hand in real time to determine whether the accurate medication-taking action has been completed;

[0098] The medication-taking action verification unit analyzes the user's medication-taking action and compares it with the standard action template to confirm whether the user has completed the correct medication behavior;

[0099] The behavior management unit generates a personalized health report based on the results of drug appearance recognition and action verification;

[0100] The data storage and synchronization unit stores the images and action data of each medication behavior and synchronizes them to cloud or local storage for subsequent viewing and analysis.

[0101] In summary, through the combination of multiple intelligent technologies, the medication behavior management module has significantly enhanced the functions and effects of the medication compliance management system. The drug appearance recognition unit, motion capture unit, medication-taking action verification unit, behavior management unit, and data storage and synchronization unit all play important roles and work together. They not only ensure the accurate verification of medication-taking behaviors but also provide practical decision-making bases for patients and doctors through personalized health reports. The intelligent analysis and real-time feedback mechanism greatly improve medication compliance, reduce medical risks caused by non-standard medication, and at the same time, the technical applications in data storage and synchronization also provide efficient and secure data guarantees for long-term health management. The implementation of this module not only promotes the progress of intelligent medication compliance management but also brings innovative solutions to the field of medical and health management.

[0102] Furthermore, the data analysis module includes the following components:

[0103] The data cleaning unit performs data cleaning operations, including noise elimination, missing value filling, and outlier detection and handling, to ensure data quality and analysis reliability;

[0104] The data integration unit realizes the integration and unification of multi-source heterogeneous data, extracts key indicators through feature engineering, and constructs a standardized analysis data set;

[0105] The pattern mining unit conducts pattern recognition and clustering analysis on users' medication-taking behaviors to discover potential behavior patterns;

[0106] The user profile modeling unit constructs multi-dimensional user feature vectors based on users' historical medication records and lifestyle data to form a personalized medication-taking behavior model;

[0107] The trend prediction unit uses time series analysis and prediction algorithms to quantitatively predict users' future medication needs, providing support for intelligent early warning and decision-making.

[0108] In summary, through a series of technical means, the data analysis module deeply mines users' medication-taking behavior data, providing strong support for personalized health management. The data cleaning unit ensures data quality, the data integration unit effectively integrates multi-source heterogeneous data, the pattern mining unit discovers potential patterns in medication-taking behaviors, the user profile modeling unit provides comprehensive user feature data for personalized health management, and the trend prediction unit provides accurate predictions for future medication needs. Each component cooperates with each other, enabling the entire data analysis module not only to provide accurate analysis results but also to dynamically adapt to changes in users' behaviors and optimize health management plans. The implementation of this module makes the monitoring, intervention, and management of medication-taking behaviors more intelligent and personalized, providing powerful technical support for improving patients' medication compliance and medical effects.

[0109] Furthermore, pattern recognition and clustering analysis are performed on users' medication behaviors to discover potential behavioral patterns, including the following steps:

[0110] Classify users' medication behaviors using the K-means algorithm by minimizing the within-cluster error: Identify potential behavioral patterns of users, where C represents the set of all identified clusters of medication behavior patterns; K represents the number of clusters of medication behavior patterns; C k represents the k-th cluster of medication behavior patterns, which contains all data points that conform to the characteristics of this cluster and reflects the similarity of medication behaviors within this cluster; x i represents the i-th medication behavior feature vector, which contains the medication behavior data of the user within a specific time period; μ k is the center point or mean vector of the k-th cluster, representing the "typical" medication behavior of this cluster;

[0111] By analyzing the relevance between the usage of different drugs and specific medication behaviors, discover rules with high support and confidence: P(A→B) = count(A∩B) / count(A), to reveal potential patterns in medication behaviors, where P(A→B) represents the conditional probability in the association rule between medication behavior patterns, that is, the probability of another behavior B occurring after the user has made behavior A, reflecting the strength of the relationship between the two; count(A∩B) represents the number of times that event A and event B occur simultaneously in the medication behavior pattern; count(A) represents the total number of times that event A occurs.

[0112] In summary, through the combination of clustering analysis and association rule mining, the medication behaviors of users can be deeply analyzed to reveal potential patterns and behavioral models. The K-means algorithm can divide users' medication behaviors into multiple pattern clusters with similar characteristics through clustering analysis of medication behaviors, thus providing a clear data basis for further analysis. And association rule mining can discover the relationships between medication behavior patterns, providing support for personalized health management and intelligent intervention. The combination of the two can not only help users optimize their medication regimens and improve medication compliance, but also provide scientific data support for doctors and improve treatment effects. In addition, through predictive analysis combined with historical data, the changing needs of users can be responded to in a timely manner, and the intelligent level of health management can be improved. This multi-dimensional data analysis method provides strong technical support and theoretical basis for modern medical and health management.

[0113] Furthermore, based on users' historical medication records and lifestyle data, construct multi-dimensional user feature vectors to form a personalized medication behavior model, including the following steps:

[0114] Collect the user's historical medication records and lifestyle data and convert them into numerical or categorical encodings;

[0115] Divide the collected data into multiple feature dimensions and construct a feature vector X for each dimension r ;

[0116] Assign weights w to each dimension r , and fuse the feature dimensions into a comprehensive user feature vector V through weighted summation: where n represents the number of feature dimensions;

[0117] Optimize the weights w r so that the predicted medication behavior model can minimize the error between the predicted and actual medication behaviors, that is, ensure that the feature vector V can accurately reflect the user's personalized medication behavior, which is expressed by the formula: where m represents the number of users; is the predicted medication behavior of the jth user according to the personalized medication behavior model; Y j is the actual medication behavior of the jth user;

[0118] Verify the accuracy and generalization ability of the model through the test set to ensure that it can predict future medication behaviors based on the user's historical medication records and lifestyle habits.

[0119] In summary, the process of constructing a personalized medication behavior model includes several key steps such as data collection, feature vector construction, weight optimization, and model verification. In this process, historical medication records and lifestyle data serve as the basis, are transformed into machine-processable data through numericalization and encoding, and the influence of different dimensions is reflected in the feature vector. By fusing each dimension into a comprehensive feature vector through weighted summation, the user's personalized medication behavior can be accurately reflected. The process of weight optimization and error minimization enables the model to continuously adjust parameters to improve prediction accuracy. Through verification and optimization, the model can provide accurate personalized medication behavior predictions for users in real applications, ensuring the safety and effectiveness of users during the medication process. Ultimately, this model can not only improve the quality of personalized medical services but also provide a scientific decision-making basis for health management.

[0120] Furthermore, the abnormal monitoring and warning module includes the following components:

[0121] An abnormal recognition unit that analyzes medication behavior data based on machine learning algorithms to achieve automatic detection and recognition of abnormal patterns, including monitoring of abnormal doses, abnormal medication times, and abnormal medication combinations;

[0122] A risk assessment unit that quantitatively analyzes the detected abnormal behaviors, evaluates their potential risks to the user's health, and classifies them into different levels;

[0123] A hierarchical warning mechanism unit that adopts a hierarchical warning mechanism according to the risk quantification results to generate warning messages of different levels;

[0124] A warning message processing unit that generates detailed warning messages based on the evaluation results and the warning mechanism;

[0125] An intervention suggestion generation unit that automatically generates targeted medication behavior correction suggestions and intervention measures based on the user profile and the risk assessment results.

[0126] In summary, the abnormal monitoring and warning module effectively helps users identify and respond to potential medication abnormal risks through five key units: an abnormal identification unit, a risk assessment unit, a hierarchical warning mechanism unit, a warning message processing unit, and an intervention suggestion generation unit. Through machine learning algorithms and multi-dimensional data analysis, it can accurately identify abnormal medication behaviors, provide hierarchical warning messages according to the severity of the abnormalities, and ensure that users can receive health reminders in a timely manner. At the same time, by generating personalized intervention suggestions, it helps users adjust their medication behaviors and improve medication safety.

[0127] Further, the quantitative analysis of the detected abnormal behaviors, the evaluation of their potential risks to the user's health, and the classification into different levels include the following steps:

[0128] For each detected abnormal behavior, evaluate its severity S according to its potential impact on health g and frequency F g ;

[0129] Based on historical data and clinical experience, evaluate the potential health impact probability P of each abnormal behavior g ;

[0130] According to the formula Calculate the comprehensive health risk score of each abnormal behavior, where R represents the overall health risk score, which is the risk result comprehensively calculated based on all detected abnormal behaviors; G represents the total number of types of abnormal behaviors; w g represents the weight of each abnormal behavior;

[0131] According to the calculated health risk score, classify it into different risk levels.

[0132] In summary, through the quantitative analysis of abnormal behaviors, the assessment of the probability of health impacts, the calculation of comprehensive health risk scores, and the classification of risk levels, the abnormal monitoring and early warning system can effectively identify and evaluate potential risks during the medication process. By comprehensively considering the types, severity, frequency, and potential health impacts of abnormal behaviors, the health risks are precisely quantified, providing important decision-making support for doctors and patients. Risk assessment not only helps patients promptly detect and correct inappropriate medication behaviors but also ensures that corresponding intervention measures can be taken for behaviors at different risk levels through hierarchical risk warnings, thereby enhancing patients' medication safety and the overall health management effect.

[0133] Furthermore, the interaction management module includes the following components:

[0134] The user interface adaptation unit dynamically adjusts the interface layout according to different devices and screen sizes to ensure the fluency and consistency of user operations;

[0135] The information display unit is responsible for displaying key information and real-time data to ensure that users can clearly and accurately obtain the required information;

[0136] The medication reminder push unit automatically pushes medication reminders according to the user's medication schedule to ensure that users take their medications on time and improve compliance;

[0137] The data visualization presentation unit converts complex data into charts, images, or other easily understandable forms, enabling users to intuitively understand the trends and patterns behind the data;

[0138] The permission management unit dynamically assigns access permissions according to different user identities and roles to ensure security and the privacy protection of information;

[0139] The interaction feedback unit captures the feedback information after user operations and makes real-time adjustments and optimizations to improve the user experience;

[0140] The multi-user support unit manages and supports the independent operations and data access of different user groups, providing a personalized interaction experience.

[0141] In summary, through the collaborative work of multiple components, the interaction management module provides a comprehensive and intelligent user experience guarantee. The user interface adaptation unit ensures the consistency and fluency of the interface on different devices. The information display unit enables users to obtain the required information clearly and accurately. The medication reminder push unit improves user compliance. The data visualization presentation unit allows users to more intuitively understand health data. The permission management unit ensures the security and privacy protection of information. The interaction feedback unit enhances the system's responsiveness and optimizes the user experience. The multi-user support unit ensures that the system can handle the personalized needs of multiple users. The careful design and optimization of each component greatly improve the practicality of the system and user satisfaction, ultimately ensuring an efficient, secure and convenient health management experience.

[0142] As Figure 2 shown, a method for managing antidepressant medication compliance based on biometric recognition includes the following steps:

[0143] Generate a unique identity feature template through multimodal biometric collection and perform real-time comparison with the pre-stored user profile to ensure the authenticity of the medication user's identity and prevent drug abuse or misuse;

[0144] Identify and classify drugs through computer vision technology to ensure the accuracy of drug types and dosages; at the same time, use motion capture technology to analyze the user's medication-taking actions and verify the standardization of medication-taking behaviors;

[0145] Clean, denoise and process outliers for the collected raw data, extract key features through the data fusion unit, and establish a personalized medication-taking behavior model to provide data support for anomaly monitoring and trend prediction;

[0146] Based on the medication-taking behavior model, monitor the user's medication-taking behavior in real time, and start a risk assessment mechanism when a deviation from the normal pattern is detected, trigger an alarm according to the risk level and generate intervention suggestions;

[0147] Dynamically adjust the interface layout and interaction method according to the user's habits and device characteristics, and provide personalized function access permissions for different user groups through permission management to ensure data security;

[0148] Automatically push personalized medication reminders according to the medication plan and real-time monitoring data, and at the same time collect user feedback to continuously improve the effect of medication compliance management.

[0149] In summary, the anti-depressant drug compliance management method based on biometric recognition combines a number of cutting-edge technologies, such as biometric recognition, computer vision, motion capture, data cleaning and fusion, risk assessment, permission management, etc., providing a brand-new solution for drug compliance management. Through multiple means such as identity authentication, drug recognition and classification, medication behavior monitoring, personalized reminders, etc., the system can achieve precise personalized medication management, which can not only avoid drug abuse and misuse, but also monitor the user's medication behavior in real time, issue early warnings in a timely manner and provide intervention suggestions. In addition, by dynamically adjusting the interface layout and optimizing the interaction design, the system can provide the most suitable functions and services for different user groups, thereby improving the overall user experience and compliance. The implementation of this method not only improves drug compliance, promotes the health management of patients, but also provides strong support for innovation in the field of intelligent health.

[0150] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An antidepressant drug compliance management system based on biometric recognition, characterized in that: Includes the following modules: The biometrics collection module collects multimodal biometric data of drug users in real time through high-definition cameras, voice sensors and fingerprint readers; The identity authentication module reliably verifies the identity of the drug user based on the collected multimodal biometric data to prevent impersonation or fraud; The medication behavior management module is responsible for real-time analysis of the user's behavior sequence, identifying the appearance characteristics of the drug and verifying the accuracy of the medication action; The data analysis module cleans, extracts features and analyzes patterns of collected data, builds a personalized medication behavior model and predicts medication trends; The abnormal monitoring and early warning module, based on the data analysis results, conducts real-time monitoring and risk assessment of medication behavior through a comprehensive risk assessment mechanism, and generates early warning information and intervention suggestions of corresponding levels; The interactive management module provides an adaptive user interface and is responsible for information display, medication reminder push, data visualization, and permission management for different user groups.

2. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The biometric acquisition module includes the following components: High-definition camera unit, which uses a high-resolution camera to collect the facial features of drug users in real time; The voice collection unit acquires the voice data of the medication user through a high-sensitivity voice sensor; Fingerprint recognition unit, using a high-precision fingerprint reader to collect fingerprint information of drug users; Infrared imaging unit, which uses infrared camera technology to monitor facial temperature distribution; A heart rate detection unit collects heart rate data through a photoplethysmography sensor; The posture sensing unit uses inertial sensors to detect the head and body posture of the user.

3. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The identity authentication module includes the following components: Identity feature integration unit, responsible for generating unique and highly reliable identity feature templates for identity authentication and prevention of fraud; The feature matching unit compares the identity feature template with the registration information in the database to determine the identity consistency; The identity confirmation unit combines the matching results and credibility scores to comprehensively evaluate the identity authentication results and make a final authentication decision; The security authentication optimization unit dynamically adjusts authentication strategies based on historical authentication records, environmental characteristics, and behavior patterns to improve the flexibility and security of identity authentication; The data encryption unit encrypts the stored and transmitted biometric data to prevent data leakage and malicious tampering, and protect user privacy and security.

4. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The medication behavior management module includes the following components: The drug appearance recognition unit uses computer vision algorithms to identify the appearance characteristics of drugs and classify them to verify whether the drugs meet the predetermined standards; The motion capture unit captures the user's hand motion trajectory in real time to determine whether the correct medication action has been completed; The medication action verification unit analyzes the user's medication action and compares it with the standard action template to confirm whether the user has completed the correct medication behavior; The behavior management unit generates personalized health reports based on the results of drug appearance recognition and action verification; The data storage and synchronization unit stores the image and motion data of each medication behavior and synchronizes them to the cloud or local storage for subsequent viewing and analysis.

5. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The data analysis module includes the following components: Data cleaning unit, which performs data cleaning operations, including noise elimination, missing value filling, and outlier detection and processing, to ensure data quality and analysis reliability; Data integration unit, which integrates and unifies multi-source heterogeneous data, extracts key indicators through feature engineering, and builds standardized analytical data sets; Pattern mining unit, which performs pattern recognition and cluster analysis on users’ medication behaviors to discover potential behavioral patterns; The user portrait modeling unit builds a multi-dimensional user feature vector based on the user's historical medication records and lifestyle data to form a personalized medication behavior model; The trend prediction unit uses time series analysis and prediction algorithms to quantitatively predict users' future medication needs, providing support for intelligent early warning and decision-making.

6. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The abnormal monitoring and early warning module includes the following components: The abnormality recognition unit analyzes medication behavior data based on machine learning algorithms to automatically detect and identify abnormal patterns, including monitoring of abnormal dosage, abnormal medication time, and abnormal medication combination; The risk assessment unit quantitatively analyzes the detected abnormal behaviors, assesses their potential risks to the user's health and classifies them into different levels; The hierarchical early warning mechanism unit adopts a hierarchical early warning mechanism to generate early warning information of different levels according to the risk quantification results; The warning message processing unit generates detailed warning information based on the evaluation results and warning mechanism; The intervention suggestion generation unit automatically generates targeted medication behavior correction suggestions and intervention measures based on user profiles and risk assessment results.

7. The antidepressant drug compliance management system based on biometric recognition according to claim 1, characterized in that: The interaction management module includes the following components: The user interface adaptation unit dynamically adjusts the interface layout according to different devices and screen sizes to ensure the smoothness and consistency of user operations; Information display unit, responsible for displaying key information and real-time data to ensure that users can obtain the required information clearly and accurately; Medication reminder push unit automatically pushes medication reminders according to the user's medication use plan to ensure that the user takes medication on time and improves compliance; Data visualization unit converts complex data into charts, images or other easy-to-understand forms, allowing users to intuitively understand the trends and patterns behind the data; The permission management unit dynamically allocates access rights according to different user identities and roles to ensure security and information privacy protection; Interactive feedback unit, which captures feedback information after user operation and makes real-time adjustments and optimizations to improve user experience; Multi-user support unit manages and supports independent operations and data access for different user groups, providing personalized interactive experience.

8. A method for managing antidepressant drug compliance based on biometric identification, implemented based on an antidepressant drug compliance management system based on biometric identification according to any one of claims 1 to 7, characterized in that: The following steps are involved: Generate a unique identity template through multi-modal biometric collection and compare it with the pre-stored user profile in real time to ensure the authenticity of the user's identity and prevent drug abuse or misuse; Computer vision technology is used to identify and classify drugs to ensure the accuracy of drug types and dosages. At the same time, motion capture technology is used to analyze the user's medication actions to verify the standardization of medication behavior. Clean, reduce noise and process outliers on the collected raw data, extract key features through the data fusion unit, and establish a personalized medication behavior model to provide data support for abnormal monitoring and trend prediction; Based on the medication behavior model, the user's medication behavior is monitored in real time, and the risk assessment mechanism is activated when deviations from the normal pattern are detected, triggering warnings and generating intervention suggestions based on the risk level; Dynamically adjust the interface layout and interaction mode according to user habits and device characteristics, and provide personalized function access rights for different user groups through permission management to ensure data security; Automatically push personalized medication reminders based on medication plans and real-time monitoring data, and collect user feedback to continuously improve medication compliance management results.

Images