Postpartum infant feeding guidance system and device for epileptic

Abstract

The invention relates to a postpartum infant feeding guidance system and device for epileptics, and relates to the technical field of medical instruments. A planning unit of the feeding guidance device is used for making a feeding plan for a user and comprises an acquisition module used for acquiring the medicine taking condition and the body information of the user and an analysis module used for generating the feeding plan according to the acquired data, and the analysis module can determine the feeding plan according to the characteristics of the medicine taken by the user. The analysis module analyzes the metabolic process of the user in the body of the user in combination with the body information of the user so as to plan the feeding time and generate an optimal feeding plan, and the analysis module can grade the feeding time according to the predicted drug concentration level and can distinguish the lactation time and the feeding forbidding time according to the grade of the feeding time. Therefore, a feeding plan is formed. The feeding guidance system comprises the feeding guidance device and the server, and the server is used for achieving data interaction with the feeding guidance device and completing database and software updating.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and in particular to a postpartum infant feeding guidance system and device for epileptic patients. Background Art

[0002] Epilepsy is a chronic brain disease characterized by recurrent abnormal discharges of brain neurons, leading to transient brain dysfunction. Such abnormal discharges may cause a variety of symptoms, including loss of consciousness, muscle twitches, sensory abnormalities, or behavioral changes, depending on which area of ​​the brain is affected and how far the discharge spreads. Epilepsy can be caused by a variety of factors, including genetic factors, brain damage, central nervous system infections, brain tumors, or metabolic disorders. Although treatments for epilepsy include medication, surgery, and lifestyle adjustments, there is currently no cure, and patients usually require long-term management to control symptoms.

[0003] Among the epilepsy patients, women of childbearing age account for a considerable proportion. Breastfeeding, as the most ideal feeding method for newborns, is crucial to the healthy growth of infants. However, for mothers with epilepsy, the use of anti-epileptic drugs (AEDs) may be passed to the baby through breast milk, posing potential health risks.

[0004] In the existing technology, there are significant deficiencies in the management of epilepsy patients during breastfeeding, especially in terms of how to effectively control and reduce the potential risks of anti-epileptic drugs (AEDs) to infants while ensuring the safety of mothers and infants. A key technical problem is the lack of effective real-time monitoring methods to assess the concentration of AEDs in the mother's body and its transmission in breast milk, which makes it difficult for mothers with epilepsy to obtain accurate breastfeeding advice. In addition, the lack of personalized feeding plans tailored to the mother's specific health status and medication use makes it difficult for mothers to find a balance between ensuring treatment efficacy and reducing the risk of drug exposure to infants. At the same time, epilepsy patients often face challenges in self-management, including poor medication compliance, insufficient understanding of the disease, and limited ability to cope with stress and emotional problems, all of which may affect the quality and safety of breastfeeding. There is currently a lack of a comprehensive intervention program to systematically address these issues, including medication management, feeding time planning, monitoring of maternal health status, and protective measures for infants.

[0005] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background technology. Summary of the invention

[0006] In view of the deficiencies of the prior art, the present invention provides a postpartum infant feeding guidance system and device for epileptic patients to solve at least part of the above-mentioned technical problems.

[0007] The present invention discloses a postpartum infant feeding guidance device for epileptic patients. The planning unit configured in the feeding guidance device is used to formulate a feeding plan for a user who is an epileptic patient. The feeding guidance device includes a collection module for obtaining the user's medication status and physical information and an analysis module for generating a feeding plan based on the collected data. The analysis module can analyze the metabolic process of the drug in the user's body based on the characteristics of the drug taken by the user and the user's physical information, so as to plan the feeding time and generate an optimal feeding plan. The analysis module can classify the feeding time according to the predicted drug concentration level, and can distinguish between the breastfeeding time and the fasting time according to the feeding time level, so as to form a feeding plan.

[0008] The feeding guidance device of the present invention can provide a personalized feeding plan for epilepsy patients to ensure the safety and scientificity of breastfeeding. By collecting the user's medication status and physical information, the analysis module can accurately simulate the metabolic process of the drug in the user's body, thereby predicting the drug concentration level. This prediction not only takes into account the characteristics of the drug, but also combines the individual differences of the user, such as weight, age, and liver and kidney function. By dividing the feeding time into levels, the device can intelligently distinguish between breastfeeding time and fasting time, reducing the risk of drugs being passed to the baby through breast milk. In addition, this intelligent feeding plan helps to improve the mother's medication compliance, reduce potential health risks caused by improper drug management, and support the healthy growth of the baby.

[0009] According to a preferred embodiment, the acquisition module includes one or more of an input component capable of acquiring data information input by a user based on human-computer interaction technology, an entry component capable of acquiring information on medications taken by a user through image acquisition, a drug detection component capable of detecting drug concentrations in a user's body in real time, and a physiological parameter monitoring component for collecting various physiological data of the user, wherein the acquisition module is communicatively connected to the analysis module.

[0010] The feeding guidance device of the present invention can comprehensively collect user data through a variety of collection modules to support the formulation of accurate feeding plans. The input component allows users to easily input data through human-computer interaction technology, while the entry component automatically obtains drug information through image acquisition technology, reducing the tediousness and error rate of manual input. The drug detection component monitors drug concentration in real time through a microneedle array and a non-invasive optical sensor, providing a painless and continuous monitoring method. The physiological parameter monitoring component collects the user's physiological data and provides additional information for the analysis of the drug metabolism process. The integration of these modules enables the device to understand the user's condition from multiple dimensions, improving the accuracy and practicality of the feeding plan.

[0011] According to a preferred embodiment, the drug detection component is configured with a microneedle array integrated with a microfluidic channel to guide the collected body fluid sample to the analysis chip, thereby achieving painless collection of body fluid samples through the skin for analysis, and / or is configured with a non-invasive optical sensor to penetrate the skin through near-infrared spectroscopy analysis technology to detect the concentration of specific drugs, thereby continuously monitoring the drug level in the user's body.

[0012] The drug detection component of the present invention can monitor the drug concentration in the user's body by means of minimally invasive or even non-invasive monitoring means. The use of microneedle arrays allows painless collection of body fluid samples, while the integration of microfluidic channels and analysis chips makes sample analysis faster and more accurate. The application of non-invasive optical sensors allows the detection of drug concentration through the skin through near-infrared spectroscopy analysis technology. This method not only reduces interference to the user, but also provides continuous drug concentration monitoring, thereby providing real-time data for the analysis module and optimizing the feeding plan.

[0013] According to a preferred embodiment, the data processing component configured in the analysis module is capable of performing preprocessing operations and feature extraction on the data information obtained by the acquisition module, so that the data in a unified standard format can be input into the pharmacokinetic model built into the data processing component, so as to simulate the absorption, distribution, metabolism and excretion process of the drug in the human body through numerical solution, thereby predicting the changes in drug concentration and information on the distribution of the drug in various organs.

[0014] The data processing component of the present invention can effectively process and analyze a large amount of collected data to predict the dynamic changes of drugs in the user's body. Through preprocessing operations and feature extraction, the data processing component ensures the quality and consistency of the data, making it suitable for the numerical solution of the pharmacokinetic model. The establishment and application of this model enables the device to accurately simulate the absorption, distribution, metabolism and excretion process of the drug, providing a scientific basis for formulating a feeding plan.

[0015] According to a preferred embodiment, the status monitoring unit of the feeding guidance device is configured with a microphone for acquiring sound data of the infant, a camera for acquiring behavior data of the infant and / or an environmental sensor for acquiring environmental data of the infant, wherein the sound data includes the frequency, duration and type of crying; the behavior data includes activity level, sleep pattern, posture and body position changes; and the environmental data includes temperature, humidity, light intensity and ambient noise level.

[0016] The status monitoring unit of the present invention can comprehensively monitor the status of infants and young children, including sound, behavior and environmental data. This comprehensive monitoring helps to timely detect the needs and abnormal conditions of infants and young children, and provide real-time feedback to mothers. For example, by analyzing the frequency and duration of crying, the device can determine whether the baby is hungry or uncomfortable; by monitoring activity levels and sleep patterns, the device can assess the health of the baby. The collection and analysis of this data provides additional support for mothers, helping them to better respond to and care for their babies.

[0017] According to a preferred embodiment, the status monitoring unit can judge the status of the infant based on the various types of sensor data obtained to confirm whether the infant's status is abnormal. When the infant's status is abnormal, the status monitoring unit can directly send an alarm signal or send the alarm signal to the planning unit, and use the built-in reminder module of the planning unit to inform the user, wherein the abnormal infant status may include a first abnormality caused by hunger factors and a second abnormality caused by other factors.

[0018] The status monitoring unit of the present invention can timely identify and respond to abnormal conditions of infants and young children. The status monitoring unit determines whether the infant and young child are abnormal, such as hungry or uncomfortable, by analyzing sound, behavior and environmental data. When an abnormality is detected, the device can immediately send out an alarm signal to remind the mother to take appropriate actions. This timely feedback mechanism not only improves the safety of infant care, but also reduces the pressure of mothers' supervision.

[0019] According to a preferred embodiment, the reminder module can remind the user to perform the feeding operation when any feeding time starting node of the feeding plan is reached, wherein the reminder module can adaptively adjust the reminder mode according to the user's daily habits and physiological rhythms based on various data information obtained by the acquisition module, and the reminder mode includes sound, vibration or vision.

[0020] The reminder module of the present invention can adaptively adjust the reminder mode according to the user's daily habits and physiological rhythm. This personalized reminder service allows mothers to receive feeding reminders in time when they are busy or resting, thereby ensuring timely feeding and meeting the needs of the baby. The realization of the adaptive reminder mode benefits from the user data obtained by the collection module, making the reminder more humane and effective.

[0021] According to a preferred embodiment, the feeding guidance device is configured with an analysis test strip for connecting to the planning unit in a pluggable manner. The analysis test strip can absorb part of the breast milk sample when the user performs the feeding operation and be inserted into a slot provided on the planning unit to reach the detection area of ​​the drug detection component of the collection module, thereby reading the color change on the analysis test strip through the drug detection component and sending it to the data processing component, so as to use the data processing component to analyze the color change and calculate the drug concentration.

[0022] The feeding guidance device of the present invention provides a simple and rapid method for drug concentration detection by using an analytical test strip. The user uses the analytical test strip to collect a breast milk sample before feeding, and reads the color change through the device to obtain real-time data on drug concentration. This method is not only convenient for user operation, but also can provide accurate drug concentration information to the data processing component, further optimizing the feeding plan.

[0023] According to a preferred embodiment, the data processing component generates a final feeding time recommendation after considering the prediction results of the pharmacokinetic model, the output of the machine learning algorithm and the pre-set safety threshold, wherein the data processing component can compare the difference between the drug concentration predicted by the model and the drug concentration obtained by analysis and calculation, thereby determining the accuracy of the model, and can train models that are below the accuracy threshold.

[0024] The data processing component of the present invention can combine pharmacokinetic model predictions, machine learning algorithms and safety thresholds to generate accurate feeding time recommendations. By comparing predicted and actual measured drug concentrations, the device can evaluate the accuracy of the model and train and optimize the model if necessary. This continuous self-learning and adjustment mechanism ensures the accuracy and reliability of feeding guidance.

[0025] The present invention also discloses a postpartum infant feeding guidance system for epileptic patients, which includes one or more of the aforementioned feeding guidance devices and a server communicatively connected to the feeding guidance device, wherein the server is used to implement data interaction with the feeding guidance device and complete database and software updates.

[0026] The feeding guidance system of the present invention realizes centralized data management and remote software update through data interaction between the server and the feeding guidance device. This systematic method not only improves the efficiency of data management, but also can continuously introduce the latest research results and improvements through software updates, thereby improving the guidance effect and user experience of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the hardware structure of a planning unit of a feeding guidance device according to a preferred embodiment of the present invention;

[0028] Figure 2 It is a working logic diagram of a planning unit of a preferred embodiment provided by the present invention;

[0029] Figure 3 It is a schematic diagram of the hardware structure of a feeding guidance device according to a preferred embodiment of the present invention;

[0030] Figure 4 This is a working logic diagram of a state monitoring unit of a preferred embodiment provided by the present invention;

[0031] Figure 5 A schematic diagram of a use scenario of a feeding guidance device according to a preferred embodiment of the present invention;

[0032] Figure 6 A hardware connection diagram of a feeding guidance system according to a preferred embodiment of the present invention.

[0033] Reference numerals list

[0034] 100: planning unit; 110: acquisition module; 111: input component; 112: entry component; 113: drug detection component; 114: physiological parameter monitoring component; 120: analysis module; 121: data processing component; 130: reminder module; 140: analysis test paper; 200: status monitoring unit; 210: microphone; 220: camera; 230: environmental sensor; 300: server. DETAILED DESCRIPTION

[0035] The following is a detailed description with reference to the accompanying drawings.

[0036] Example 1

[0037] The present invention provides a postpartum infant feeding guidance device, which is aimed at infants from birth to 1 year old and infants from 1 year old to 3 years old. Since whether to adopt breastfeeding depends on the choice of the family and various practical factors, there are cases where some children are weaned early in the infant stage, and there may be cases where some children continue to eat breast milk when they are 2 years old or even older (this situation is called "long-term breastfeeding" or "natural weaning"). Therefore, the feeding guidance device of the present invention is mainly aimed at (postpartum) infants and young children who are breastfed, and is not limited to newborns. Further, the users of the feeding guidance device of the present invention are usually breastfeeding mothers, and especially patients with special diseases, such as epilepsy patients. Among them, these patients with special diseases usually need to take corresponding drugs for a long time to control or alleviate the disease and prevent the onset of the disease. However, some drugs will enter the body of infants and young children through breast milk, causing potential health hazards to them. However, if the dosage of the drugs is reduced or even stopped, it will also cause great harm to their own health. Therefore, these users can use the feeding guidance device of the present invention to ensure that they can use the drugs normally while reducing or even avoiding the harm to infants and young children. Furthermore, in addition to epilepsy, the above-mentioned special diseases may also include thyroid disease, depression, hypertension, etc. For example, mothers with hyperthyroidism or hypothyroidism may need to take some drugs, such as propylthiouracil, to regulate thyroid hormone levels. These drugs can be passed to infants through breast milk and may affect the thyroid function of infants; mothers with depression can pass antidepressants, especially selective serotonin reuptake inhibitors (SSRIs), into breast milk when taking them, and may have side effects on infants; some drugs used to treat hypertension, such as ACE inhibitors, beta-blockers and calcium channel blockers, may be secreted in breast milk and may have an impact on infants, especially at higher doses.

[0038] Based on this, the postpartum infant feeding guidance device of the present invention can be particularly suitable for patients with special diseases such as epilepsy, including Figure 1 The planning unit 100 for formulating a feeding plan is shown, wherein the planning unit 100 can collect data information related to the user, especially the user's medication status, and analyze the metabolic process of the medication in the user's body according to the characteristics of the medication taken by the user and the user's physical information, so as to plan the feeding time and generate the optimal feeding plan. The present invention will be illustrated by taking epilepsy patients as an example, but patients with other special diseases are also the target group of the feeding guidance device of the present invention.

[0039] Preferably, the planning unit 100 can be configured as a wearable device, such as a smart watch or a smart bracelet, so that the user can wear the planning unit 100 on the body. Figure 2The schematic diagram of the working logic of the planning unit 100 in a preferred embodiment is shown in FIG.

[0040] Preferably, the planning unit 100 may be configured with a collection module 110 for acquiring one or more data information, wherein the collection module 110 may include one or more of an input component 111 , an entry component 112 , a drug detection component 113 and a physiological parameter monitoring component 114 .

[0041] Preferably, the input component 111 can obtain data information input by the user in a variety of different ways based on human-computer interaction technology, wherein the input component 111 may include a (high-resolution) capacitive touch screen and a voice recognition chip, the touch screen may use multi-touch technology to support gesture operations, thereby improving input efficiency; the voice recognition chip may use a deep learning algorithm to accurately recognize multiple languages ​​and dialects, and support natural language processing, so that users can input information orally.

[0042] Preferably, the input component 112 can obtain information such as the type and dosage of the medicine taken by the user through image acquisition. The input component 112 may include a (high-resolution) camera lens and a (high-speed) image processing chip. The camera lens uses autofocus technology to quickly capture clear images under different light conditions; the image processing chip can decode barcodes and two-dimensional codes in various formats in real time, including common EAN, UPC, QR codes, etc. This design not only improves the accuracy of data input, but also greatly reduces the time of manual input, and is particularly suitable for postpartum mothers.

[0043] Preferably, the drug detection component 113 can detect the drug concentration in the user's body in real time in a variety of ways, one of which is to use a microneedle array made of biocompatible materials, which is usually between 200 and 300 microns in length, can penetrate the surface of the skin but will not touch the nerve endings, so it will not cause pain. The microneedle array is integrated with a microfluidic channel, which can guide the collected body fluid samples to the analysis chip, thereby realizing the painless collection of trace body fluid samples through the skin for analysis; another method is to use a non-invasive optical sensor integrated on a wearable device to penetrate the skin through near-infrared spectroscopy analysis technology to detect the concentration of a specific drug, so that the user can continuously monitor the drug level without affecting daily activities. Preferably, the drug detection component 113 of the present invention can be configured with one or all of the microneedle array and the non-invasive optical sensor.

[0044] Preferably, the microneedle array can be manufactured using a microelectromechanical system (MEMS) process and made of biocompatible materials such as silicon, gold or titanium, wherein the tip diameter of each microneedle is less than 1 micron to ensure that nerve endings are not stimulated during puncture. Preferably, the surface of the microneedle array is coated with a hydrophilic material, which can absorb trace amounts of body fluids through capillary action. Preferably, the microfluidic channel uses precision etching technology, and the inner wall of the microfluidic channel is coated with anti-coagulation material to ensure that the sample can flow smoothly to the analysis chip. Preferably, the analysis chip integrates a micro-electrochemical sensor and a micro-optical sensor, wherein the electrochemical sensor uses enzyme electrode technology to specifically detect specific drug molecules; the optical sensor uses the principle of fluorescent immunoassay to identify target drug molecules through specific antibodies. The combination of these two technologies improves the accuracy and sensitivity of detection and can simultaneously detect the concentrations of multiple drugs.

[0045] Preferably, the non-invasive optical sensor can use near-infrared spectroscopy (NIRS) to achieve detection by utilizing the absorption characteristics of different molecules to near-infrared light of a specific wavelength. Preferably, the non-invasive optical sensor can use multiple light-emitting diodes (LEDs) as light sources, covering a wavelength range of 750 to 2500 nm, and its receiving end can use a highly sensitive photodiode array to detect weak light signals passing through the skin, and its signal processing chip can use algorithms such as Fourier transform and principal component analysis to extract drug concentration information from complex spectral data.

[0046] Preferably, the physiological parameter monitoring component 114 can be used to collect various physiological data of the user, which may affect drug metabolism. For example, changes in heart rate and activity may affect the metabolic rate of drugs, and changes in body temperature may affect the absorption and distribution of certain drugs. Further, the physiological parameter monitoring component 114 may include one or more sensors such as a heart rate monitor, a blood pressure monitor, a body temperature sensor, and an activity sensor. These sensors can be integrated into wearable devices to achieve 24-hour uninterrupted monitoring.

[0047] Preferably, the heart rate monitor may be configured with a green LED and a photoelectric sensor to detect changes in blood volume using photoplethysmography (PPG) to measure heart rate, while also providing blood oxygen saturation information. Preferably, the sphygmomanometer may be configured with a micro air pump and a pressure sensor to achieve continuous non-invasive blood pressure monitoring using the oscillation principle, thereby avoiding the inconvenience of traditional cuff-type sphygmomanometers. Preferably, the body temperature sensor uses thermistor technology, which is in direct contact with the skin and measures body temperature through the principle of heat conduction, with fast response and high precision. Preferably, the activity sensor may be configured with a three-axis accelerometer and a gyroscope to accurately capture the user's motion state and posture changes, wherein these data are crucial for evaluating drug metabolism rates because exercise affects blood circulation and metabolic processes.

[0048] Preferably, the planning unit 100 may be configured with an analysis module 120 for generating a feeding plan according to the data information acquired by the acquisition module 110, wherein the analysis module 120 may include a data processing component 121, a data storage component and a data transmission component.

[0049] Preferably, the data processing component 121 may first perform pre-processing operations such as cleaning and standardization on the data information acquired by the acquisition module 110, wherein the pre-processing operations may include removing outliers, processing missing data, and unifying data from different sources into a standard format. For example, the data processing component 121 may use a moving average method to smooth the heart rate data, or use an interpolation technique to fill in missing drug concentration data points. In addition, the data processing component 121 may also perform feature extraction to extract features meaningful for subsequent analysis from the raw data, such as peak and valley values ​​of drug concentration, daily variation patterns of heart rate, etc.

[0050] Preferably, the data processing component 121 of the present invention can adopt a multi-stage data processing process when performing preprocessing to ensure data quality and consistency. Preferably, the data processing component 121 can first use the Z-score method to identify outliers and mark data points exceeding 3 standard deviations as potential anomalies; then, the Kalman filter algorithm is applied to smooth these outliers, which can reduce the impact of noise while retaining data trends. Further, for missing data, the data processing component 121 can use multiple imputation methods to create multiple possible data sets to estimate missing values, thereby considering the uncertainty of the data. Specifically, the data processing component 121 can use the Markov chain Monte Carlo (MCMC) method to generate multiple interpolation data sets, and then combine these results to obtain the final estimated value. Preferably, the data processing component 121 can use the min-max normalization method to standardize the data, so as to convert data of different scales into the [0,1] interval. This method maintains the distribution characteristics of the original data and facilitates the integrated analysis of different types of data. Preferably, during the feature extraction process, the data processing component 121 may use wavelet transform to analyze the time series data, wherein wavelet transform can provide information in both the time and frequency domains, and is particularly suitable for capturing the periodic changes of drug concentrations and physiological parameters. In addition, the data processing component 121 may also apply principal component analysis (PCA) to reduce the dimension, thereby extracting the most representative features.

[0051] The data processing component 121 of the present invention not only improves the accuracy of subsequent analysis through this complex preprocessing process, but also identifies potential data collection problems, such as sensor failure or user operation errors, thereby improving the reliability of the entire device.

[0052] Preferably, the data processing unit 121 of the present invention may be built with a pharmacokinetic model, which simulates the absorption, distribution, metabolism and excretion process of drugs in the human body based on the compartment model theory. Further, the pharmacokinetic model takes into account a variety of factors, including the physicochemical properties of the drug (such as molecular weight, fat solubility), route of administration, dosage and individual factors such as age, body weight, liver and kidney function, etc., wherein the pharmacokinetic model may use a set of differential equations to describe the transfer and elimination of drugs between different tissues, and by solving these equations by numerical methods, it is possible to predict the changes in drug concentrations in blood and other tissues over time.

[0053] Preferably, the pharmacokinetic model can adopt the physiological pharmacokinetic (PBPK) method, abstracting the human body into multiple interconnected compartments, wherein each compartment represents an organ or tissue group, such as blood, liver, kidney, etc. The pharmacokinetic model can use the following set of ordinary differential equations to describe the transfer and metabolism of drugs between compartments:

[0054]

[0055] Among them, C i is the drug concentration in compartment i, k ij is the transfer rate constant of drug from compartment i to compartment j, k ei is the elimination rate constant of the drug from compartment i, I i (t) is the input function.

[0056] Preferably, the input function I i (t) represents the rate at which the drug enters the body. It describes the process of the drug entering the body from the route of administration (such as intravenous injection, oral administration, etc.). Different routes of administration may correspond to different input functions, which can be obtained by measuring the concentration data of the drug in the body and then using the deconvolution method to derive the corresponding input function.

[0057] Preferably, the pharmacokinetic model takes into account the physicochemical properties of the drug, such as the distribution coefficient (logP), the ionization constant (pKa), etc. These parameters affect the distribution of the drug in different tissues, wherein the distribution coefficient refers to the distribution ratio of neutral drug molecules between two immiscible solvents (usually n-octanol and water), which reflects the affinity of the drug molecules between lipids and water; the ionization constant refers to the equilibrium constant of the weak electrolyte in the solution at ionization equilibrium, and for drugs with ionizable groups, the ionization constant value determines the ratio of its ionization state at physiological pH. In addition, the effect of half-life can also be considered. Preferably, some antiepileptic drugs (such as phenobarbital, sodium valproate) have nonlinear pharmacokinetic characteristics, and the relationship between drug concentration and dose is not a simple linear relationship, and a more complex model is required for prediction. Preferably, antiepileptic drugs are often used in combination with other drugs, so the pharmacokinetic model also takes into account the metabolic changes that may be caused by drug interactions. For example, some antiepileptic drugs can induce or inhibit liver enzymes, thereby affecting the metabolism of other drugs. At the same time, due to the large differences in patients' responses to anti-epileptic drugs, accurate individualized drug concentration prediction is particularly important. Therefore, the pharmacokinetic model also integrates individual factors, including liver and kidney function indicators (such as creatinine clearance), body weight, age, etc., and achieves personalized prediction by adjusting the corresponding rate constants. In particular, epilepsy patients may develop drug resistance after long-term medication, and the clearance rate may change over time, so it is necessary to dynamically adjust the relevant parameters in the model.

[0058] Preferably, the data processing component 121 can use the Runge-Kutta method with an adaptive step size when performing numerical solution using the pharmacokinetic model, which can improve the calculation efficiency while ensuring accuracy. The data processing component 121 of the present invention can not only accurately predict the drug concentration by performing numerical solution on the established pharmacokinetic model, but also provide information on the distribution of the drug in various organs, which is particularly important for evaluating the effect of the drug on a specific organ (such as the breast).

[0059] Preferably, the data processing component 121 may be integrated with a variety of machine learning algorithms to optimize the pharmacokinetic model and improve the prediction accuracy. These algorithms can learn from historical data and continuously optimize the parameters of the pharmacokinetic model. For example, a support vector machine (SVM) can be used to predict the drug clearance rate of an individual, or a random forest algorithm can be used to evaluate the degree of influence of different factors on drug metabolism. In addition, deep learning techniques such as long short-term memory networks (LSTMs) can also be used to capture complex patterns of drug concentration changes over time.

[0060] Preferably, the data processing component 121 may use a support vector machine (SVM) to predict individual drug clearance, wherein the SVM may use a radial basis function kernel that can capture nonlinear relationships, and the model uses grid search and cross-validation to optimize hyperparameters, such as C value and gamma value.

[0061] Preferably, the data processing component 121 can use a random forest algorithm to evaluate the effects of different factors on drug metabolism, wherein the algorithm can construct multiple decision trees, each tree uses bootstrap samples and random feature subsets, and identifies the most critical influencing factors by calculating feature importance scores.

[0062] Preferably, the data processing component 121 can use a long short-term memory network (LSTM) to capture the time series pattern of drug concentration, wherein the LSTM network includes multiple hidden layers, each layer has 64 neurons, and uses dropout technology to prevent overfitting. The network training can use the Adam optimizer with a learning rate of 0.001.

[0063] Preferably, the data processing component 121 can generate a final feeding time recommendation after considering the prediction results of the pharmacokinetic model, the output of the machine learning algorithm and the pre-set safety threshold, wherein the data processing component 121 can use fuzzy logic to handle uncertainty when generating the feeding time recommendation. For example, it can classify the feeding time into three levels of "safe", "possibly safe" and "unsafe" according to the predicted drug concentration level. Preferably, the data processing component 121 can give priority to designating time periods at the "safe" level as feeding times. Furthermore, the data processing component 121 can also consider other practical factors when generating feeding time recommendations, such as the hunger state of infants and young children and the mother's work and rest schedule, so as to generate a feeding plan that is both safe and practical.

[0064] Preferably, the feeding plan generated by the data processing component 121 can be adjusted in real time according to the actual feeding situation. For example, when an infant or young child enters a hungry state during an unplanned feeding time, causing the user to perform an unplanned feeding operation, the user can collect the actual situation of this feeding operation through the input component 111 and send it to the data processing component 121, so that it can update the feeding plan.

[0065] Preferably, the data processing component 121 can store the calculated data in the data storage component, wherein the data storage component can use a high-speed flash memory chip for local data storage and use an AES-256-bit encryption algorithm to protect sensitive data. Preferably, the data transmission component of the analysis module 120 can realize data interaction with other functional units through a communication connection, wherein the data transmission component can use wired and / or wireless methods for data transmission, for example, low-power Bluetooth 5.0 technology and / or Wi-Fi 6 (802.11ax) standards can be used to ensure high-speed and stable data transmission.

[0066] Preferably, if Figure 1 As shown, a reminder module 130 may be provided in the planning unit 100 to remind the user to perform the feeding operation when any feeding time starting node of the feeding plan is reached, wherein the reminder module 130 may adaptively adjust the reminder mode according to the user's daily habits and physiological rhythms. For example, the reminder module 130 may learn which reminder mode (sound, vibration or visual, etc.) the user is more likely to respond to, and use that mode when appropriate. Furthermore, the reminder module 130 may also be integrated with contextual awareness technology, and select the most appropriate reminder mode by analyzing various data information such as ambient noise, light, and user activity status that can be obtained through the acquisition module 110. For example, a gentle vibration reminder may be selected when the user is asleep, and a sound reminder may be selected when the user is active.

[0067] Preferably, if Figure 3 As shown, the feeding guidance device may be provided with a state monitoring unit 200 for obtaining the state of the infant in real time, wherein when the state of the infant is abnormal, the state monitoring unit 200 may directly send an alarm signal or send an alarm signal to the planning unit 100, and inform the user by using the reminder module 130 built into the planning unit 100. Further, the state monitoring unit 200 may be communicatively connected with the data transmission component of the planning unit 100.

[0068] Preferably, if Figure 3 As shown, the state monitoring unit 200 may be configured with various sensor hardware such as a microphone 210 for acquiring infant sound data (especially crying), a camera 220 for acquiring infant behavior data, and an environmental sensor for acquiring infant environment data, wherein the sound data may include crying frequency, duration and type, etc.; the behavior data may include activity level (amount of exercise), sleep pattern (sleep time, sleep stage), posture and body position changes, etc.; the environmental data may include temperature, humidity, light intensity and environmental noise level, etc. Preferably, as Figure 4As shown, the state monitoring unit 200 can judge the state of the infant based on the various types of sensor data obtained to confirm whether the state of the infant is abnormal, wherein the abnormal state of the infant may include a first abnormality caused by hunger factors and a second abnormality caused by other factors. Further, when the first abnormality occurs and the current time is an unplanned feeding time, the user usually needs to perform an unplanned feeding operation. At this time, the analysis module 120 can export the data stored in the data storage component to the data processing component 121 to determine the level of the current time, wherein if it is at the "safe" level, direct feeding can be performed; if it is at the "unsafe" level, direct feeding cannot be performed; if it is at the "possibly safe" level, the user can decide the feeding method.

[0069] Preferably, under normal circumstances, breastfeeding can be carried out by direct feeding and indirect feeding, wherein indirect feeding refers to the method of first using manual milking or a breast pump to extract breast milk in advance and reasonably store it, and then feeding it to infants and young children. Therefore, the user can first extract breast milk in advance in the above manner and reasonably store it during the feeding time of the feeding plan, so that in some special circumstances, the indirect feeding method can be used to provide breast milk to infants and young children, wherein the above special circumstances may be the situation where an unplanned feeding operation needs to be performed, but direct feeding cannot be performed at this time, or the situation where it is inconvenient for the user to perform direct feeding, or even the situation where the user wants to perform indirect feeding, etc.

[0070] Preferably, considering the particularity of the user using the feeding guidance device, the feeding guidance device of the present invention can be built with a breast pump or can be used in conjunction with any breast pump on the market to avoid the situation where breastfeeding is needed but direct feeding is not possible or inconvenient, resulting in the infant being unable to eat breast milk in time. When the user does not use a breast pump, manual milking can be used instead. In particular, when the user feels breast fullness and discomfort, at least part of the breast milk can be taken out by manual milking or a breast pump according to the level at the current time. If the current time is at an "unsafe" level, light milking can be performed to relieve discomfort without completely emptying the breast. If the breast milk taken out does not meet the standard, it can be discarded.

[0071] Preferably, if Figure 1 and Figure 3As shown, the feeding guidance device of the present invention may also include an analysis test paper 140, which can absorb a trace amount of breast milk sample, so that a portion of the breast milk sample is extracted on the analysis test paper 140 during direct feeding or indirect feeding, and is inserted into a slot provided on the planning unit 100 (i.e., a wearable device) to reach the detection area of ​​the drug detection component 113 of the acquisition module 110, so that the color change on the analysis test paper 140 can be read by the optical sensor of the drug detection component 113 and sent to the data processing component 121, so that the color change can be analyzed by the data processing component 121 to calculate the drug concentration.

[0072] Preferably, the data processing component 121 can export the data information related to the current time stored in the data storage component to compare the difference between the drug concentration predicted by the model and the drug concentration obtained by analysis and calculation, so as to determine the accuracy of the model. Further, when the accuracy of the model is low, the data processing component 121 can train the established model. Preferably, the analysis test paper 140 can be used at intervals, wherein the interval period can be flexibly adjusted according to the accuracy of the model, and when the interval period passes, the reminder module 130 can remind the user.

[0073] Figure 5 This is a schematic diagram of a use scenario of a feeding guidance device according to a preferred embodiment of the present invention, wherein the appearance of the status monitoring unit 200 is only an example, and it can be configured in any shape.

[0074] Example 2

[0075] This embodiment is a further improvement on Embodiment 1, and the repeated contents will not be repeated here.

[0076] The present invention provides a postpartum infant feeding guidance system, and in particular a postpartum infant feeding guidance system for epileptic patients, which comprises the feeding guidance device as described in Example 1. Preferably, Figure 6 As shown, the feeding guidance system may also include: a server 300, which is used to communicate with the feeding guidance devices used by different users to achieve data interaction and complete database and software updates. Further, the updated database may include drug types, especially when new drugs are on the market; the updated software may include models and algorithms, etc.

[0077] Preferably, the feeding guidance system may further include a medical terminal operated by medical personnel, so as to provide professional advice and guidance to each user by logging into the server 300 .

[0078] It should be noted that the above-mentioned specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention specification and its drawings are illustrative and do not constitute a limitation of the claims. The scope of protection of the present invention is defined by the claims and their equivalents. The present invention specification contains multiple inventive concepts, such as "preferably" or "according to a preferred embodiment", which means that the corresponding paragraph discloses an independent concept, and the applicant reserves the right to file a divisional application based on each inventive concept. Throughout the text, the features guided by "preferably" are only an optional method and should not be understood as a must-have setting. Therefore, the applicant reserves the right to abandon or delete the relevant preferred features at any time.

Claims

1. A feeding guidance device for postpartum infants and young children of epileptic patients, It is characterized in that The planning unit (100) configured in the feeding guidance device is used to formulate a feeding plan for a user who is an epileptic patient, and comprises a collection module (110) for obtaining the user's medication status and physical information and an analysis module (120) for generating a feeding plan based on the collected data, wherein: The analysis module (120) can analyze the metabolic process of the drug in the user's body based on the characteristics of the drug taken by the user and the user's physical information, so as to plan the feeding time and generate an optimal feeding plan. The analysis module (120) can classify the feeding time according to the predicted drug concentration level, and can distinguish between breastfeeding time and fasting time according to the feeding time level, thereby forming a feeding plan.

2. The feeding guidance device according to claim 1, It is characterized in that The acquisition module (110) includes one or more of an input component (111) capable of acquiring data information input by a user based on human-computer interaction technology, an input component (112) capable of acquiring information about medications taken by a user by means of image acquisition, a medication detection component (113) capable of detecting the drug concentration in the user's body in real time, and a physiological parameter monitoring component (114) for collecting various physiological data of the user, wherein the acquisition module (110) is communicatively connected to the analysis module (120).

3. The feeding guidance device according to claim 1 or 2, It is characterized in that The drug detection component (113) is configured with a microneedle array integrated with a microfluidic channel to guide the collected body fluid sample to the analysis chip, thereby realizing painless collection of body fluid samples through the skin for analysis, and / or is configured with a non-invasive optical sensor to penetrate the skin through near-infrared spectroscopy analysis technology to detect the concentration of specific drugs, thereby continuously monitoring the drug level in the user's body.

4. The feeding guidance device according to any one of claims 1 to 3, It is characterized in that The data processing component (121) configured in the analysis module (120) is capable of performing preprocessing operations and feature extraction on the data information acquired by the acquisition module (110), so that the data in a unified standard format can be input into a pharmacokinetic model built into the data processing component (121), so as to simulate the absorption, distribution, metabolism and excretion process of the drug in the human body through numerical solution, thereby predicting the change in drug concentration and information on the distribution of the drug in various organs.

5. The feeding guidance device according to any one of claims 1 to 4, It is characterized in that The state monitoring unit (200) configured in the feeding guidance device is configured with a microphone (210) for acquiring infant sound data, a camera (220) for acquiring infant behavior data, and / or an environmental sensor (230) for acquiring infant environment data, wherein the sound data includes crying frequency, duration and type; the behavior data includes activity level, sleeping pattern, posture and body position changes; and the environmental data includes temperature, humidity, light intensity and environmental noise level.

6. The feeding guidance device according to any one of claims 1 to 5, It is characterized in that The state monitoring unit (200) can judge the state of the infant based on the various types of sensor data obtained to confirm whether the state of the infant is abnormal. When the state of the infant is abnormal, the state monitoring unit (200) can directly send an alarm signal or send the alarm signal to the planning unit (100) and use the reminder module (130) built into the planning unit (100) to inform the user, wherein the abnormal state of the infant may include a first abnormality caused by hunger and a second abnormality caused by other factors.

7. The feeding guidance device according to any one of claims 1 to 6, It is characterized in that The reminder module (130) can remind the user to perform a feeding operation when any feeding time starting node of the feeding plan is reached, wherein the reminder module (130) can adaptively adjust the reminder mode according to the user's daily habits and physiological rhythm based on various types of data information obtained by the acquisition module (110), and the reminder mode includes sound, vibration or vision.

8. The feeding guidance device according to any one of claims 1 to 7, It is characterized in that The feeding guidance device is provided with an analysis test paper (140) for being connected to the planning unit (100) in a pluggable manner. The analysis test paper (140) can absorb part of the breast milk sample when the user performs a feeding operation and be inserted into a slot provided on the planning unit (100) to reach the detection area of ​​the drug detection component (113) of the collection module (110), so that the color change on the analysis test paper (140) is read by the drug detection component (113) and sent to the data processing component (121), so that the color change is analyzed by the data processing component (121) to calculate the drug concentration.

9. The feeding guidance device according to any one of claims 1 to 8, It is characterized in that The data processing component (121) generates a final feeding time recommendation after considering the prediction results of the pharmacokinetic model, the output of the machine learning algorithm and a pre-set safety threshold, wherein the data processing component (121) can compare the difference between the drug concentration predicted by the model and the drug concentration obtained through analysis and calculation, thereby determining the accuracy of the model, and can train the model below the accuracy threshold.

10. A feeding guidance system for postpartum infants and young children of epileptic patients, It is characterized in that It comprises one or more feeding guidance devices according to any one of claims 1 to 9 and a server (300) communicatively connected to the feeding guidance device, wherein the server (300) is used to implement data interaction with the feeding guidance device and complete database and software updates.

Images