Early cardiovascular and cerebrovascular disease detection method based on sleep data

Abstract

The invention relates to the technical field of disease detection, in particular to an early cardiovascular and cerebrovascular disease detection method based on sleep data. The method comprises the following steps: acquiring sensor data in a night sleep process of a user, and preprocessing the data to obtain intermediate features; constructing a data set based on historical data and establishing a probability distribution hypothesis, and performing parameter learning on a probability density function by using methods such as maximum likelihood estimation and the like to form a training model; extracting intermediate features from newly collected sleep data, inputting the intermediate features into the model, calculating the probability of occurrence of the intermediate features in the distribution, and judging whether the user has the risk of cardiovascular and cerebrovascular diseases according to the probability; meanwhile, the new features are stored in an individual database, and an individualized probability distribution function is constructed; and integrating the existing model and the individualized model, calculating the risk probability of the user, and outputting early warning information when the risk probability meets a preset condition. According to the invention, early risk prompting of cardiovascular and cerebrovascular diseases can be realized by using night sleep data.

Description

TECHNICAL FIELD

[0001] The present application relates to a disease detection method, more particularly, to an early cardiovascular and cerebrovascular disease detection method based on sleep data. BACKGROUND

[0002] The formation of cardiovascular and cerebrovascular diseases is not formed overnight, but is a long-term cumulative process, but its onset is often extremely sudden, with strong concealment. For patients, the disease before the outbreak of the lesion period is the golden stage of preventing disease deterioration.

[0003] But at present in the diagnosis link, many problems need to be solved. Cardiovascular and cerebrovascular diseases are gradually formed in silence, and patients themselves are almost difficult to detect signs of illness, and when the disease breaks out, the best treatment opportunity has often been missed. The existing clinical diagnosis means has poor accuracy in judging the stability of the plaque. Taking the common carotid artery plaque ultrasonic examination as an example, it can only roughly judge the degree of vascular stenosis, and it is difficult to provide accurate basis for the stability of the plaque. SUMMARY

[0004] In view of the deficiencies of the prior art, the purpose of the present application is to provide a disease detection method which can effectively prompt the user of the risk of cardiovascular and cerebrovascular diseases.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] An early cardiovascular and cerebrovascular disease detection method based on sleep data, comprising:

[0007] Step 1, collecting sensor data of the user during night sleep, and pre-processing the data to obtain intermediate features;

[0008] Step 2, constructing a historical data set based on the intermediate features, and establishing a probability distribution hypothesis of the data according to the historical data set to obtain a probability density function form;

[0009] Step 3, using the historical data set to learn the parameters of the probability density function to obtain a trained probability density function model;

[0010] Step 4, processing the newly collected sleep data of the user to obtain new intermediate features;

[0011] Step 5, inputting the new intermediate features into the probability density function model, calculating the occurrence probability thereof in the model, and judging whether the user has a risk of cardiovascular and cerebrovascular diseases according to the occurrence probability;

[0012] Step 6, storing the new intermediate features to an individual database corresponding to the user, and establishing an individualized probability distribution function based on the individual database.

[0013] Step 7, the risk probability of the user suffering from cardiovascular and cerebrovascular diseases is calculated by integrating the probability density function model and the individualized probability distribution function, and early warning information is output when the risk probability meets a preset condition.

[0014] Further, in step 1, the measurement device collects the body change basic data given by the sensor every night, and calculates the Yuragi index of vascular strain and the PWV value according to the given data, and stores the calculated data as historical data to be arranged into a data set.

[0015] Further, in step 3, the joint probability density is calculated according to the probability density function, and the parameters of the probability density function are calculated by maximum likelihood estimation.

[0016] Further, in step 4, the user collects the night sleep data during the night sleep process using the capacitive sensor, and processes the collected data into the same intermediate features as in step 1.

[0017] Further, in step 5, the distribution of the heart rate and respiratory rate data is determined by the parameters calculated in step 3; and the probability of the occurrence of the data in the distribution is calculated according to the newly appeared heart rate data and respiratory rate data.

[0018] Further, in step 6, the probability density function conforming to the user's own database is calculated in the same way as in step 3 according to the individual database.

[0019] Further, in step 7, the probabilities of the occurrence of the data are calculated according to the probability density functions obtained in steps 3 and 6, respectively; two probability sums are integrated, and the smaller value of the two probabilities is taken, and when the smaller value is smaller than a threshold value, it is considered that the user has a risk of cardiovascular and cerebrovascular diseases, and the user is warned.

[0020] By adopting the above technical solution, the present application has the following advantages:

[0021] 1. Early risk discovery: by analyzing the user's night sleep data, potential risks can be discovered when cardiovascular and cerebrovascular diseases are still in the early stage, which can significantly advance the intervention opportunity and improve the effectiveness of prevention and treatment compared with the traditional diagnosis method after the occurrence of symptoms.

[0022] 2. Non-invasive collection: the present application uses a capacitive microphone sensor to collect heart rate, respiratory rate and other data during sleep, without the need to wear complex equipment or perform traumatic detection, and the use experience is friendly and convenient for long-term monitoring.

[0023] 3. Personalized modeling: By establishing a unique database for each user and generating a personalized probability distribution function, the method of this invention can adaptively adjust the risk assessment criteria, avoiding misjudgments or omissions caused by physiological differences between different individuals, and improving diagnostic accuracy.

[0024] 4. Dynamic update mechanism: As users' sleep data continues to accumulate, this invention can continuously update and optimize the individual database and probability distribution model, making the risk assessment results more accurate and suitable for long-term health management. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the structure of an embodiment of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Early detection methods for cardiovascular and cerebrovascular diseases based on sleep data include:

[0028] Step 1: Collect sensor data during the user's sleep at night and preprocess the data to obtain intermediate features;

[0029] Step 1.1: Collect basic data on changes in human body state from a capacitive microphone sensor every night using a measuring device. At this time, the data is continuous analog data.

[0030] Step 1.2: Calculate indicators related to cardiovascular and cerebrovascular diseases using basic human body change data provided by sensors—process the basic human body change data obtained in Step 1.1 to calculate the user's vascular strain Yuragi index and PWV value (vascular structure index).

[0031] Step 1.3: After organizing the data collected by the sensors, store it in the database as historical data;

[0032] Step 1.4: Repeat steps 1.1 and 1.2 until the required amount of data is collected.

[0033] Step 2: Construct a historical dataset based on the intermediate features, and establish a probability distribution hypothesis for the data based on the historical dataset to obtain the probability density function form;

[0034] The collected feature data is analyzed, and the probability density function of the feature data is assumed according to the distribution of the collected sample data; that is, the form of the probability density function f(x) of the historical data obtained in step 1 is assumed.

[0035] Step 3, parameter learning of the probability density function is performed using the historical data set, and a trained probability density function model is obtained.

[0036] Step 3.1, according to the historical data in step 1 and the probability density function f(x) assumption in step 2, the joint probability density is calculated; that is, according to the historical data and the assumed probability density function, the joint probability density function is calculated:

[0037]

[0038] Where n is the number of historical data.

[0039] Step 3.2, according to the joint probability density function, the parameters of the probability density function f(x) are calculated by maximum likelihood estimation; that is, the joint probability density in step 3.1 is maximum likelihood estimated: The parameters w of the probability density function are calculated.

[0040] Step 4, the newly collected user sleep data is processed to obtain new intermediate features;

[0041] Step 4.1, during the user's night sleep, a capacitive microphone sensor is used to collect night sleep data;

[0042] Step 4.2, the collected data is processed into the same intermediate features as in step 1; that is, according to the collected user night sleep data, the user's blood vessel strain Yuragi index and PWV value (blood vessel structure index) are calculated.

[0043] Step 5, the new intermediate features are input into the probability density function model, the occurrence probability of the new intermediate features in the model is calculated, and whether the user has a risk of cardiovascular and cerebrovascular diseases is judged according to the occurrence probability.

[0044] Step 5.1, the parameters of the probability density function calculated in step 3 are substituted into the assumed probability density function to determine the distribution function f(x, w) of the respiratory rate and heart rate data.

[0045] Step 5.2, for the newly appeared data , the probability P of the new data appearing in the original data distribution is calculated:

[0046]

[0047] Step 5.3: Use the calculated probability to determine whether the newly appearing data is within the normal range of the original data distribution. If it is within the normal range, no reminder is given and step 5 is skipped. If it is not within the normal range, a reminder is given. That is, determine whether to give a reminder about the human body condition based on the probability. When an event with a very low probability occurs, a warning is given to the user.

[0048] Step 6: Store the new intermediate features in the individual database corresponding to the user, and establish an individualized probability distribution function based on the individual database;

[0049] Save the user's intermediate feature data to a separate database. Create a separate database corresponding to this user and store the calculated vascular strain Yuragi index and PWV value (vascular structure index) in this database.

[0050] That is, based on the created separate database, the probability density function that conforms to the user's own database is calculated using the same method as in step 3. .

[0051] Step 7: Combine the probability density function model and the individualized probability distribution function to calculate the user's risk probability of developing cardiovascular and cerebrovascular diseases, and output early warning information when the risk probability meets the preset conditions.

[0052] Step 7.1: Collect the user's nighttime sleep data using a condenser microphone;

[0053] Step 7.2: Preprocess the data to calculate intermediate feature data, namely the calculated vascular strain Yuragi index and PWV value (vascular structure index).

[0054] Step 7.3, based on the probability density function obtained in steps 3 and 6. and Calculate the probability of the data appearing respectively. and ;

[0055] Step 7.4, combine the two probabilities and We take the smaller of the two probabilities, P. When the value of P is less than the threshold, we consider that the user has a risk of cardiovascular and cerebrovascular diseases and issue a warning to the user.

[0056] This invention can provide early diagnosis and warning of cardiovascular and cerebrovascular diseases based on the user's sleep data, enabling users to detect the risk of cardiovascular and cerebrovascular diseases at an early stage.

[0057] In the early warning process, this invention uses a method of modeling based on the user's historical data, which enables the method to adaptively adjust the judgment parameters according to different users.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early detection of cardio-cerebrovascular diseases based on sleep data, characterized in that, The method comprises the following steps: Step 1, collecting sensor data of a user during night sleep and pre-processing the data to obtain intermediate features; Step 2, constructing a historical data set based on the intermediate features and establishing a probability distribution hypothesis of the data according to the historical data set to obtain a probability density function form; Step 3, performing parameter learning on the probability density function using the historical data set to obtain a trained probability density function model; Step 4, processing newly collected user sleep data to obtain new intermediate features; Step 5, inputting the new intermediate features into the probability density function model, calculating the occurrence probability of the new intermediate features in the model, and judging whether the user has a risk of cardiovascular and cerebrovascular diseases according to the occurrence probability; Step 6, storing the new intermediate features in an individual database corresponding to the user, and establishing an individualized probability distribution function based on the individual database; Step 7, combining the probability density function model and the individualized probability distribution function to calculate the risk probability of the user suffering from cardiovascular and cerebrovascular diseases, and outputting a warning information when the risk probability meets a preset condition.

2. The method for early detection of cardiovascular and cerebrovascular diseases based on sleep data according to claim 1, characterized in that, In step 1, the measuring device collects the basic data of human body changes given by the sensor every night, calculates the Yuragi index and PWV value of blood vessel strain according to the given data, and stores the calculated data as historical data to be arranged into a data set. 3.The early cardiovascular and cerebrovascular disease detection method based on sleep data of claim 1, wherein, In step 3, the joint probability density is calculated according to the probability density function, and the parameters of the probability density function are calculated by maximum likelihood estimation.

4. The method for early detection of cardiovascular and cerebrovascular diseases based on sleep data according to claim 1 or 2, characterized in that, In step 4, during night sleep, the user collects night sleep data using a capacitive sensor, and processes the collected data into the same intermediate features as in step 1.

5. The method for early detection of cardiovascular and cerebrovascular diseases based on sleep data according to claim 3, characterized in that, In step 5, the distribution of the respiratory rate and heart rate data is determined by the parameters calculated in step 3; and the probability of the occurrence of the data in the distribution is calculated according to the newly appeared heart rate data and respiratory rate data.

6. The method for early detection of cardiovascular and cerebrovascular diseases based on sleep data according to claim 3, characterized in that, In step 6, the probability density function conforming to the user's own database is calculated according to the individual database in the same way as in step 3.

7. The method for early detection of cardiovascular and cerebrovascular diseases based on sleep data according to claim 6, characterized in that, In step 7, the probability density functions obtained in steps 3 and 6 are calculated, respectively, and the probabilities of the occurrence of the data are calculated. The two probability sums are combined, and the smaller value of the two probabilities is taken; when the smaller value is smaller than a threshold value, it is considered that the user has a risk of cardiovascular and cerebrovascular diseases, and a warning is given to the user.

Images