CN106108846A - A kind of intelligent drug risk monitoring method and system - Google Patents

Abstract

The embodiment of the present invention discloses an intelligent drug risk monitoring method and system, wherein the method includes: collecting the user's physical sign data through a wearable device; scanning the medicine taken when taking medicine to obtain drug data; The data is sampled at preset time intervals; the side effect probability of the drug is obtained based on the neural network model. In the embodiment of the present invention, according to the changes in the physical signs of heart disease patients and the ingredients of the medicines taken, combined with the neural network, the side effects of the user after taking the medicine are intelligently analyzed, and the risk of taking the medicine is conveniently and quickly evaluated. A large number of people's medication data are used for training, fully taking care of individual differences, quickly judging the patient's physical condition, predicting and timely monitoring the user's medication safety.

Description

An intelligent drug risk monitoring method and system

technical field

The invention relates to the field of home care technology, in particular to an intelligent drug risk monitoring method and system.

Background technique

Any medicine has side effects. On the one hand, medicine has a therapeutic effect, and on the other hand, it has side effects. There are many types of drugs that have side effects on heart patients, and some drugs are not very stable. Therefore, heart patients must follow the doctor's advice and do not use them without authorization. However, some patients do not strictly follow the doctor's advice when taking medicine, or do not return for consultation or check-up when their condition changes, and take one or more medicines for a long time, resulting in toxic side effects and even aggravating the original heart disease. Therefore, the detection of side effects caused by taking drugs has become particularly important. Today, wearable health devices are developing rapidly and can be used to monitor the physiological indicators of the human body for a long time.

In terms of wearable devices for detecting signs related to heart disease, there are currently commercialized products. For example, the cardiac monitor provided by CardioNet Company of the United States can collect the patient's electrocardiogram data in real time and send it back to the company for analysis. Isansys wireless heart rate sensor, wireless pulse oximeter and wireless blood pressure monitor.

In terms of drug side effect prediction, the existing "clinical drug decision system" refers to databases, such as adverse drug reaction databases, formula databases, and adverse drug reaction databases, to participate in prescription decisions during the prescription process, helping doctors to give Reasonable formula.

Combine wearable devices with medical care, use wearable devices to record human body signs for statistical analysis, and give early warnings for abnormal values.

At present, wearable devices only participate in the detection of heart disease signs, and most monitoring systems only have simple data collection and recording functions. CardioNet's equipment sends information back to the monitoring center to manually judge the patient's physical condition, but it still has the disadvantage of poor timeliness. In addition, for the judgment and prediction of abnormal physical signs, most of them use simple statistical data, such as blood pressure is less than a certain value, the standard is rigid, and individual differences are not well taken care of.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides an intelligent drug risk monitoring method and system, which can train a large number of people's medication data through a neural network, fully take care of individual differences, and quickly judge the patient's risk. Physical condition, to achieve drug monitoring.

In order to solve the above problems, the present invention proposes an intelligent drug risk monitoring method, the method comprising:

Collect user's vital signs data through wearable devices;

Scan the medicines taken when taking medicines to obtain medicine data;

Sampling of sign data at preset time intervals;

Obtain the side effect probability of the drug based on the neural network model.

Preferably, after the step of obtaining the side effect probability of the medicine based on the neural network model, it also includes:

Alert the user based on the probability of side effects.

Preferably, the method also includes:

Set a risk threshold, and when the probability of side effects exceeds the risk threshold, a reminder will be sent to the user.

Correspondingly, the present invention also provides an intelligent drug risk monitoring system, said system comprising:

Wearable devices, used to collect user's vital signs data;

The scanning module is used to scan the medicine taken when taking medicine to obtain medicine data;

A sampling module, configured to sample the sign data at preset time intervals;

The calculation module is used to obtain the side effect probability of the medicine based on the neural network model.

Preferably, the system further includes: a reminder module, configured to send a reminder to the user according to the probability of side effects.

Preferably, the system further includes: a setting module, configured to set a risk threshold; when the probability of side effects exceeds the risk threshold, the reminder module sends a reminder to the user.

In the embodiment of the present invention, according to the changes in the physical signs of heart disease patients and the ingredients of the medicines taken, combined with the neural network, the side effects of the user after taking the medicine are intelligently analyzed, and the risk of medication is conveniently and quickly evaluated. A large number of people's medication data are used for training, fully taking care of individual differences, quickly judging the patient's physical condition, predicting and timely monitoring the user's medication safety.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is a schematic flow chart of an intelligent drug risk monitoring method according to an embodiment of the present invention;

Fig. 2 is a schematic diagram of the structure and composition of an intelligent drug risk monitoring system according to an embodiment of the present invention.

detailed description

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of an intelligent drug risk monitoring method according to an embodiment of the present invention. As shown in Fig. 1, the method includes:

S1, collect the user's physical signs data through the wearable device;

S2, scan the medicine taken when taking the medicine, and obtain the medicine data;

S3, sampling the sign data at preset time intervals;

S4, obtaining the side effect probability of the drug based on the neural network model.

In the embodiment of the present invention, the user's vital signs data can be collected through the following wearable devices.

Table 1 List of wearable devices corresponding to detection signs

a.OFweek wearable devices:

The InBody Band provides users with an intelligent way to manage their physical health by measuring body composition and allowing users to easily track their weight, fitness activity and overall health. In addition to the "standard" functions of sports trackers such as motion tracking, heart rate monitoring, sleep pattern analysis, and vibration reminders, it can analyze the body's muscle and fat content, and can also monitor the calories consumed and consumed by the user. It can measure the user's health according to the user's heart rate. Fatigue, which measures the user's heart rate using an electrocardiogram.

b. HealthPals products:

A wearable health monitor powered by the body's own energy. Health monitoring consists of a bracelet and a ring equipped with sensors, temperature, humidity, etc. to measure arterial blood oxygen saturation blood pressure sensors, sensors for white-collar electrocardiogram, sensors for heart monitoring, accelerator and respiration, monitoring with sensors signal It consists of an EEG headset and an EKG sensor for more accurate ECG data collection. They monitor body temperature, blood pressure, brain waves, heart rate, etc. respectively. It is mainly used in patients with cardiovascular and cerebrovascular diseases, sleep disorders, hypertension, epilepsy and stroke.

c. Scanadu Scout:

A device capable of autonomously detecting vital signs can measure blood pressure, heart rate, body temperature and body temperature just by placing it on the forehead for a few seconds. It is also possible to directly use existing sign detection equipment such as smart wristbands and other sign detection equipment that can detect heart rate, body temperature, perspiration, etc.

d. Wearable pulse oximeter

The wearable pulse oximeter continuously monitors the user's blood oxygen saturation and pulse rate. It is equipped with a fingertip sensor and is connected to the wrist device, which can monitor the user's blood oxygen saturation during normal daily activities or at night. The device is useful for detecting symptoms such as obstructive sleep apnea syndrome, chronic obstructive pulmonary disease and sleep apnea.

e. Respiration monitoring smartphone earphones

It can be converted into a value according to the speed, depth, coherence, full of panting or clear breathing sound in the sample, that is, "breathing index". The MyBreath mobile app, which tracks and measures the breathing of users using regular headphones, is also part of Aetna's CarePass mobile health platform. With such a new headset, combined with the MyBreath mobile application, it is possible to fully monitor the user's breathing.

The corresponding signs of drug side effects are shown in Table 2.

Table 2 Corresponding signs of drug side effects

At present, the diagnosis of heart disease is mainly by analyzing the electrocardiogram (ECG) to judge the physiological function of the heart. The electrocardiograph and the Holter dynamic electrocardiograph are commonly used clinically to record the ECG.

After scanning the QR code, get the drug information, and then match the drug library to find out what side effects the drug will cause, and then constrain the detected side effect results to make the results more accurate.

Some drug characteristics are shown in Table 3.

Table 3 Partial list of drug characteristics

Further, after S4, it also includes: sending a reminder to the user according to the probability of side effects.

Further, it also includes:

Set a risk threshold, and when the probability of side effects exceeds the risk threshold, a reminder will be sent to the user.

In the method embodiment of the present invention, the user wears a wearable device, and the wearable device continuously collects the user's vital signs data, and transmits the data to the device (such as a mobile phone) where the system is located through wireless transmission; Personal data (such as age, height, weight, gender, etc.); when taking medication, enter the data of the medication taken into the system by scanning the QR code. Sampling at a certain time interval (such as 10 minutes), and the user data of the previous step are used as data input to the system; the system uses the trained neural network model to fit the current or imminent side effect probability according to the data, and reminds the user.

The neural network model requires a large amount of data for training. The single training data includes 24 hours before the medication, the drugs used, the patient's physical symptoms collected within 72 hours after the medication, and the patient's personal data. After the neural network inputs the data, it fits the current or future side effects and corresponding probabilities in the data as the output.

The wearable device always records the user's signs related to heart disease (including: the patient's electrocardiogram sign value, heart rate, pulse, blood pressure, blood oxygen level, body temperature, weight, blood oxygen saturation, fatigue, breathing), when the patient needs medication At the same time, by scanning the QR code, the drug information (drug name, possible side effects of the drug) is detected and obtained, and the patient's signs are continuously detected and analyzed through the neural network to determine whether side effects occur. When the side effects occur, through the disease The probability and importance of the onset, using the idf method, calculates the score of the disease. The higher the score, the more worthy of attention. Through the threshold of the pre-equipment, when the score exceeds the threshold, the user is reminded that the side effect may occur.

The input format is triplet (A, L, D) data, A represents the attributes of the patient, such as age, gender, height, weight, L represents signs related to heart disease, L is obtained through wearable device detection, Record the patient's physical signs the day before taking the medicine as L1, and the corresponding physical signs at the current moment after taking the medicine as L2, then L=(L1,L2), and D represents the medicine taken by the patient.

Output vector C=(c1,c2,...,ck), ci in the base means the probability of getting disease i is ci, and c1 means the probability of not getting sick, so the neural network calculates the probability of P(C|A,L,D).

The neural network is divided into three layers. The first layer is the input layer, the second layer is the hidden layer, and the third layer is the output layer. The hidden layer uses the hyperbolic tangent function tanh as the activation function.

The neural network workflow is as follows:

The triplet (A, L, D) is input as a vector, let X=(A, L, D), calculate tanh(d+HX) in the hidden layer, calculate b+Utanh(d+HX) in the output layer, and pass The drug obtained by the QR code is D, and the side effect corresponding to D is (d1, d2,...dm), m≤k, d1=c1, and m values are extracted from C to form a new vector, because the m values For the desired result, normalization processing is required so that the probability sum is 1. Therefore, the softmax function is applied to the m values, and the output layer is normalized, as follows:

Note y _i =b+Utanh(d+HX), then

After the calculation is completed, the probability pi of each side effect is obtained, and the importance degree wi of the side effect is obtained by querying the drug database, and the score fi=pi*wi is calculated. If fi exceeds the threshold of the pre-device, the user is reminded that the side effect may occur .

Correspondingly, the embodiment of the present invention also provides an intelligent drug risk monitoring system, as shown in Figure 2, the system includes:

The wearable device 1 is used to collect the physical sign data of the user;

The scanning module 2 is used to scan the medicine taken when taking medicine to obtain medicine data;

The sampling module 3 is used to sample the sign data at preset time intervals;

The calculation module 4 is used to obtain the side effect probability of the medicine based on the neural network model.

Further, the system also includes: a reminder module, configured to send a reminder to the user according to the probability of side effects.

The setting module is used to set the risk threshold; when the probability of side effects exceeds the risk threshold, the reminder module sends a reminder to the user.

For the functions of each functional module in the system embodiment of the present invention, refer to the flow processing in the method embodiment of the present invention, and will not be repeated here.

In the embodiment of the present invention, according to the changes in the physical signs of heart disease patients and the ingredients of the medicines taken, combined with the neural network, the side effects of the user after taking the medicine are intelligently analyzed, and the risk of medication is conveniently and quickly evaluated. A large number of people's medication data are used for training, fully taking care of individual differences, quickly judging the patient's physical condition, predicting and timely monitoring the user's medication safety.

Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above-mentioned embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium can include: Read Only Memory (ROM, Read Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk, etc.

In addition, the intelligent drug risk monitoring method and system provided by the embodiments of the present invention are described above in detail. In this paper, specific examples are used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only used to help Understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation and scope of application. In summary, the content of this specification is not It should be understood as a limitation of the present invention.

Claims (6)

1. An intelligent drug risk monitoring method, characterized in that the method comprises: Collect user's vital signs data through wearable devices; Scan the medicines taken when taking medicines to obtain medicine data; Sampling of sign data at preset time intervals; Obtain the side effect probability of the drug based on the neural network model. 2. The intelligent drug risk monitoring method according to claim 1, further comprising: Alert the user based on the probability of side effects. 3. The intelligent medicine risk monitoring method as claimed in claim 2, is characterized in that, described method also comprises: Set a risk threshold, and when the probability of side effects exceeds the risk threshold, a reminder will be sent to the user. 4. An intelligent drug risk monitoring system, characterized in that the system comprises: Wearable devices, used to collect user's vital signs data; The scanning module is used to scan the medicine taken when taking medicine to obtain medicine data; A sampling module, configured to sample the sign data at preset time intervals; The calculation module is used to obtain the side effect probability of the medicine based on the neural network model. 5. The intelligent drug risk monitoring system according to claim 4, further comprising: a reminder module, configured to send reminders to users according to the probability of side effects. 6. The intelligent drug risk monitoring system according to claim 5, characterized in that, the system further comprises: a setting module for setting a risk threshold; when the probability of side effects exceeds the risk threshold, the reminder module sends a message to the user remind.