Medication management method and system of intelligent medicine box

Abstract

The invention discloses a medication management method and system for an intelligent medicine chest, and the method comprises the following steps: S1, carrying out the cooperative starting of hardware; s2, data acquisition and primary processing; s3, carrying out data cross validation and identification confirmation; s4, intelligently dividing a storage area; s5, performing real-time monitoring and early warning on the period of validity; s6, generating and executing a personalized medicine dispensing scheme; s7, medicine taking process supervision and data recording. The medicine taking management system of the intelligent medicine box comprises a medicine box hardware control module; a data acquisition and identification module; a data storage module; a medicine box matching applet function module; an intelligent medical care platform web end function module; an intelligent analysis module; and a safety and maintenance module. According to the invention, through cooperative work of the image acquisition device, the spectrum analyzer and the RFID reader, drug information is acquired from multiple dimensions, and the accuracy of drug identification is greatly improved by using a deep learning algorithm and a professional spectrum comparison technology.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medicine cabinets, and particularly to a medication management method and system for an intelligent medicine cabinet. Background Art

[0002] The incidence of chronic diseases has been increasing year by year. The elderly population often suffers from multiple chronic diseases and needs to take multiple medications simultaneously, resulting in a significant increase in the demand for medication management.

[0003] In traditional intelligent medicine cabinets, their functions exhibit relatively obvious single - characteristic features. In daily use scenarios, they mainly focus on the key link of taking medicine, and the services they can provide are only limited to the measurement of medicines and the reminder of taking medicine time. However, once out of the taking - medicine link, in many important aspects such as the accurate identification of medicines, the tracking of drug efficacy, and the in - depth association with users' health data, traditional intelligent medicine cabinets are unable to cope and are difficult to meet the growing diversified and refined health management needs of users.

[0004] Accordingly, this application proposes a medication management method and system for an intelligent medicine cabinet. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a medication management method and system for an intelligent medicine cabinet.

[0006] To achieve the above - mentioned purpose, the present invention adopts the following technical solutions:

[0007] A medication management method for an intelligent medicine cabinet includes the following steps:

[0008] S1. Hardware collaborative startup: Put the medicine into the intelligent medicine cabinet. The sensors in the medicine cabinet immediately sense the entry of the item, and the medicine cabinet quickly starts the image acquisition device, the spectral analyzer, and the RFID reader. The image acquisition device takes pictures of the medicine package from multiple angles, the spectral analyzer performs spectral scanning on the medicine, and at the same time, the RFID reader reads the RFID tag information on the medicine package;

[0009] S2. Data acquisition and preliminary processing: Transmit the acquired image data to the image recognition module, and transmit the spectral data collected by the spectral analyzer to the spectral analysis module. By comparing with the standard spectral database, obtain the ingredient information of the medicine to assist in confirming the type of medicine, and directly transmit the tag data read by the RFID reader to the data integration module;

[0010] S3. Data cross - verification and identification confirmation: The data integration module summarizes and cross - verifies the information obtained from image recognition, spectral analysis, and RFID tag reading;

[0011] S4. Intelligent division of storage areas, which are divided into normal temperature area, cool area, and refrigerated area. When used for the first time or when new drugs are put in each time, the system automatically distributes the drugs to the corresponding storage areas according to the recognized drug storage requirements.

[0012] S5. Real-time monitoring and warning of expiration date. The medicine box system monitors the expiration date of drugs in real time and prominently displays it on the medicine box display screen and mobile application in a countdown manner.

[0013] S6. Generation and execution of personalized medicine distribution plan. Input personal health information, medical history, allergy history, and the prescribed medication plan by the doctor through the mobile application or the interaction interface of the medicine box to generate a personalized medicine distribution plan.

[0014] S7. Supervision of the medication process and data recording. When taking medicine, the built-in camera of the medicine box is turned on, and with the help of image recognition technology, it monitors the user's medication actions and transmits the medication situation to the mobile application in real time for easy viewing by the user and family members.

[0015] Preferably, in step S1, the spectral analyzer emits light of a specific wavelength to perform spectral scanning on the drug to obtain spectral characteristic information at the molecular level of the drug.

[0016] The RFID reader quickly emits radio frequency signals to read the RFID tag information on the drug package. Key data such as the name, specification, production date, expiration date, and production batch of the drug are pre-stored in the tag.

[0017] Preferably, in step S3, after the data integration module receives the image recognition, spectral analysis, and RFID tag information, when the key information such as the drug name and ingredients are consistent among the three, the medicine box system quickly confirms the recognition result, stores the detailed drug information in the local database, and synchronizes it to the cloud for backup.

[0018] Preferably, in step S3, after the data integration module receives the image recognition, spectral analysis, and RFID tag information, when the key information such as the drug name and ingredients are consistent among the three, the medicine box system quickly confirms the recognition result, stores the detailed drug information in the local database, and synchronizes it to the cloud for backup.

[0019] Preferably, in step S5, the medicine box system is built with a high-precision time synchronization module to monitor the expiration date of drugs in real time and prominently display the remaining expiration date of the drugs on the medicine box display screen and the associated mobile application in a countdown manner.

[0020] Preferably, in step S6, personal health information, medical history, allergy history, and the doctor's medication plan can be entered on the mobile application or the interaction interface of the medicine box. After the medicine box system receives it, it uses intelligent algorithms, combines the drug characteristics and individual conditions, and generates a personalized medicine distribution plan to clarify the taking time, dosage, and frequency of each drug.

[0021] A medication management system for a smart medicine cabinet, comprising:

[0022] Medicine cabinet hardware control module: Responsible for driving the screen display of the medicine cabinet to ensure that various types of information are clearly presented to the user, and managing and controlling sensor data and operations;

[0023] Data acquisition and identification module: Using image acquisition devices, spectrometers, and RFID readers, collect drug information from multiple angles, transmit the collected image data to the image recognition module, spectral data to the spectral analysis module, and RFID tag data to the data integration module. Through operations such as comparing with the standard database, achieve accurate identification and information confirmation of drugs;

[0024] Data storage module: Based on the MySQL database, store various types of data such as user information, prescription information, medication records, vital sign monitoring values, and device operation parameters to ensure the integrity and security of the data;

[0025] Function module of the small program supporting the medicine cabinet and function module of the web end of the intelligent medical care platform: Provide functions such as login, password modification, and system exit for community administrators, realize the function of monitoring the vital sign information and medication records of community elderly people, so that community administrators can timely understand the health status of the elderly, support community administrators to manage the information of community elderly people and their family members, communicate with the medicine cabinet, and obtain and manage relevant data;

[0026] Intelligent analysis module: Analyze the user's medication pattern, understand the user's medication habits through mining data such as medication records, evaluate the medication effect, combine the vital sign monitoring values and medication records, judge the impact of drugs on the user's health status, and provide personalized medication suggestions for the user according to the analysis results, or assist community administrators and medical staff in adjusting the medication plan;

[0027] Security and maintenance module: Implement password protection for database access, and managers can change passwords to prevent data leakage; Realize regular automatic backup of the database to ensure data security and recoverability; Provide a system operation status monitoring function to promptly detect and solve system failures.

[0028] Preferably, the medicine cabinet hardware control module is used to control sensors for detecting the entry of items and vital sign monitoring sensors to process and analyze the collected data, and control the operation of the automatic medicine dispensing device to achieve accurate drug distribution.

[0029] Preferably, the data collection and recognition module is used to store user information and prescription information. The stored user information includes relevant data of community elders, their family members, and community administrators. The recorded drug information covers detailed information such as the name, ID, and ISBN of the drugs. The stored prescription information includes the drug ID, dosage, and start and end times of taking the medicine in the prescription.

[0030] The present invention has the following beneficial effects:

[0031] 1. Through the cooperation of the image acquisition device, the spectral analyzer, and the RFID reader, drug information is collected in multiple dimensions. Combining deep learning algorithms with professional spectral comparison techniques, the drug recognition accuracy is greatly improved, avoiding medication errors caused by incorrect recognition and ensuring the medication safety of users. In addition, the storage area is intelligently divided and the expiration date is monitored in real time to warn of drugs approaching expiration, ensuring a suitable drug storage environment, reducing waste, and guaranteeing quality.

[0032] 2. After the user inputs personal health information, medical history, and other data, the medicine cabinet system uses intelligent algorithms to generate a personalized medicine dispensing plan, accurately determining the time, dosage, and frequency of taking the medicine to meet the individual needs of different users. This personalized service helps improve the therapeutic effect of drugs and enhance the pertinence and effectiveness of user health management.

[0033] 3. Image recognition technology is used to monitor the user's medication-taking actions. Once there is a situation of not taking the medicine on time or operating errors, an immediate reminder is given, effectively improving the user's medication compliance. At the same time, the system transmits the medication-taking situation and vital sign monitoring data in real time, facilitating users, their family members, and community administrators to promptly know the health information and providing data support for medical decision-making to achieve dynamic tracking and management of the user's health status.

[0034] 4. The medicine cabinet is equipped with a mini-program and a web version of the intelligent healthcare platform, creating practical functions for different users. Community elders and their family members can easily manage medication-taking and query information using the mini-program, and remote medical consultation is more convenient for the elders to seek medical treatment. Community administrators can efficiently manage user information and monitor the health data of the elders through the web version, realizing remote care and comprehensively improving service efficiency and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the overall process of a medication management method for a smart medicine cabinet proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0037] A method for managing medication in a smart medicine box comprises the following steps:

[0038] S1. Hardware collaborative startup, put the medicine into the smart medicine box, use a 10-inch TFT-LCD screen with a resolution of 1280×800 pixels, and display large font icons for the elderly to view. It is also equipped with a Qualcomm Snapdragon 665 processor and supports the Android 10 system to ensure stable operation of the program. The medicine box has a built-in 5000mAh rechargeable lithium battery with a full battery life of more than 7 days. The outer shell is made of ABS plastic, which is durable and hygienic, meets medical standards, and is integrated injection molding and easy to clean.

[0039] The sensor in the medicine box immediately senses the entry of the object, and the medicine box quickly activates the image acquisition device, spectrum analyzer and RFID reader. The image acquisition device is equipped with multiple high-definition cameras to quickly and accurately shoot the drug packaging from multiple angles such as horizontal, vertical, and tilted to ensure that clear images of all sides of the drug packaging are obtained, and key information such as text, patterns, and logos on the packaging are fully captured;

[0040] The spectrometer performs spectral scanning on the drug. The spectrometer emits light of a specific wavelength to perform spectral scanning on the drug to obtain spectral characteristic information at the drug molecular level;

[0041] The RFID reader reads the RFID tag information on the drug packaging. The RFID reader quickly transmits radio frequency signals to read the RFID tag information on the drug packaging. The tag pre-stores key data such as the drug name, specification, production date, expiration date, production batch, etc.

[0042] S2. Data collection and preliminary processing: The acquired image data is transmitted to the image recognition module, where the image recognition module uses a convolutional neural network learning model designed for grid structure data to intelligently recognize and analyze the text, patterns and other information in the image, and carefully compares the acquired spectral data with the precise and large standard spectral database, and deeply analyzes the spectral characteristics with the help of professional algorithms and models to obtain the composition information of the drug;

[0043] The spectral data collected by the spectrum analyzer is transmitted to the spectrum analysis module, which compares the acquired spectral data with the large and accurate standard spectrum database, and uses professional algorithms and models to deeply analyze the spectral characteristics and obtain the composition information of the drug;

[0044] Assist in confirming the type of medicine, the tag data read by the RFID reader is directly transmitted to the data integration module;

[0045] S3. Data cross-verification and identification confirmation. The data integration module summarizes and cross-verifies the information obtained from image recognition, spectral analysis, and RFID tag reading.

[0046] After receiving the information from image recognition, spectral analysis, and RFID tags, if the key information such as drug name and ingredients is consistent among the three, the medicine cabinet system quickly confirms the recognition result, stores the detailed drug information in the local database, and synchronizes it to the cloud for backup.

[0047] If there are differences in the information, the system automatically marks the suspicious points and initiates secondary identification. During this process, the image acquisition device and the spectral analyzer are re-called to collect data finely from multiple angles, and more information is obtained from the cloud extended database for in-depth comparison.

[0048] S4. Intelligent division of storage areas, which are divided into normal temperature area, cool area, and refrigerated area. When used for the first time or when new drugs are placed each time, the system automatically allocates the drugs to the corresponding storage areas according to the identified drug storage requirements. Each area is equipped with independent temperature and humidity adjustment devices and high-precision sensors.

[0049] The temperature in the normal temperature area is controlled at 10°C to 30°C, and the humidity is controlled at 35% to 75%. The temperature in the cool area does not exceed 20°C, and the humidity is controlled at 45% to 75%. The temperature in the refrigerated area is controlled at 2°C to 8°C, and the humidity is controlled at 35% to 75%.

[0050] S5. Real-time monitoring and warning of expiration date. The medicine cabinet system monitors the expiration date of drugs in real-time and prominently displays it on the medicine cabinet display screen and the mobile application in a countdown manner.

[0051] The medicine cabinet system is built-in with a high-precision time synchronization module to monitor the expiration date of drugs in real-time. In a countdown manner, it prominently displays the remaining expiration date of the drugs on the medicine cabinet display screen and the associated mobile application.

[0052] When the drug is 3 months away from the expiration date, the mobile application pushes a yellow warning message to remind the user to pay attention to the expiration date of the drug. When it is 1 month away from the expiration date, the medicine cabinet emits a sharp sound alarm, and at the same time, the display screen flashes a red warning message. For drugs approaching expiration, the medicine cabinet system provides handling suggestions on the mobile application.

[0053] S6. Generation and execution of personalized drug dispensing plan. Enter personal health information, medical history, allergy history, and the prescribed medication plan by the doctor through the mobile application or the interactive interface of the medicine cabinet to generate a personalized drug dispensing plan.

[0054] You can input your personal health information, medical history, allergy history and doctor's medication plan on the mobile phone application or medicine box interactive interface. After receiving the information, the medicine box system uses intelligent algorithms to combine the characteristics of the medicine and individual conditions to generate a personalized medication plan, clarifying the time, dosage and frequency of taking each medicine.

[0055] S7. Medication process supervision and data recording. When taking medicine, the built-in camera in the medicine box is turned on, and the image recognition technology is used to monitor the user's medication actions, and the medication status is transmitted to the mobile phone application in real time for the user and family members to view.

[0056] A medicine management system for a smart medicine box, comprising:

[0057] Medicine box hardware control module: responsible for driving the screen display of the medicine box, ensuring that all kinds of information are clearly presented to the user, and managing and controlling sensor data and work;

[0058] The medicine box hardware control module is used to control the sensors that detect the entry of items and the vital signs monitoring sensors to process and analyze the collected data, control the operation of the automatic medicine dispensing device, and achieve accurate distribution of medicines;

[0059] Data collection and identification module: Use image acquisition equipment, spectrum analyzer, and RFID reader to collect drug information from multiple angles, transmit the collected image data to the image recognition module, transmit the spectrum data to the spectrum analysis module, and transmit the RFID tag data to the data integration module. Through operations such as comparison with the standard database, accurate drug identification and information confirmation can be achieved;

[0060] The data collection and identification module is used to store user information and prescription information. The stored user information includes relevant data of community elders, their family members, and community administrators. The recorded drug information includes detailed information such as the drug name, ID, ISBN, etc. The stored prescription information includes the drug ID, dosage, and start and end time of the drug included in the prescription.

[0061] Data storage module: Based on MySQL database, it stores user information, prescription information, medication records, vital signs monitoring values, equipment operating parameters and other data to ensure data integrity and security;

[0062] Functional module of the medicine box supporting applet Functional module of the smart medical care platform web terminal: Provide community administrators with login, password modification, and system exit functions, and realize the function of monitoring the vital signs and medication records of the elderly in the community, so that community administrators can understand the health status of the elderly in a timely manner, support community administrators to manage the information of the elderly in the community and their families, communicate with the medicine box, and obtain and manage relevant data;

[0063] Intelligent Analysis Module: Analyze the user's medication pattern, understand the user's medication habits through mining data such as medication records, evaluate the medication effect, combine the vital sign monitoring values and medication records to judge the impact of the medication on the user's health condition, and provide personalized medication advice for the user or assist community administrators and medical staff in adjusting the medication plan according to the analysis results;

[0064] Security and Maintenance Module: Implement password protection for database access. Managers can change the password to prevent data leakage; realize automatic regular backup of the database to ensure data security and recoverability; provide system operation status monitoring function to detect and solve system failures in a timely manner.

[0065] Among them, the intelligent medicine box is placed in the community elderly care scenario. For the medication management method, the above methods and systems are used, and compared with the implementation cases using other different methods and systems;

[0066] Example 1:

[0067] Step 1: The community administrator puts the medications that the elderly usually take into the intelligent medicine box in turn. When the sensors in the medicine box sense the entry of an item, they immediately prompt the image acquisition device, spectral analyzer, and RFID reader to start synchronously. The image acquisition device takes pictures of the medicine packaging from different angles through multiple cameras to comprehensively capture important information such as text, patterns, and logos on the packaging. The spectral analyzer emits light of a specific wavelength to perform a spectral scan of the medicine to obtain the spectral characteristics at the molecular level of the medicine.

[0068] Meanwhile, the RFID reader quickly emits radio frequency signals to read the key data such as the medicine name, specification, production date, expiration date, production batch, etc. pre-stored in the RFID tag of the medicine packaging.

[0069] Step 2: The image data obtained by the image acquisition device is transmitted to the image recognition module. According to the convolutional neural network (CNN) feature extraction formula, the input image I is an m×n matrix, the convolutional kernel K is a p×q matrix, the stride is s, and the padding is f. The calculation formula for the output feature map O after convolutional operation is:

[0070]

[0071] This module automatically extracts the features in the image through multiple convolutional layers, pooling layers, and fully connected layers, intelligently identifies and analyzes information such as text and patterns in the image, and extracts preliminary information such as the medicine name, dosage form, and manufacturer;

[0072] The spectral data collected by the spectral analyzer is transmitted to the spectral analysis module. Using the similarity matching algorithm, the obtained spectral data S is compared with the spectrum S in the standard spectral database stdPerform a comparison. The similarity is calculated by cosine similarity, and the formula is:

[0073]

[0074] By deeply analyzing the spectral characteristics, obtain the component information of the drug to assist in confirming the drug type. The tag data read by the RFID reader is directly transmitted to the data integration module to provide basic data for subsequent comprehensive analysis.

[0075] Step 3: The data integration module summarizes the information obtained from image recognition, spectral analysis, and RFID tag reading. According to the consistency judgment formula of data cross-validation, let the drug name obtained from image recognition be N rfid , and the component information be C img , C spec , C rfid . The consistency judgment formula is:

[0076]

[0077] If the key drug information obtained by the three technologies is consistent, the medicine cabinet system confirms the drug recognition result and stores the detailed drug information in the local database, and at the same time synchronizes it to the cloud database for backup. If there are differences in the information, the system starts the secondary recognition process.

[0078] Step 4: The system allocates the drug to the corresponding storage area according to the storage requirements of the recognized drug; taking the temperature control of the refrigerated area as an example, the current temperature of the refrigerated area is T cur , the set temperature range is [T min , T max , the refrigeration power of the temperature adjustment device is, the refrigeration time is t, and the heat capacity of the refrigerated area is C. The temperature adjustment formula is:

[0079]

[0080] When T new < T min , stop refrigerating; when T new > T max , start refrigerating.

[0081] Step 5: When the medicine cabinet system starts, the built-in high-precision time synchronization module automatically connects to the network time server (NTP), obtains the accurate standard time through the network time server, and calibrates and synchronizes the medicine cabinet system time with it.

[0082] After that, at regular time intervals (such as every 15 minutes), the time synchronization module requests time updates from the NTP server again to ensure that the time of the medicine cabinet system always maintains high-precision accuracy. When the calculated remaining validity period of the medicine is in the form of a countdown, it is simultaneously prompted on the display screen of the medicine cabinet and the associated mobile application, and will be displayed in different colors according to the length of the remaining validity period. For example, it is displayed in green when the remaining validity period is more than 3 months, yellow when it is 3 - 1 month, and red when it is less than 1 month.

[0083] Step 6: Log in to the system through the mobile application and enter in detail the personal health information, past medical history, drug allergy history of the elderly, and the medication plan prescribed by the doctor. After the medicine cabinet system receives the information entered by the family member, combined with the user health information vector H, drug characteristic vector D, and historical medication data vector P, a basic medicine distribution plan B is obtained through the machine learning model M, and then combined with the rule-based adjustment function R. The final personalized medicine distribution plan S is:

[0084] B = M(H, D, P), S = R(B, H, D)

[0085] After the medicine distribution plan is generated, it is synchronously displayed on the display screen of the medicine cabinet and the mobile application of the family member.

[0086] Step 7: According to the algorithm for setting the medication reminder time, let the medication time interval be:

[0087] Tinterual, and the last medication time be Tlast, then the next medication reminder time is:

[0088] Tnext = Tlast + Tinterual

[0089] When the medication time arrives, the built-in speaker of the medicine cabinet emits a voice reminder, and the screen synchronously displays a prominent medication reminder interface. The mobile application of the elderly's family member will also receive a push notification. At the same time, the built-in camera of the medicine cabinet is activated, and image recognition technology is used to monitor the elderly's medication actions, continuously shooting the pictures of the medication process and transmitting the image data to the medicine cabinet system for analysis.

[0090] If the elderly do not take the medicine on time, the medicine cabinet will remind again after a period of time; if the medication operation is incorrect, the medicine cabinet will promptly issue a correction reminder and record the relevant information in the local database. After the elderly complete taking the medicine, the medicine cabinet system will transmit the medication situation to the mobile application and the web end of the intelligent medical care platform of the community administrator in real time. The administrator can log in to the web end to view the elderly's medication records in real time.

[0091] Community doctors regularly log in to the platform, combine the vital sign monitoring values of the elderly and the medication records, evaluate the medication effect, and adjust the medication plan if necessary. The medicine cabinet system synchronously updates the medicine distribution plan.

[0092] After the implementation of a series of operation processes, the intelligent medicine box works in coordination with an image acquisition device, a spectral analyzer, and an RFID reader, and uses algorithms such as convolutional neural networks and spectral comparison to achieve precise identification and management of drug information. At the same time, it reasonably partitions according to storage requirements and monitors the expiration date in real time. The system can generate a personalized medicine dispensing plan based on user input, fully considering drug characteristics and individual differences. When taking medicine simultaneously, the built-in camera monitors the user's actions to improve medication compliance, and the recorded data also provides a basis for doctors to adjust the treatment plan.

[0093] Example Two:

[0094] Community workers put the elderly's medicines into the intelligent medicine box. The medicine box uses a high-definition camera combined with OCR technology to identify the text on the medicine package to obtain basic information, and uses the NFC function to read detailed drug information. Then, the data is transmitted to the processing module for integrated analysis and stored in local and cloud databases. Next, according to the elderly's medication plan and drug characteristics, the medicine box automatically partitions. The built-in voice assistant can answer the elderly's questions about medication time, remaining drug quantity, etc. through voice interaction. When giving a medication reminder, the LED indicator light flashes, and the voice assistant makes a sound. The pressure sensor monitors the opening and closing of the medicine box. If the medicine is not taken, it will remind again and notify the family members.

[0095] At the same time, the medicine box is connected to intelligent health monitoring devices to obtain the elderly's vital sign data, providing a medication effect evaluation report for community doctors to adjust the medication plan.

[0096] Example Three:

[0097] Community workers put the elderly's medicines into the medicine box. The high-definition image acquisition device on the top of the medicine box takes a full-view photo of the package, and with the help of advanced artificial intelligence image recognition algorithms, quickly identifies information such as the drug name, specifications, and manufacturer. The identified data is immediately uploaded to the intelligent cloud platform for comparison and verification with the drug database on the platform to ensure accuracy, and at the same time, a local database backup is made.

[0098] The intelligent cloud platform formulates a medication plan according to the doctor's prescription and the elderly's needs. When the medication time arrives, the medicine box lights up and makes a sound to remind, and also pushes a notification to the mobile phone bound to the elderly. The medicine box can be connected to health monitoring devices such as intelligent bracelets and sphygmomanometers to collect the elderly's health data in real time and upload it to the cloud platform.

[0099] It should be noted that in the comparative example implementation, in the use of the community elderly care scenario, in Comparative Example One, traditional QR code scanning and local storage management are adopted. In Comparative Example Two, mainly a high-definition image acquisition device and image recognition algorithms are used to automatically extract key drug data, and a comparison is made with Example One, Example Two, and Example Three, as shown in Table One:

[0100] Table One: Comparison Table for Different Drug Identifications

[0101]

[0102] As compared with other embodiments and comparative examples in Table 1, Example 1 has significant advantages in many aspects: in terms of the number of algorithms, it leads by 4 algorithms. Multiple algorithms such as convolutional neural networks and spectral comparison operate in coordination, providing diverse processing strategies and being able to flexibly handle complex packaging situations. In contrast, Example 2 has only 2 algorithms, with slightly insufficient functional coverage, and Example 3 and the two comparative examples have only 1 algorithm each, with limited means for handling complex tasks. The significant difference in the number of algorithms significantly affects the overall performance and adaptability. In terms of recognition accuracy, the recognition accuracies of Example 1, Example 2, and Example 3 are 98%, 95%, and 96% respectively. Example 1 has the highest accuracy, which is crucial for ensuring the accuracy and safety of drug management. Regarding the diversity of drug packaging, Example 1 and Example 3 perform excellently in the recognition ability of special-shaped packaging, both reaching 98%, being able to keenly capture unique shape features and ensuring that drugs of different shapes are not misjudged or missed. In the recognition of reflective material packaging, Example 3 performs extremely prominently, reaching 98%, and Example 1 also has a certain ability to handle it, which is 80%. For the adaptability to different language packaging, Example 1 stands out with a high adaptability of 93%. Relying on advanced image recognition, translation, and text parsing algorithms, it can effectively recognize label information in multiple languages. In terms of the recognition ability of micro-font labels, Example 1 also performs excellently, leading with 90%. By optimizing the image acquisition resolution and enhancement algorithms, combined with advanced text recognition technology, it ensures that key drug information is not omitted.

[0103] It should be noted that in the comparative examples, in terms of information processing speed, drug ingredient analysis ability, and drug dosage form recognition ability, a comparison is made with Example 1, Example 2, Comparative Example 1, and Comparative Example 2, as shown in Table 2:

[0104] Table 2: Comparison Table of Information Processing Speed, Drug Ingredient Analysis Ability, and Drug Dosage Form Recognition Ability

[0105]

[0106] As shown in Table 2, although Example 1 is relatively slow in information processing speed at 2.5 seconds, not as fast as Example 2 (1.5 seconds, with a simple algorithm and limited information sources resulting in a simple process), Comparative Example 1 (theoretically 0 seconds, with almost real-time completion of simple input but limited functions), Example 3 (2 seconds, with a powerful deep learning algorithm but large computational volume), and Comparative Example 2 (2.25 seconds, with a slightly more complex process due to image recognition combined with mobile application assistance to improve information), its technical logic and design concept endow it with unparalleled advantages in drug management accuracy. It innovatively adopts multi-dimensional data fusion, comprehensively using image acquisition devices, spectral analyzers, and RFID readers to obtain detailed drug information from multiple aspects such as appearance features, molecular structure characteristics, and identity identification.

[0107] In terms of the ability of pharmaceutical ingredient analysis, Example 1 leads by 92%. By virtue of the spectrometer to obtain unique spectral characteristics, relevant algorithms are used to deeply analyze and interpret a large amount of spectral data. Combining with the multi-dimensional data cross-validation mechanism, spectral analysis is combined with image recognition and RFID information, greatly improving the analysis accuracy.

[0108] In terms of the ability of pharmaceutical dosage form recognition, Example 1 leads far ahead with a recognition rate of 96%. Multi-dimensional recognition means are combined with complex algorithms. The image acquisition device comprehensively captures the appearance, the spectral analyzer provides auxiliary information, and the RFID information is mutually verified. The convolutional neural network algorithm is used to deeply extract and analyze features for the image, and the spectral data is finely analyzed. After the data of different dimensions are processed by their respective algorithms, they are integrated through the data fusion algorithm to achieve accurate recognition of various conventional and special dosage forms.

[0109] Generally speaking, although Example 1 does not have an advantage in terms of information processing speed, it has obvious advantages in the accuracy of pharmaceutical ingredient analysis and dosage form recognition, and can provide more comprehensive and reliable support for the intelligent medicine cabinet to manage drugs, and is more suitable for scenarios with higher requirements for the accuracy of drug management.

[0110] The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered within the protection scope of the present invention.

Claims

1. A method for managing medication in a smart medicine box, characterized in that: The following steps are involved: S1. Hardware collaborative startup. Put the medicine into the smart medicine box. The sensor in the medicine box immediately senses the entry of the object. The medicine box quickly starts the image acquisition device, spectrum analyzer and RFID reader. The image acquisition device takes pictures of the medicine packaging from multiple angles. The spectrum analyzer performs spectrum scanning on the medicine. At the same time, the RFID reader reads the RFID tag information on the medicine packaging. S2, data acquisition and preliminary processing, the acquired image data is transmitted to the image recognition module, the spectrum data collected by the spectrum analyzer is transmitted to the spectrum analysis module, and the component information of the drug is obtained by comparing with the standard spectrum database to assist in confirming the type of drug, and the label data read by the RFID reader is directly transmitted to the data integration module; S3, data cross-validation and identification confirmation, the data integration module summarizes and cross-validates the information of image recognition, spectral analysis and RFID tag reading; S4. Intelligent division of storage areas into normal temperature area, cool area, and refrigerated area. When a new drug is put into the system for the first time or each time, the system automatically allocates the drug to the corresponding storage area according to the identified drug storage requirements; S5. Real-time monitoring and early warning of validity period: the medicine box system monitors the validity period of the medicine in real time and displays it prominently on the medicine box display and mobile phone application in a countdown manner; S6. Personalized medication plan generation and execution: input personal health information, medical history, allergy history, and medication plan prescribed by the doctor through the interactive interface of the mobile phone application or medicine box to generate a personalized medication plan; S7. Medication process supervision and data recording. When taking medicine, the built-in camera in the medicine box is turned on, and the image recognition technology is used to monitor the user's medication actions, and the medication status is transmitted to the mobile phone application in real time for the user and family members to view.

2. The medication management method of a smart medicine box according to claim 1 is characterized in that: In step S1, the spectrum analyzer emits light of a specific wavelength to perform spectral scanning on the drug to obtain spectral characteristic information at the drug molecular level; The RFID reader quickly transmits radio frequency signals to read the RFID tag information on the drug packaging. The tag pre-stores key data such as the drug's name, specifications, production date, expiration date, production batch, etc.

3. The medication management method of a smart medicine box according to claim 1 is characterized in that: In step S2, the image recognition module uses a convolutional neural network learning model designed specifically for grid structure data to intelligently recognize and analyze text, patterns and other information in the image, and carefully compares the acquired spectral data with the precise and large standard spectral database, and deeply analyzes the spectral characteristics with the help of professional algorithms and models to obtain the composition information of the drug.

4. The medication management method of a smart medicine box according to claim 1 is characterized in that: In step S3, after the data integration module receives the image recognition, spectral analysis and RFID tag information, the key information such as the drug name and ingredients are consistent. The medicine box system quickly confirms the recognition result, stores the detailed drug information in the local database, and synchronizes it to the cloud for backup.

5. The medication management method of a smart medicine box according to claim 1 is characterized in that: In step S5, the medicine box system has a built-in high-precision time synchronization module to monitor the expiration date of the medicine in real time, and uses a countdown method to prominently display the remaining expiration date of the medicine on the medicine box display screen and the mobile phone application associated therewith.

6. The medication management method of a smart medicine box according to claim 1 is characterized in that: In step S6, personal health information, medical history, allergy history and doctor's medication plan can be entered into the mobile phone application or the medicine box interactive interface. After receiving the information, the medicine box system uses an intelligent algorithm to combine the characteristics of the drug and the individual situation to generate a personalized medication plan to specify the time, dosage and frequency of taking each drug.

7. A smart medicine box medication management system, characterized in that: include: Medicine box hardware control module: responsible for driving the screen display of the medicine box, ensuring that all kinds of information are clearly presented to the user, and managing and controlling sensor data and work; Data collection and identification module: Use image acquisition equipment, spectrum analyzer, and RFID reader to collect drug information from multiple angles, transmit the collected image data to the image recognition module, transmit the spectrum data to the spectrum analysis module, and transmit the RFID tag data to the data integration module. Through operations such as comparison with the standard database, accurate drug identification and information confirmation can be achieved; Data storage module: Based on MySQL database, it stores user information, prescription information, medication records, vital signs monitoring values, equipment operating parameters and other data to ensure data integrity and security; Functional module of the medicine box supporting applet Functional module of the smart medical care platform web terminal: Provide community administrators with login, password modification, and system exit functions, and realize the function of monitoring the vital signs and medication records of the elderly in the community, so that community administrators can understand the health status of the elderly in a timely manner, support community administrators to manage the information of the elderly in the community and their families, communicate with the medicine box, and obtain and manage relevant data; Intelligent analysis module: Analyze the user's medication patterns, understand the user's medication habits through mining medication records and other data, evaluate the medication effect, and determine the impact of the drug on the user's health status by combining vital sign monitoring values ​​and medication records. Based on the analysis results, provide users with personalized medication recommendations, or assist community administrators and medical staff in adjusting medication plans; Security and maintenance module: Implement password protection for database access, and managers can change passwords to prevent data leakage; implement regular automatic database backup to ensure data security and recoverability; provide system operation status monitoring function to detect and resolve system failures in a timely manner.

8. The smart medicine box medication management system according to claim 7, characterized in that: The medicine box hardware control module is used to control the sensors for detecting the entry of items and the vital signs monitoring sensors to process and analyze the collected data, control the operation of the automatic medicine dispensing device, and achieve accurate distribution of medicines.

9. The smart medicine box medication management system according to claim 7, characterized in that: The data collection and identification module is used to store user information and prescription information. The stored user information includes relevant data of community elderly people, their families, and community administrators. The recorded drug information covers detailed information such as the drug's name, ID, ISBN, etc. The stored prescription information includes the drug ID, dosage, and start and end time of the medication contained in the prescription.

Images