Data processing method and device, computer equipment and storage medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to a data processing method, which comprises the following steps: acquiring multi-dimensional physiological data and drug combination data of a target patient; calling a target prediction model issued by the central server; performing prediction processing on the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain corresponding adverse reaction evaluation data; judging whether the adverse reaction evaluation data is smaller than a preset dynamic safety threshold value or not; if yes, target drugs corresponding to the drug combination data are grabbed based on a drug dispensing tool; and generating medicine taking information corresponding to the target medicine, and displaying the medicine taking information. The invention further provides a data processing device, computer equipment and a storage medium. In addition, the medicine taking information can be stored in the block chain. The method can be applied to a medication management scene in the field of medical science and technology, and improves the efficiency and accuracy of medication risk assessment processing and medicine distribution processing.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology and can be applied to the field of medical technology, particularly to data processing methods, devices, computer equipment and storage media. Background Technology

[0002] In the traditional field of medication management, the current mainstream operating model is to conduct medication risk assessments according to pre-set fixed rules after generating patient medication data. This assessment method is like a "one-size-fits-all" approach, failing to fully consider the rich and diverse individual characteristics of patients. However, these key factors all have a significant impact on medication risk, but the traditional model ignores them, resulting in low efficiency and unsatisfactory accuracy in medication risk assessment, which is seriously out of step with the current advocacy for personalized precision medicine.

[0003] Meanwhile, in the medication dispensing process, traditional methods rely excessively on manual operation. The manual dispensing process is not only cumbersome and time-consuming, but also highly susceptible to human factors such as operator fatigue and concentration, leading to dispensing errors. Furthermore, the lack of intelligent auxiliary tools makes it difficult for traditional dispensing methods to achieve efficient and accurate dispensing, undoubtedly posing a risk to patient medication safety and potentially negatively impacting treatment outcomes. Summary of the Invention

[0004] The purpose of this application is to provide a data processing method, apparatus, computer equipment, and storage medium to solve the technical problems of low processing efficiency and accuracy in the existing field of medication management, specifically in medication risk assessment and dispensing.

[0005] Firstly, a data processing method is provided, including:

[0006] Acquire multidimensional physiological data and drug combination data of target patients;

[0007] Invoke the target prediction model issued by the preset central server;

[0008] Based on the target prediction model, the multidimensional physiological data and the drug combination data are predicted and processed to obtain the corresponding adverse reaction assessment data;

[0009] Determine whether the adverse reaction assessment data is less than a preset dynamic safety threshold;

[0010] If so, the target drug corresponding to the drug combination data is captured based on the preset drug distribution tool;

[0011] Generate drug dispensing information corresponding to the target drug, and display and process the drug dispensing information.

[0012] Secondly, a data processing apparatus is provided, comprising:

[0013] The first acquisition module is used to acquire multi-dimensional physiological data and drug combination data of the target patient;

[0014] The calling module is used to call the target prediction model issued by the preset central server;

[0015] The first prediction module is used to predict and process the multidimensional physiological data and the drug combination data based on the target prediction model to obtain the corresponding adverse reaction assessment data.

[0016] The first judgment module is used to determine whether the adverse reaction assessment data is less than a preset dynamic safety threshold.

[0017] The capture module is used to capture the target drug corresponding to the drug combination data based on a preset drug distribution tool if the condition is met.

[0018] The first processing module is used to generate drug dispensing information corresponding to the target drug and to display and process the drug dispensing information.

[0019] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described data processing method.

[0020] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described data processing method.

[0021] In the above-described data processing method, apparatus, computer equipment, and storage medium, the following steps are taken: First, multi-dimensional physiological data and drug combination data of the target patient are acquired. Then, a target prediction model issued by a preset central server is invoked. Next, the multi-dimensional physiological data and drug combination data are predicted based on the target prediction model to obtain corresponding adverse reaction assessment data. Subsequently, it is determined whether the adverse reaction assessment data is less than a preset dynamic safety threshold. If so, a target drug corresponding to the drug combination data is retrieved using a preset drug dispensing tool. Finally, medication dispensing information corresponding to the target drug is generated and displayed. Based on the above automated processing flow, this application acquires multi-dimensional physiological data and drug combination data of the target patient, then uses a target prediction model issued by a central server to predict adverse reaction assessment data based on the multi-dimensional physiological data and drug combination data. Only when the adverse reaction assessment data is detected to be less than the dynamic safety threshold is the target drug retrieved based on the drug dispensing tool, thereby generating medication dispensing information corresponding to the target drug and displaying the medication dispensing information. Thus, this application, by using a target prediction model to predict and process multi-dimensional physiological data and drug combination data, can achieve efficient and accurate medication risk assessment, improving the efficiency and accuracy of medication risk assessment. Furthermore, when adverse reaction assessment data is detected to be below a preset dynamic safety threshold, the use of a drug dispensing tool can quickly identify the target drug corresponding to the drug combination data, generate and display the corresponding drug dispensing information, effectively reducing manual drug dispensing operations and thus improving the efficiency and accuracy of drug dispensing processing. Attached Figure Description

[0022] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;

[0024] Figure 2 This is a flowchart of an embodiment of the data processing method according to this application;

[0025] Figure 3 This is a schematic diagram of the structure of an embodiment of the data processing apparatus according to this application;

[0026] Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0028] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0029] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0030] like Figure 1 As shown, system architecture 100 may include terminal device 101, network 102, and server 103. Terminal device 101 may be a laptop 1011, tablet 1012, or mobile phone 1013. Network 102 is used as a medium to provide a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0031] Users can use terminal device 101 to interact with server 103 via network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0032] Terminal device 101 can be various electronic devices with a display screen and support web browsing. In addition to laptops 1011, tablets 1012, or mobile phones 1013, terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer, and a desktop computer, etc.

[0033] Server 103 can be a server that provides various services, such as a backend server that provides support for the pages displayed on terminal device 101.

[0034] It should be noted that the data processing method provided in the embodiments of this application is generally executed by a server / terminal device, and correspondingly, the data processing device is generally located in the server / terminal device.

[0035] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0036] Continue to refer to Figure 2 A flowchart illustrating an embodiment of the data processing method according to this application is shown. The order of steps in the flowchart can be changed, and some steps can be omitted, depending on different needs. The data processing method provided in this application embodiment can be applied to any scenario requiring data processing, and therefore can be applied to products in these scenarios, such as medication management products in the medical technology field. The data processing method includes the following steps:

[0037] Step S201: Obtain multi-dimensional physiological data and drug combination data of the target patient.

[0038] In this embodiment, the data processing method runs on an electronic device (e.g., Figure 1The server / terminal device shown can acquire multi-dimensional physiological data and drug combination data of the target patient via wired or wireless connection. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultrawideband) connections, and other currently known or future-developed wireless connection methods. The implementing entity of this application is specifically a data processing system, which may be simply referred to as the system. The aforementioned multi-dimensional physiological data includes characteristic data of the target patient, which may include at least eGFR values ​​and genotypes (such as CYP450 genotypes), and may also include data such as age, gender, genotype, and medical history. The aforementioned drug combination data includes data such as the type and dosage of drugs.

[0039] Step S202: Invoke the target prediction model issued by the preset central server.

[0040] In this embodiment, the construction process of the above-mentioned target prediction model includes: 1. Problem definition: The goal is to build a prediction model with patient characteristics (such as age, gender, genes, medical history, etc.) and drug information (such as type, dosage, etc.) as inputs, and drug response (such as efficacy, side effects, etc.) as outputs. 2. Participants: Hospitals, clinics, research institutions, etc. (each institution is referred to as a client), and a central coordination server. 3. Data distribution: Each client holds local patient data, which cannot leave the local machine. The data is non-independent and identically distributed (Non-IID), meaning that the data distribution of different clients may be different (e.g., patient groups in different regions may differ). 4. Model initialization: The central server initializes a global model (e.g., a deep neural network or gradient boosting tree) and sends the initial model parameters to all clients. 5. Federated learning training loop: a. Client selection: In each round of training, the central server randomly selects a portion of clients to participate. b. Local training: Each selected client trains using local data, calculating updates to the model parameters (gradients or weight increments). The training process can use traditional supervised learning, such as minimizing the loss function between the predicted drug response and the actual response. c. Parameter Upload: The client encrypts the updated parameters (or gradients) and sends them to the central server, but does not send the original data. d. Aggregate Update: The central server uses an aggregation algorithm (such as FedAvg) to aggregate the received parameter updates and update the global model. e. Model Deployment: The central server sends the updated global model parameters to all clients. 6. Model Validation: During training, model performance can be monitored using public datasets held by the central server (such as publicly available drug response datasets) or through local validation on the client (and then aggregating evaluation metrics on the server). 7. Termination Condition: Training stops when the model converges or reaches the preset number of training epochs. 8. Model Deployment: After training is complete, the global model (i.e., the target prediction model described above) can be deployed to various clients to predict drug responses in new patients.

[0041] Step S203: Based on the target prediction model, perform prediction processing on the multi-dimensional physiological data and the drug combination data to obtain the corresponding adverse reaction assessment data.

[0042] In this embodiment, the specific implementation process of predicting the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain the corresponding adverse reaction assessment data will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0043] Step S204: Determine whether the adverse reaction assessment data is less than a preset dynamic safety threshold.

[0044] In this embodiment, the adverse reaction assessment data mentioned above refers to the adverse reaction probability. When the adverse reaction probability (PADR) is below 5% (i.e., 0.05), the drug combination is generally considered safe. However, the specific dynamic safety threshold needs to be determined based on the type of drug, the patient's health condition, and the policies of the medical institution. In some cases, physicians may be willing to accept a higher adverse reaction probability in exchange for the therapeutic effect of the drug; while in other cases, it may be necessary to strictly control the adverse reaction probability within a low range. Therefore, the dynamic safety threshold is a dynamic parameter that is adjusted according to specific circumstances.

[0045] One method is to compare the adverse reaction assessment data with the aforementioned dynamic safety threshold to determine whether the adverse reaction assessment data is less than the dynamic safety threshold.

[0046] Step S205: If yes, use a preset drug distribution tool to capture the target drug corresponding to the drug combination data.

[0047] In this embodiment, the specific implementation process of the above-mentioned drug sorting tool for capturing the target drug corresponding to the drug combination data will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0048] Step S206: Generate drug dispensing information corresponding to the target drug, and display the drug dispensing information.

[0049] In this embodiment, the medication information may include "Patient XX, please collect your medication." The system can prompt the patient to collect the medication via light, voice, or screen. Additionally, it may include visual assistance features: the screen displays an image of the tablet, its name, and instructions for use (e.g., "To be swallowed whole"). Voice guidance is also included: for visually impaired patients, the device provides voice instructions for medication administration (e.g., "Please take 2 white round tablets").

[0050] Based on the above automated processing flow, this application acquires multi-dimensional physiological data and drug combination data of the target patient. Then, using a target prediction model issued by a central server, it predicts adverse reaction assessment data based on this data. Only when the adverse reaction assessment data is detected to be below a dynamic safety threshold is the target drug corresponding to the drug combination data retrieved using a drug dispensing tool. This generates and displays the corresponding medication information. Thus, by using a target prediction model to predict and process multi-dimensional physiological data and drug combination data, this application can efficiently and accurately complete medication risk assessment, improving the efficiency and accuracy of medication risk assessment. Furthermore, when the adverse reaction assessment data is detected to be below a preset dynamic safety threshold, the drug dispensing tool can quickly retrieve the target drug corresponding to the drug combination data, generate and display the corresponding medication information, effectively reducing manual drug dispensing and improving the efficiency and accuracy of drug dispensing.

[0051] In some alternative implementations, step S203 includes the following steps:

[0052] The preset adverse reaction probability formula is invoked based on the target prediction model.

[0053] In this embodiment, the above-mentioned adverse reaction probability formula specifically includes: in, β is the Sigmoid activation function, which maps the linear combination result to the probability interval [0,1]. i This refers to the adverse reaction coefficient of drug i, used to quantify the contribution of drug i to adverse reactions, and is obtained by training on the chemical properties of the drug and clinical data. iThis refers to whether drug i is in the prescription (1 for presence, 0 for absence) or the standardized value of the dosage. α refers to the liver and kidney function attenuation coefficient, or attenuation coefficient, reflecting the impact of the patient's liver and kidney function on drug metabolism, calculated from kidney function indicators (such as eGFR) or liver function indicators (such as ALT). LiverFunc refers to the liver and kidney function score (such as eGFR value or Child-Pugh classification), used to directly quantify the patient's liver and kidney function status; the worse the liver function (the lower the score), the higher the weight of α. γ refers to the gene susceptibility coefficient, used to reflect the individual differences in drug metabolism caused by the patient's genotype (such as CYP450 enzyme polymorphism). GeneFactor refers to the gene risk score (such as CYP2C19 slow metabolizer = 1, fast metabolizer = 0), used to identify whether the patient carries a high-risk gene variant, directly affecting drug metabolism efficiency. δ refers to the medical history adjustment coefficient, used to reflect the impact of the patient's past medical history (such as allergy history, liver disease history) on the current medication risk. HistoryFactor refers to a medical history risk score (e.g., allergy history = 1, no allergy history = 0), which is used to quantify a patient's sensitivity to a specific drug using historical data.

[0054] The formula logic and dynamic adjustment mechanism of the above adverse reaction probability formula include: Linear combination layer: summarizing drug characteristics Patient physiological status (α·LiverFunc), gene risk (γ·GeneFactor), and medical history risk (δ·HistoryFactor). For example, if a patient has a low eGFR value (poor renal function), α increases, and the overall risk increases; if the patient carries the CYP2D6 slow metabolizer gene, γ increases, and the risk is further upregulated. Sigmoid activation layer: Compresses linear results to a probability range and outputs an interpretable PADR value (adverse reaction probability, or adverse reaction data). For example, the linear output is 2.5 → σ(2.5) ≈ 0.92, which is a 92% adverse reaction risk. Dynamic threshold adjustment: Base threshold: The default is set to 5% (i.e., PADR > 5% triggers an alert). Dynamic adjustment rules: Drug type: The threshold for anticoagulants (such as warfarin) is reduced to 3% (due to high bleeding risk); the threshold for antibiotics (such as penicillin) is increased to 7% (allergic reactions are easier to control). Patient characteristics: The threshold for elderly patients (>65 years old) is reduced to 4% (decreased metabolic capacity); the threshold for children is increased to 6% (larger room for dose adjustment). Institutional policy: The threshold for ICU patients has been lowered to 2% (critically ill patients requiring stricter management).

[0055] Obtain a preset coefficient corresponding to the adverse reaction probability formula; wherein the preset coefficient includes at least an adverse reaction coefficient and a decay coefficient.

[0056] In this embodiment, the preset coefficients include adverse reaction coefficients and attenuation coefficients, and may also include gene susceptibility coefficients and medical history adjustment coefficients.

[0057] The multidimensional physiological data and the drug combination data are calculated and processed based on the adverse reaction probability formula to obtain the corresponding calculation results.

[0058] In this embodiment, data analysis can be performed on the aforementioned multi-dimensional physiological data and drug combination data. Based on the parameter types required by the aforementioned adverse reaction probability formula, relevant parameters corresponding to the target patient can be obtained, including standardized dose values, liver and kidney function scores, genetic risk scores, and medical history risk scores. Then, all obtained parameters and the aforementioned preset coefficients are substituted into the corresponding positions within the aforementioned adverse reaction probability formula for calculation. The generated calculation result (adverse reaction probability) is used as the final adverse reaction assessment data.

[0059] The calculation results are used as the adverse reaction assessment data.

[0060] Based on the above processing flow, this application calls a preset adverse reaction probability formula based on the target prediction model, then obtains the preset coefficients corresponding to the adverse reaction probability formula, and then calculates and processes multi-dimensional physiological data and drug combination data based on the use of the adverse reaction probability formula. This enables the efficient and accurate generation of corresponding adverse reaction assessment data, effectively improving the generation efficiency of adverse reaction assessment data and ensuring the accuracy of the generated adverse reaction assessment data.

[0061] In some optional implementations of this embodiment, the dispensing tool includes a target detection model and a target robotic arm; step S205 includes the following steps:

[0062] Based on the target detection model, the drug information corresponding to the drug combination data is identified from the preset drug dispensing station.

[0063] In this embodiment, the dispensing station is a pre-built location in a hospital for storing medications. A high-definition camera is installed above the dispensing station to continuously capture images of the tablets at a fixed angle (e.g., a 45-degree overhead view), ensuring that each tablet is fully visible. The camera is equipped with a ring light to eliminate shadows and enhance the contrast of the tablet edges (e.g., white tablets are more easily identifiable against a black background).

[0064] Furthermore, the aforementioned target detection model can specifically employ a detection model built upon a lightweight real-time detection algorithm. For example, a pre-trained YOLOv8 model can be used to obtain a preset target control algorithm. The pre-trained target detection model is used to segment the image and identify the contour position of each tablet. Geometric analysis is then performed on the detected tablet contours: circular: aspect ratio close to 1:1, uniform edge curvature; elliptical: aspect ratio > 1.2:1, symmetrical elliptical edges; irregular: edges with obvious protrusions or depressions (e.g., heart-shaped, triangle-shaped). Subsequently, the system automatically adds electronic tags to the tablets (e.g., "circular - aspirin," "irregular - warfarin") and stores them in a database for subsequent matching. Based on the target detection model, drug information, such as drug images, matching the aforementioned drug combination data is identified from the dispensing station. In addition, dynamic calibration can be performed: before each batch of dispensing, the system randomly selects 3 tablets for manual review. If the recognition error rate > 0.5%, model fine-tuning is triggered (e.g., adjusting the contour detection threshold).

[0065] The target control algorithm generates a target capture strategy corresponding to the drug information.

[0066] In this embodiment, the target control algorithm can adopt the PID control algorithm. The above-mentioned target grasping strategy refers to the grasping strategy applied to the target robotic arm, and the corresponding strategy includes: 1. Grasping force control: 1) Force sensor integration: A six-axis force sensor is installed at the end of the robotic arm to monitor the vertical pressure (Z-axis) and lateral friction force (X / Y-axis) of the tablet during grasping in real time. 2) PID control algorithm: Proportional term (P): Quickly adjusts the grasping speed according to the current error (target force - actual force) (e.g., accelerates the downward pressure when the error is large); Integral term (I): Eliminates long-term errors (e.g., the tablet surface is smooth, causing continuous slippage); Derivative term (D): Predicts the trend of error change (e.g., slows down in advance when the tablet is about to break). Example: When grasping a round tablet, the system sets the target force to 0.5N. If the actual force is 0.3N, the PID algorithm will output a command to make the robotic arm press down at a speed of 0.2m / s until the force reaches the target. 3) Material Adaptive: The gripping strategy is dynamically adjusted according to the tablet material (such as sugar coating, film coating): Sugar-coated tablets: the initial gripping speed is reduced (from 0.5m / s to 0.3m / s) to avoid sugar coating peeling off; Film-coated tablets: the holding time after gripping is increased (from 0.2 seconds to 0.5 seconds) to ensure stable adsorption.

[0067] 2. Vibration Suppression: 1) Vibration Monitoring: Piezoelectric ceramic sensors are installed at the joints of the robotic arm to collect vibration frequency (e.g., 10-100Hz) and amplitude (e.g., 0.1-1mm) in real time. 2) Reverse Vibration Compensation: The system generates a signal with the opposite phase to the detected vibration (e.g., if the original vibration is a sine wave, the compensation signal is an anti-phase sine wave), which is output by the robotic arm motor to cancel the vibration. Example: If a 15Hz, 0.5mm vibration is detected in the Z-axis direction, the system will drive the motor to generate a 15Hz, -0.5mm reverse vibration, making the combined vibration approach zero. 3) Drug Dispensing Channel Optimization: The inner wall of the channel is made of a low-friction material (e.g., polytetrafluoroethylene) and a spiral guide groove is designed to guide the tablets to slide down naturally, reducing jamming caused by vibration.

[0068] 3. Medication dispensing efficiency optimization and motion trajectory planning: 1) Path shortestization: The A* algorithm is used to plan the robotic arm's grasping path, prioritizing densely packed areas of tablets (such as the upper layer of the pillbox) and straight-line motion trajectories to reduce idle travel time. Example: From pillbox A (coordinates [10,20]) to pillbox B (coordinates [30,40]), the system will choose a straight line instead of a broken line, saving 0.3 seconds / movement. 2) Parallel grasping strategy: For irregularly shaped tablets, a "dual-claw cooperative" mode is adopted: the main claw grasps the main body of the tablet, and the secondary claw assists in fixing the protruding part (such as the top of a heart-shaped tablet) to prevent rotation or slippage.

[0069] Based on the target robotic arm, the target drug corresponding to the drug combination data is grasped from the dispensing station according to the target grasping strategy.

[0070] In this embodiment, the target robotic arm can be controlled to automatically extract the target drug that matches the drug combination data from the dispensing station according to the strategy content of the target grasping strategy.

[0071] The system also provides an anomaly handling mechanism for drug grabbing, including: 1. Damaged tablet detection: After grabbing, a second image is taken using a camera, and image thresholding is used to detect whether the tablet edges are intact (e.g., fragments of a damaged tablet will exceed the normal outline boundary). If damage is detected, the robotic arm discards the tablet into the waste bin and marks the batch in the database as requiring manual verification. 2. Mis-dispensed tablet correction: An RFID reader is installed at the dispensing channel exit to verify the tablet label against the prescription information (e.g., drug name, dosage). If the information does not match (e.g., ibuprofen is grabbed when aspirin should be dispensed), the system immediately stops dispensing and triggers an alarm, awaiting manual intervention. 3. Efficiency target achievement: Dispensing speed: Through optimized trajectory and parallel grabbing, 30 tablets / minute is achieved (i.e., one tablet is grabbed and placed every 2 seconds). 4. Abnormality rate control: Breakage rate: Through force control and material self-adaptation, the number of broken tablets is controlled to ≤1 tablet per 10,000 tablets (0.01%); Misclassification rate: Through RFID secondary verification, the number of misclassified tablets is ensured to be ≤0.01%; Medication jamming rate: Through vibration suppression and channel optimization, the probability of irregularly shaped medication jamming is <0.01%.

[0072] In addition, the medication dispensing tool receives a list of safe medication combinations, i.e., a list of target medications (such as "Aspirin 100mg × 30 tablets + Clopidogrel 75mg × 30 tablets"), and only dispenses medications on the list. After dispensing, a medication traceability code (including tablet shape, batch number, and dispensing time) is generated and printed on the medication bag for patients to scan and verify, ensuring a closed-loop traceability of "dispensing-taking".

[0073] Based on the above processing flow, this application identifies drug information corresponding to drug combination data from the dispensing station using a target detection model, then generates a target grasping strategy corresponding to the drug information using a target control algorithm, and finally, based on the use of a target robotic arm and the target grasping strategy, can automatically and accurately grasp the target drug corresponding to the drug combination data from the dispensing station, effectively improving the grasping efficiency of the target drug and ensuring the accuracy of the grasped target drug.

[0074] In some optional implementations, the target drug is temporarily stored in a pre-set smart drug storage compartment; the target drug has an embedded target tag; after step S206, the above-mentioned electronic device may also perform the following steps:

[0075] Determine whether a medication retrieval operation triggered by the target patient has been received from the smart pharmacy.

[0076] In this embodiment, the target drug is temporarily stored in a pre-set smart drug warehouse, and the target drug is embedded with a target tag, which is specifically a miniature passive RFID tag.

[0077] The tag embedding process for the target medicine includes: Miniature passive RFID tag embedding: During tablet pressing, a 1mm × 1mm miniature passive RFID chip is embedded in the center of the tablet (thickness ≤ 0.1mm, not affecting efficacy). The tag uses the high frequency (HF) 13.56MHz or ultra-high frequency (UHF) 860-960MHz frequency band, with a cost controlled at 0.02 yuan / tablet, reducing costs through large-scale production. Drug compartment reader configuration: A loop coil reader (5cm diameter, 2mm thickness) is installed at the bottom of the medicine box or inside the smart drug compartment, at a distance ≤ 1cm from the tablets to ensure signal coverage of all tablets. The reader emits a continuously modulated radio frequency signal (power ≤ 1W, compliant with electromagnetic safety standards) to activate the tag and receive the reflected signal.

[0078] If so, obtain the drug features corresponding to the target tag based on the preset card reader.

[0079] In this embodiment, the extraction process of the above-mentioned drug features includes: 1. Tag activation and signal reflection: When a drug retrieval operation is received from the smart pharmacy, the card reader emits electromagnetic waves, and the tag modulates the reflected signal (including a unique ID and stored preset data) through the load. The reflected signal strength (RSSI) attenuates with distance, and the card reader dynamically adjusts the power (e.g., from 1W to 0.1W) to optimize the reading range. 2. Multi-tag reading strategy: An anti-collision algorithm (e.g., time-slotted ALOHA protocol) is adopted to avoid signal collisions from multiple tablets reflecting simultaneously; the card reader performs time-division scanning (50ms interval) to ensure that the signal of each tablet is completely acquired. Multi-dimensional feature analysis and comparison. 3. Feature extraction. Signal fundamental parameter analysis: Amplitude: The reflected signal strength (RSSI) is related to the distance to the tablet and its material (e.g., tablets packaged in metal foil attenuate RSSI faster); Phase: The phase shift of the tag's reflected signal (typical value 0-2π) reflects the tablet's position and minor label deformation; Frequency: The passive tag's reflection frequency is consistent with the reader's transmission frequency (13.56MHz or 915MHz), but label anomalies can be detected through harmonic analysis (e.g., counterfeit tags may generate additional harmonics). Tablet physical feature reconstruction: Dielectric constant derivation: Different materials (e.g., sugar coating ε≈3, capsule shell ε≈2) are distinguished by the degree of signal attenuation after penetrating the tablet (e.g., -3dB attenuation corresponds to a dielectric constant ε=4); Shape and size recognition: The tablet's outline (e.g., a circular shape with a diameter of 8mm or an elliptical shape with a major axis of 10mm) is reconstructed using signal scattering patterns (e.g., edge reflection intensity).

[0080] The drug characteristics are verified based on a preset feature database.

[0081] In this embodiment, the number of extracted tablets can be counted for preliminary verification. If the number of reflected signals is counted (e.g., the prescription requires 2 tablets, but 3 independent label signals are detected), an "abnormal quantity" alarm is triggered; stray signals are filtered out by signal strength stratification (e.g., RSSI > -50dBm is considered a valid label).

[0082] Then, feature database comparison processing is performed: the feature database stores the characteristic fingerprints of each type of tablet (e.g., "Aspirin: round, 8mm in diameter, dielectric constant 4.2, reflection phase shift 1.2rad"); by comparing the real-time extracted drug features with the feature database item by item, the matching threshold is set to 90% similarity (e.g., dielectric constant error ≤ 0.5, shape error ≤ 0.3mm), and the corresponding comparison results are generated. The comparison results include whether the drug features pass the verification or not.

[0083] If the drug characteristics pass the verification, the content of the drug characteristics is compared based on a preset prescription database.

[0084] In this embodiment, the aforementioned prescription database is a pre-built electronic prescription database. The process of comparing the drug features includes: if the RFID feature match is successful (e.g., a "ibuprofen 200mg" tag is detected), the system retrieves the patient's current medication information from the electronic prescription database; the comparison content includes: drug name (e.g., the prescription is "acetaminophen," but the RFID tag shows "ibuprofen"); dosage (e.g., the prescription is 200mg, but the RFID tag shows a dosage of 100mg); and administration time (e.g., the current time is not within the prescription's allowed time window), and generates corresponding content comparison results based on the obtained comparison content. The content comparison results include whether the drug feature passes the content comparison or fails. Additionally, a processing database can be used to record the RFID features of different batches of the same drug (e.g., "Aspirin batch A: phase offset 1.2rad, batch B: 1.3rad") to avoid false alarms due to batch differences.

[0085] If the drug characteristics pass the content comparison, the smart pharmacy is unlocked so that the target patient can retrieve the target drug from the smart pharmacy.

[0086] In this embodiment, the drug characteristics are only deemed to have passed verification when the drug characteristics are detected and verified through content comparison. In this way, the smart pharmacy is unlocked so that the target patient can retrieve the target drug from the smart pharmacy.

[0087] Based on the above processing flow, when this application receives a medication retrieval request from a target patient, it automatically extracts the drug characteristics corresponding to the target tag based on the use of the card reader. It then intelligently verifies the drug characteristics using a combination of the characteristic database and the prescription database, improving the accuracy of drug verification. Furthermore, the smart pharmacy is only unlocked when the drug characteristics pass verification, effectively reducing medical accidents caused by drug confusion and improving the intelligence and accuracy of the medication retrieval operation.

[0088] In some optional implementations, after the step of verifying the drug features based on a preset feature database, the electronic device may further perform the following steps:

[0089] If the drug characteristics fail the content comparison, corresponding voice prompts and alarm messages will be generated.

[0090] In this embodiment, when a drug characteristic fails the content comparison, such as a discrepancy between the tablet name or dosage and the prescription (e.g., a patient should take "antihypertensive drug A," but the RFID displays "vitamin C"), the system will automatically generate a voice prompt and alarm message corresponding to suspending dispensing from the pharmacy. The content of the voice prompt is not specifically limited and can be set according to actual business needs; for example, it could include "Please wait for confirmation from medical personnel." Similarly, the content of the alarm message is not specifically limited and can be set according to actual business needs; for example, it could include "There is an error with the medication; please handle it immediately."

[0091] The voice prompts are then broadcast.

[0092] In this embodiment, the aforementioned voice prompts can be broadcast to provide risk warnings to the target patient.

[0093] Obtain the pre-set communication information of medical staff.

[0094] In this embodiment, the aforementioned medical personnel may refer to the hospital's responsible nurse. The aforementioned communication information may be based on the hospital's HIS system for information notification.

[0095] Based on the communication information, the alarm information is sent to the medical staff.

[0096] In this embodiment, alarm information generated can be sent to medical staff through selected communication information, such as the hospital's HIS system. At the same time, information such as the patient's identity and abnormal medication information can be pushed to the responsible nurse so that the nurse can remotely view the RFID data of the pills related to the target patient through a mobile terminal and decide whether to allow medication or change the medication.

[0097] Based on the above processing flow, when the drug characteristics fail the content comparison, this application automatically and intelligently generates corresponding voice prompts and alarms, improving the efficiency and intelligence of voice prompt and alarm generation. The voice prompts are then broadcast, and alarms are sent to medical staff using communication information. This allows medical staff to review the alarms and correct any abnormalities in the target drug, thereby ensuring the accuracy and safety of medication dispensing for the target patient.

[0098] In some optional implementations of this embodiment, after step S204, the electronic device may further perform the following steps:

[0099] If the adverse reaction assessment data is greater than the dynamic safety threshold, a preset drug adjustment strategy is obtained.

[0100] In this embodiment, the drug adjustment strategy includes three strategies, ranked from highest to lowest priority: Strategy 1. Replace high-risk drugs. Matching alternative drugs: Query drug knowledge bases (such as Micromedex, Lexicomp) to screen for drugs with the same pharmacological effects as the current prescription but different metabolic pathways. Example: If a patient has a high risk of rhabdomyolysis due to statins + amiodarone, the system suggests replacing amiodarone with dronedarone (a similar antiarrhythmic drug, but without CYP3A4 metabolic conflict). Avoiding contraindications: Check whether the alternative drug conflicts with other comorbidities of the patient (such as diabetes, hypertension). For example, avoid prescribing nonsteroidal anti-inflammatory drugs (NSAIDs) for patients with renal failure. Strategy 2. Adjust dosage or frequency. Dosage gradient optimization: Adjust the dosage based on eGFR or liver enzyme activity. For example: reduce the metformin dose by 50% for patients with renal failure (eGFR < 30 mL / min); reduce the isoniazid dose from 300 mg / day to 150 mg / day for patients with slow acetylation genotype. Optimize dosing time: Adjust the timing of drug administration to reduce interactions. For example, when cimetidine (which inhibits CYP2D6) is taken with digoxin, it is recommended to change cimetidine to be taken after meals, with a 4-hour interval between cimetidine and digoxin. Strategy 3. Increase adjuvant medications or monitoring. Antagonize high-risk reactions: For drug combinations known to potentially aggravate adverse reactions, it is recommended to add an antagonist. For example, lithium salts + thiazide diuretics may increase the risk of lithium poisoning; it is recommended to supplement with triamterene and monitor blood lithium concentration. Enhance laboratory monitoring: For cases where medication cannot be replaced, increase the monitoring frequency. For example, when vancomycin + aminoglycosides are used together, it is recommended to test renal function and hearing every 2 days. Strategy 4. Phased adjustment (complex cases) Gradual replacement strategy: If the patient is taking multiple medications for a long time, it is systematically recommended to adjust in stages to avoid withdrawal reactions caused by sudden discontinuation. For example, prioritize replacing the drug with the strongest interaction (e.g., warfarin → rivaroxaban); after the patient is stable, adjust the secondary medication (e.g., halve the dose of a beta-blocker). Transitional plan: Provide temporary alternative medications (such as short-term use of low molecular weight heparin to replace warfarin, and switch back after the INR stabilizes).

[0101] The drug combination data is adjusted based on the drug adjustment strategy to obtain the corresponding specified drug combination data.

[0102] In this embodiment, a higher-priority strategy can be selected from the aforementioned drug adjustment strategies for adjustment processing, ensuring that the adverse reaction assessment data generated by the finally selected drug adjustment strategy meets the condition of being less than the aforementioned dynamic safety threshold. Specifically, the drug combination data is adjusted according to the strategy content corresponding to the selected drug adjustment strategy, and the corresponding specified drug combination data is obtained.

[0103] Based on the target prediction model, the data of the specified drug combination are predicted to obtain the corresponding specified adverse reaction assessment data.

[0104] In this embodiment, the above-described process of predicting the specified drug combination data based on the target prediction model to obtain the corresponding specified adverse reaction assessment data can be referred to the specific process of predicting the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain the corresponding adverse reaction assessment data, which will not be elaborated further here.

[0105] If the specified adverse reaction assessment data is less than the dynamic safety threshold, a decision report corresponding to the specified drug combination data is generated.

[0106] In this embodiment, if the target prediction model outputs a specified adverse reaction assessment data that is less than the aforementioned dynamic safety threshold, a decision report will be automatically constructed based on the specified drug combination data. The content of the constructed decision report may include: recommended actions: specific drug changes (e.g., "discontinue amlodipine and switch to nifedipine controlled-release tablets"); basis: cited clinical guidelines (e.g., FDA black box warning, NICE consensus); precautions: such as "blood pressure fluctuations need to be monitored after adjustment" or "the new drug may cause dizziness." Alternative options (e.g., 3 different adjustment paths for doctors to choose from) are provided.

[0107] The designated drugs corresponding to the designated drug combination data, along with the decision report, will be sent to the target patient.

[0108] In this embodiment, the process of distributing the designated drug can refer to the aforementioned drug collection process, and will not be elaborated further here. Specifically, the target patient can be notified of the drug change via email or SMS, and the aforementioned decision report can be pushed to them.

[0109] For example, for a hypertensive patient (eGFR = 45 mL / min, CYP2C92 / 3 genotype) currently prescribed amlodipine + loxoprofen, the PADR predicts an 18% risk of kidney injury. Adjustment process: Risk identification: Loxoprofen is metabolized by CYP2C9, and the patient is a slow metabolizer, potentially leading to elevated blood drug concentrations, and adverse reaction assessment data exceeding the pre-set dynamic safety threshold. Drug replacement strategies employed included: recommending discontinuation of loxoprofen and switching to celecoxib (low CYP2C9 dependence); if the patient has a history of peptic ulcers, further recommendations included acetaminophen (with liver function monitoring). Dosage optimization: The amlodipine dose was reduced from 10 mg / day to 5 mg / day to avoid further decrease in renal blood flow. Enhanced monitoring: Serum creatinine was measured weekly for 4 consecutive weeks after adjustment.

[0110] Based on the above processing flow, when the original adverse reaction assessment data is detected to be greater than the dynamic safety threshold, this application will intelligently adjust the drug combination data based on the use of the drug adjustment strategy to obtain the specified drug combination data. By comparing the adverse reaction assessment data with the threshold again, when the specified adverse reaction assessment data is detected to be less than the dynamic safety threshold, a decision report corresponding to the specified drug combination data will be automatically generated. Then, the specified drug corresponding to the specified drug combination data and the decision report will be sent to the target patient. This can achieve intelligent and accurate adjustment of the drug combination data, improve the safety of drug compatibility processing, effectively prevent adverse drug reactions in patients, and provide a safe foundation for subsequent drug dispensing and administration.

[0111] In some optional implementations of this embodiment, the target drug has an embedded target label; after step S206, the electronic device may further perform the following steps:

[0112] During the process of the target patient taking the medication corresponding to the target drug, the hand trajectory and swallowing characteristics of the target patient are obtained based on the preset millimeter-wave radar.

[0113] In this embodiment, a 24GHz / 77GHz millimeter-wave radar is installed in the patient's medication area (such as above the bedside table or on the wall). The radar's elevation angle (e.g., 30°) and horizontal coverage area (e.g., a 1-meter radius) are adjusted to ensure that the patient's head, hands, and medication box are within the monitoring field of view. The radar transmits frequency-modulated continuous wave (FMCW), and the target distance is calculated by measuring the signal round-trip time. The scanning frequency is set to 10Hz (data updated 10 times per second).

[0114] The process of detecting human proximity signals includes: continuous radar analysis of a range-Doppler map: the distance axis identifies static objects (such as medicine boxes or water cups) and dynamic objects (such as patient movement) within 0.5-1.5 meters; the Doppler axis detects the speed of the object's movement through frequency shift (e.g., a patient walking towards a medicine box at a speed of 0.5 m / s). When a human body is detected entering a preset area (e.g., within 0.8 meters of a medicine box), the radar locks onto the head area (based on the distance difference between the head and torso). The activation process of hand micro-movement monitoring includes: when the hand enters the radar's field of view (within 0.3 meters of the head), the system switches to micro-Doppler analysis mode: the radar emits a higher resolution signal (bandwidth extended to 1 GHz) to capture the frequency modulation (typical frequency shift range: 10 Hz-1 kHz) generated by minute hand movements (e.g., fingers pinching a pill, wrist rotation). The signal is converted into a time-frequency map through short-time Fourier transform (STFT) to extract the hand movement trajectory (e.g., a parabolic path from the medicine box to the mouth).

[0115] In addition, the anti-jamming design includes: the radar adopts a MIMO (Multiple Input Multiple Output) antenna array, and uses beamforming technology to suppress background interference (such as swaying curtains or moving pets); dynamic filtering is performed on static objects (such as medicine boxes) to retain only the signals of moving targets.

[0116] In addition, the swallowing motion feature extraction process includes: laryngeal motion monitoring: the radar is focused on the patient's neck area (0.1-0.2 meters from the head), and small displacements of the laryngeal tissues are detected through high-resolution distance imaging (HRRP) (such as the Adam's apple moving down 2-3 cm during swallowing). Temporal features (such as duration 0.5-1.5 seconds) and frequency domain features (such as vibration frequency 3-8 Hz, corresponding to the rhythm of laryngeal muscle contraction) of the swallowing motion are extracted.

[0117] Obtain the pill reflection data corresponding to the medication taken by the target patient.

[0118] In this embodiment, the detection process of the above-mentioned tablet reflection data includes: at the moment the tablet enters the oral cavity, the radar detects the dielectric constant (reflecting the material, such as the dielectric constant of sugar-coated tablets ≈ 3-5, and that of film-coated tablets ≈ 5-8) and the radar cross-section (RCS) (reflecting the size, such as the RCS of an 8mm diameter tablet ≈ 0.5cm). 2 ).

[0119] The hand trajectory, swallowing features, and pill reflex data are fused together to obtain corresponding fused data.

[0120] In this embodiment, the above-mentioned fusion process includes: associating the hand trajectory (such as the path of "medicine box → hand → mouth") with the timestamps of swallowing action and drug reflex data: if the interval between the time when the hand reaches the mouth (T1) and the time when swallowing begins (T2) is less than 0.5 seconds, and the swallowing duration is within the normal range (0.8±0.3 seconds), then the action continuity is determined to be passed.

[0121] The fused data is verified based on a preset verification strategy.

[0122] In this embodiment, the above verification process includes: comparing the generated fusion data with data in a preset tablet feature library, with an allowable error range of: size: ±0.5mm; dielectric constant: ±10%, and generating a verification result for the fusion data based on the obtained comparison result. If the comparison result is a pass, the fusion data is determined to have passed verification; otherwise, if the comparison result is a fail, the fusion data is determined to have failed verification.

[0123] If the fused data passes verification, a medication record corresponding to the target patient is generated, and the medication record is recorded.

[0124] In this embodiment, when the hand trajectory, swallowing features, and pill features are all verified, the system will generate a medication event record (including timestamp, pill type, and patient ID), encrypt the data and upload it to the electronic health record (EHR) for doctors to review later, and display a "medication successful" prompt to the patient via the bedside screen or mobile APP (such as a green checkmark icon + voice feedback).

[0125] In addition, the handling of related abnormal situations includes: 1. No swallowing detected: The system waits 10 seconds and then re-detects. If there is still no swallowing signal, a tiered alarm is triggered: Level 1 alarm (first abnormality): A reminder is pushed to the patient's mobile phone (e.g., "Please confirm whether you have taken your medication?"); Level 2 alarm (unresolved within 30 minutes): Medical staff are notified (e.g., by sending a text message to the nurses' station through the hospital's HIS system). 2. Pill characteristics mismatch: The current medication record is immediately frozen, and the patient is asked to take the medication again; if the re-detection still results in a mismatch, the system marks it as a "high-risk event" and initiates a manual review process (e.g., the nurse confirms the patient's status via video call).

[0126] Based on the above processing flow, this application acquires the hand trajectory and swallowing features of the target patient extracted based on millimeter-wave radar, as well as the pill reflection data corresponding to the drug taken by the target patient, during the process of the target patient taking the corresponding medication. The hand trajectory, swallowing features, and pill reflection data are then fused to obtain fused data. The fused data is automatically verified based on the use of a verification strategy. This enables non-contact monitoring to cover the entire medication behavior process, resolves privacy disputes associated with traditional cameras, improves the accuracy of medication behavior verification, helps ensure that patients take medication as prescribed, and provides data support for medication adherence management.

[0127] In some alternative implementations, the user information obtained is subject to user consent and complies with relevant laws and policies.

[0128] Furthermore, any software tools or components not belonging to our company that appear in the embodiments of this application are merely illustrative examples and do not represent actual use.

[0129] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0130] It should be emphasized that, in order to further ensure the privacy and security of the above-mentioned medication information, the medication information can also be stored in a blockchain node.

[0131] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0132] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0133] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of a data processing apparatus, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0134] like Figure 3 As shown, the data processing device 300 described in this embodiment includes: a first acquisition module 301, a calling module 302, a first prediction module 303, a first judgment module 304, a capture module 305, and a first processing module 306. Wherein:

[0135] The first acquisition module 301 is used to acquire multi-dimensional physiological data and drug combination data of the target patient;

[0136] Module 302 is used to invoke the target prediction model issued by the preset central server;

[0137] The first prediction module 303 is used to predict and process the multidimensional physiological data and the drug combination data based on the target prediction model to obtain the corresponding adverse reaction assessment data.

[0138] The first judgment module 304 is used to determine whether the adverse reaction assessment data is less than a preset dynamic safety threshold.

[0139] The capture module 305 is used to capture the target drug corresponding to the drug combination data based on a preset drug distribution tool if the condition is met.

[0140] The first processing module 306 is used to generate drug dispensing information corresponding to the target drug and to display the drug dispensing information.

[0141] In some optional implementations of this embodiment, the first prediction module 303 includes:

[0142] The submodule is invoked to invoke a preset adverse reaction probability formula based on the target prediction model;

[0143] The first acquisition submodule is used to acquire a preset coefficient corresponding to the adverse reaction probability formula; wherein, the preset coefficient includes at least an adverse reaction coefficient and a decay coefficient;

[0144] The calculation submodule is used to calculate and process the multidimensional physiological data and the drug combination data based on the adverse reaction probability formula to obtain the corresponding calculation results;

[0145] A submodule is defined to use the calculation results as the adverse reaction assessment data.

[0146] In some optional implementations of this embodiment, the dispensing tool includes a target detection model and a target robotic arm; the grasping module 305 includes:

[0147] The identification submodule is used to identify drug information corresponding to the drug combination data from the preset drug dispensing station based on the target detection model;

[0148] The second acquisition submodule is used to acquire the preset target control algorithm;

[0149] A generation submodule is used to generate a target capture strategy corresponding to the drug information based on the target control algorithm.

[0150] The grasping submodule is used to grasp the target drug corresponding to the drug combination data from the dispensing station based on the target robotic arm and according to the target grasping strategy.

[0151] In some optional implementations of this embodiment, the target drug is temporarily stored in a preset smart drug warehouse; the target drug has an embedded target tag; the data processing device further includes:

[0152] The second judgment module is used to determine whether a drug retrieval operation triggered by the target patient for the smart pharmacy has been received;

[0153] The second acquisition module is used to acquire, if so, drug features corresponding to the target tag extracted by a preset card reader;

[0154] The verification module is used to verify the drug characteristics based on a preset feature database;

[0155] The comparison module is used to perform content comparison of the drug features based on a preset prescription database if the drug features pass the verification.

[0156] The unlocking module is used to unlock the smart pharmacy if the drug characteristics pass the content comparison, so that the target patient can take the target drug out of the smart pharmacy.

[0157] In some optional implementations of this embodiment, the data processing apparatus further includes:

[0158] The first generation module is used to generate corresponding voice prompts and alarm messages if the drug features fail the content comparison.

[0159] The broadcast module is used to broadcast the voice prompt information.

[0160] The third acquisition module is used to acquire the preset communication information of medical staff;

[0161] The sending module is used to send the alarm information to the medical staff based on the communication information.

[0162] In some optional implementations of this embodiment, the data processing apparatus further includes:

[0163] The fourth acquisition module is used to acquire a preset drug adjustment strategy if the adverse reaction assessment data is greater than the dynamic safety threshold.

[0164] The adjustment module is used to adjust the drug combination data based on the drug adjustment strategy to obtain the corresponding specified drug combination data.

[0165] The second prediction module is used to perform prediction processing on the specified drug combination data based on the target prediction model to obtain the corresponding specified adverse reaction assessment data.

[0166] The second generation module is used to generate a decision report corresponding to the specified drug combination data if the specified adverse reaction assessment data is less than the dynamic safety threshold.

[0167] The distribution module is used to distribute the designated drugs corresponding to the designated drug combination data, as well as the decision report, to the target patient.

[0168] In some optional implementations of this embodiment, the target drug has an embedded target label; the data processing device further includes:

[0169] The fifth acquisition module is used to acquire the hand trajectory and swallowing features of the target patient based on a preset millimeter-wave radar during the process of the target patient taking the medication corresponding to the target drug.

[0170] The sixth acquisition module is used to acquire the pill reflection data corresponding to the medicine taken by the target patient;

[0171] The fusion module is used to fuse the hand trajectory, the swallowing features, and the pill reflex data to obtain corresponding fused data;

[0172] The verification module is used to perform verification processing on the fused data based on a preset verification strategy.

[0173] The second processing module is used to generate a medication record corresponding to the target patient if the fused data passes verification, and to record and process the medication record.

[0174] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0175] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0176] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0177] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may also include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for data processing methods. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0178] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, for example, to execute computer-readable instructions for the data processing method.

[0179] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0180] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the data processing method described above.

[0181] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0182] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A data processing method, characterized by, The method comprises the following steps: acquiring multi-dimensional physiological data and drug combination data of a target patient; calling a target prediction model issued by a preset central server; performing prediction processing on the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain adverse reaction evaluation data; determining whether the adverse reaction evaluation data is less than a preset dynamic safety threshold; if yes, based on a preset drug dispensing tool, a target drug corresponding to the drug combination data is grabbed; generating and displaying drug taking information corresponding to the target drug.

2. The data processing method according to claim 1, characterized in that, The step of performing prediction processing on the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain adverse reaction evaluation data comprises the following steps: calling a preset adverse reaction probability formula based on the target prediction model; acquiring a preset coefficient corresponding to the adverse reaction probability formula; wherein the preset coefficient comprises at least an adverse reaction coefficient and a decay coefficient; performing calculation processing on the multi-dimensional physiological data and the drug combination data based on the adverse reaction probability formula to obtain a calculation result; using the calculation result as the adverse reaction evaluation data.

3. The data processing method of claim 1, wherein, The drug dispensing tool comprises a target detection model and a target mechanical arm. The step of grabbing the target drug corresponding to the drug combination data based on the preset drug dispensing tool comprises the following steps: identifying drug information corresponding to the drug combination data from a preset drug dispensing table based on the target detection model; acquiring a preset target control algorithm; generating a target grabbing strategy corresponding to the drug information based on the target control algorithm; grabbing the target drug corresponding to the drug combination data from the drug dispensing table based on the target mechanical arm according to the target grabbing strategy.

4. The data processing method according to claim 1, characterized in that, The target drug is temporarily stored in a preset intelligent medicine cabinet. The target drug has a target label embedded therein. After the step of generating and displaying the drug taking information corresponding to the target drug, the following steps are further included: determining whether a drug taking operation triggered by the target patient on the intelligent medicine cabinet is received; if yes, acquiring a drug feature corresponding to the target label extracted based on a preset card reader; performing checking processing on the drug feature based on a preset feature database; if the drug feature passes the checking, performing content comparison on the drug feature based on a preset prescription database; if the drug feature passes the content comparison, performing unlocking processing on the intelligent medicine cabinet so that the target patient can take the target drug from the intelligent medicine cabinet.

5. The data processing method according to claim 4, characterized in that, After the step of performing checking processing on the drug feature based on the preset feature database, the following steps are further included: if the drug feature does not pass the content comparison, generating corresponding voice prompt information and alarm information; playing the voice prompt information; acquiring communication information of a medical staff; based on the communication information, sending the alarm information to the medical staff.

6. The data processing method of claim 1, wherein, After the step of determining whether the adverse reaction evaluation data is less than the preset dynamic safety threshold, the following steps are further included: If the adverse reaction evaluation data is greater than the dynamic safety threshold, a preset drug adjustment strategy is acquired; Based on the drug adjustment strategy, the drug combination data is adjusted to obtain corresponding specified drug combination data; Based on the target prediction model, the specified drug combination data is predicted to obtain corresponding specified adverse reaction evaluation data; If the specified adverse reaction evaluation data is less than the dynamic safety threshold, a decision report corresponding to the specified drug combination data is generated; The specified drug corresponding to the specified drug combination data and the decision report are issued to the target patient.

7. The data processing method of claim 1, wherein, The target drug has a target tag embedded therein; After the step of generating the drug taking information corresponding to the target drug and performing the display processing on the drug taking information, the method further comprises: During the execution of the drug taking process corresponding to the target drug by the target patient, a hand trajectory and a swallowing feature of the target patient extracted based on a preset millimeter wave radar are acquired; Pill reflection data corresponding to the drug taken by the target patient are acquired; The hand trajectory, the swallowing feature, and the pill reflection data are fused to obtain corresponding fusion data; The fusion data are verified based on a preset verification strategy; If the fusion data pass the verification, a drug taking archive corresponding to the target patient is generated, and the drug taking archive is recorded.

8. A data processing apparatus, characterized by, Comprise: A first acquisition module is configured to acquire multi-dimensional physiological data of a target patient and drug combination data; A calling module is configured to call a target prediction model issued by a preset central server; A first prediction module is configured to predict the multi-dimensional physiological data and the drug combination data based on the target prediction model to obtain corresponding adverse reaction evaluation data; A first judgment module is configured to judge whether the adverse reaction evaluation data is less than a preset dynamic safety threshold; A grabbing module is configured to, if yes, grab a target drug corresponding to the drug combination data based on a preset drug distribution tool; A first processing module is configured to generate drug taking information corresponding to the target drug and perform display processing on the drug taking information.

9. A computer device, comprising: A memory and a processor are included, the memory has computer readable instructions stored therein, and the processor implements the steps of the data processing method of any one of claims 1 to 7 when executing the computer readable instructions.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium has computer readable instructions stored thereon, and the computer readable instructions are executed by the processor to implement the steps of the data processing method of any one of claims 1 to 7.

Images