Medical intelligent monitoring system and method based on intelligent bracelet

Abstract

The invention relates to the technical field of medical monitoring, and discloses a medical intelligent monitoring system and method based on an intelligent bracelet, and the method comprises the steps: calculating a fuzzy ideal solution corresponding to M groups of initial monitoring data, calculating the deviation distance between each group of monitoring data and the fuzzy ideal solution, screening out N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, the method comprises the steps of calculating an external influence degree value based on external influence data, correcting N groups of target monitoring data based on the external influence degree value to obtain N groups of accurate monitoring data, finally generating a membership judgment matrix based on the N groups of accurate monitoring data, and inputting the N groups of accurate monitoring data and the membership judgment matrix into a pre-constructed prediction evaluation model to obtain an evaluation result. Through the multi-level data processing and optimizing method, inaccurate initial monitoring data can be effectively filtered out, intelligent correction is carried out in combination with external influence factors, and the influence of noise interference on a subsequent evaluation result is reduced.

Description

Technical Field

[0001] The present invention relates to the field of medical monitoring technology, and more specifically, to a medical intelligent monitoring system and method based on a smart bracelet. Background Art

[0002] With the popularity of smart wearable devices, they are gradually used in fields such as education, aviation and medical care. In the medical field, wearable devices can be used not only by hospital patients, but also by healthy people at home and patients with chronic diseases. The use of these devices can save the cost and time of going to the hospital. The most common device is a medical smart bracelet, which can monitor the patient's health status in real time, such as heart rate and blood oxygen level. However, errors may occur in the data during the monitoring process, and few existing technologies can effectively correct these data, affecting the detection accuracy.

[0003] For example, the patent with publication number CN104055499A provides a wearable smart bracelet and method for continuously monitoring human physiological signs. The patent adopts a unique data analysis method to analyze the collected human physiological signs, and can obtain a comprehensive and accurate human health status. The patent with publication number CN108742563A provides a health monitoring system based on a smart bracelet. The patent includes a storage module, a health assessment and analysis terminal, and the health assessment and analysis terminal performs a comprehensive health status analysis on the physiological signals collected by the physiological parameter detection device and the environmental parameters collected by the environmental parameter detection device.

[0004] Although the existing technology collects physiological data through smart bracelets for analysis, it lacks optimization and correction of the physiological data. Since the smart bracelets are easily interfered by external factors during the collection process, noise exists in the physiological data, and the specific values ​​are not accurate enough, which leads to large deviations in the subsequent analysis results.

[0005] In view of this, the present invention proposes a medical intelligent monitoring system and method based on a smart bracelet to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a medical intelligent monitoring system and method based on a smart bracelet.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The first aspect is a medical intelligent monitoring method based on a smart bracelet, including:

[0009] Obtain M groups of initial monitoring data within a preset time period, calculate the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculate the deviation distance between each group of initial monitoring data and the fuzzy ideal solution, and select N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1;

[0010] Acquire external impact data within a preset time period, calculate external impact degree values ​​based on the external impact data, and correct N groups of target monitoring data based on the external impact degree values ​​to obtain N groups of accurate monitoring data;

[0011] A membership judgment matrix is ​​generated based on N groups of precise monitoring data, and the N groups of precise monitoring data and the membership judgment matrix are input into a pre-built prediction and evaluation model to obtain an evaluation result.

[0012] Furthermore, the fuzzy ideal solution includes a positive fuzzy ideal solution and a negative fuzzy ideal solution. The method for calculating the fuzzy ideal solution corresponding to the M groups of initial monitoring data includes:

[0013] Traverse M groups of initial monitoring data, combine the corresponding element indicators in the M groups of initial monitoring data to construct S fuzzy sets, determine the optimal element indicators and the worst element indicators in the S fuzzy sets based on the preset state range values, combine the S optimal element indicators to construct a positive fuzzy ideal solution, and combine the S worst element indicators to construct a negative fuzzy ideal solution. The initial monitoring data includes S element indicators.

[0014] Furthermore, the deviation distance includes a positive deviation distance and a negative deviation distance, and the method for calculating the positive deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes:

[0015]

[0016] In the formula, is the positive deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, x ij is the jth element index in the i-th group of initial monitoring data, is the optimal element index corresponding to the positive fuzzy ideal solution.

[0017] Furthermore, the method for calculating the negative deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes:

[0018]

[0019] In the formula, is the negative deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, is the worst element index corresponding to the negative fuzzy ideal solution.

[0020] Furthermore, the method of selecting N groups of target monitoring data from M groups of initial monitoring data according to the deviation distance includes:

[0021] The real-time fit value corresponding to each set of initial monitoring data is calculated according to the deviation distance, and the absolute value of the difference between the real-time fit value and the standard fit value is calculated. The initial monitoring data whose absolute value of the difference is greater than the preset difference threshold is eliminated, so as to screen out N groups of target monitoring data.

[0022] Furthermore, the method for calculating the real-time fit value includes:

[0023]

[0024] In the formula, C i is the real-time fit value corresponding to the i-th group of initial monitoring data.

[0025] Furthermore, the method for correcting N groups of target monitoring data based on the external influence value includes:

[0026] Input the external impact value and N groups of target monitoring data into the pre-built monitoring correction model to obtain N groups of accurate monitoring data;

[0027] The construction method of the monitoring and correction model includes:

[0028] Acquire a sample data set, wherein the sample data set includes historical external impact values, historical target monitoring data, and historical precise monitoring data;

[0029] Divide the sample data set into a sample training set and a sample test set, and build a regression network;

[0030] The historical external influence values ​​and historical target monitoring data in the sample training set are used as the input data of the regression network, and the historical accurate monitoring data in the sample training set is used as the output data of the regression network. The regression network is trained to obtain an initial regression network for predicting accurate monitoring data.

[0031] The initial regression network is tested using a sample test set, and the initial regression network with an output smaller than a preset error value is used as a monitoring and correction model.

[0032] Furthermore, the method for generating a membership judgment matrix based on N groups of precise monitoring data includes:

[0033] Traverse N groups of precise monitoring data, normalize each element indicator in each group of precise monitoring data, determine the corresponding membership based on the preset state range value and the normalized element indicator, and construct a membership judgment matrix based on the membership corresponding to the N groups of precise monitoring data.

[0034] In the second aspect, a medical intelligent monitoring system based on a smart bracelet is used to implement the above-mentioned medical intelligent monitoring method based on a smart bracelet, including:

[0035] Filtering module: used to obtain M groups of initial monitoring data within a preset time period, calculate the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculate the deviation distance between each group of monitoring data and the fuzzy ideal solution, and filter out N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1;

[0036] Correction module: used to obtain external impact data within a preset time period, calculate the external impact degree value based on the external impact data, and correct N groups of target monitoring data based on the external impact degree value to obtain N groups of accurate monitoring data;

[0037] Evaluation module: Generate a membership judgment matrix based on N groups of precise monitoring data, input the N groups of precise monitoring data and the membership judgment matrix into the pre-built prediction evaluation model to obtain the evaluation results.

[0038] In a third aspect, an electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned medical intelligent monitoring method based on the smart bracelet when executing the computer program.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The present invention first calculates the fuzzy ideal solution corresponding to M groups of initial monitoring data, calculates the deviation distance between each group of monitoring data and the fuzzy ideal solution, screens out N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, calculates the external influence value based on the external influence data, corrects the N groups of target monitoring data based on the external influence value, obtains N groups of precise monitoring data, and finally generates a membership judgment matrix based on the N groups of precise monitoring data, inputs the N groups of precise monitoring data and the membership judgment matrix into a pre-built predictive evaluation model to obtain an evaluation result. The present invention can effectively filter out inaccurate initial monitoring data through the above-mentioned multi-level data processing and optimization method, and perform intelligent correction in combination with external influencing factors, thereby significantly improving the accuracy of smart bracelet data acquisition, reducing the impact of noise interference on subsequent evaluation results, and providing more accurate and reliable health assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of the medical intelligent monitoring method based on the smart bracelet in the present invention;

[0042] Figure 2 It is a structural schematic diagram of the medical intelligent monitoring system based on the smart bracelet in the present invention;

[0043] Figure 3 It is a flow chart of the method for calculating the fuzzy ideal solution in the present invention;

[0044] Figure 4It is a schematic diagram of the process of generating a membership judgment matrix in the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Example 1

[0047] See also Figure 1 As shown, this embodiment discloses a medical intelligent monitoring method based on a smart bracelet, including:

[0048] S10: acquiring M groups of initial monitoring data within a preset time period, calculating the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculating the deviation distance between each group of initial monitoring data and the fuzzy ideal solution, and selecting N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1;

[0049] In this embodiment, the preset time period can be set according to prior knowledge or expert experience. For example, the preset time period can be 10 minutes, then a set of initial monitoring data is obtained every half minute. The initial monitoring data represents the physiological characteristics of the user. For example, the initial monitoring data may include heart rate value, blood pressure value, blood oxygen saturation value, and respiratory rate value.

[0050] Fuzzy ideal solutions include positive fuzzy ideal solutions and negative fuzzy ideal solutions, such as Figure 3 As shown, the method for calculating the fuzzy ideal solution corresponding to the M groups of initial monitoring data includes:

[0051] Traverse M groups of initial monitoring data, combine the corresponding element indicators in the M groups of initial monitoring data to construct S fuzzy sets, determine the optimal element indicators and the worst element indicators in the S fuzzy sets based on the preset state range values, combine the S optimal element indicators to construct a positive fuzzy ideal solution, and combine the S worst element indicators to construct a negative fuzzy ideal solution. The initial monitoring data includes S element indicators.

[0052] It should be noted that the element index refers to a specific measurement parameter used to characterize the physiological characteristics of the user. For example, the initial monitoring data includes four element indicators, namely, heart rate value, blood pressure value, blood oxygen saturation value and respiratory rate value. The corresponding element indicators mentioned above are just the same element indicators in the M groups of initial monitoring data. For example, the M groups of initial monitoring data center frequency values ​​are combined to construct the corresponding fuzzy set. The state range value in the above represents the health state range value, for example:

[0053] The state range corresponding to the heart rate value is 60-100bpm, and the state range of the blood oxygen saturation value is 95%-100%. Therefore, the above-mentioned optimal element index represents the best value within the health state range, and the worst element index represents the worst value outside the health state range.

[0054] Exemplarily, in this embodiment, M is 5 and S is 4, then the initial monitoring data includes heart rate value, blood pressure value, blood oxygen saturation value and respiratory rate value, as shown in Table 1:

[0055] Table 1 Initial monitoring data table

[0056]

[0057] Then, the state range value corresponding to the heart rate value is 60-100bpm, the state range value corresponding to the blood pressure value is 90 / 60-120 / 80mmHg, the state range value of the blood oxygen saturation value is 95%-100%, and the state range value corresponding to the respiratory rate value is 14-20 times / minute. Then, according to the above content, the optimal element indicator corresponding to the heart rate value is 80bpm, the worst element indicator corresponding to the heart rate value is 105bpm, the optimal element indicator corresponding to the blood pressure value is 120 / 80mmHg, and the worst element indicator corresponding to the blood pressure value is 130 / 85mmHg. The same is true for the blood oxygen saturation value and the respiratory rate value. This embodiment will not go into too much detail on this. The positive fuzzy ideal solution is a set of S optimal element indicators, and the negative fuzzy ideal solution is a set of S worst element indicators.

[0058] The deviation distance includes positive deviation distance and negative deviation distance. The method for calculating the positive deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes:

[0059]

[0060] In the formula, is the positive deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, x ij is the jth element index in the i-th group of initial monitoring data, is the optimal element index corresponding to the positive fuzzy ideal solution.

[0061] Similarly, the method for calculating the negative deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes:

[0062]

[0063] In the formula, is the negative deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, is the worst element index corresponding to the negative fuzzy ideal solution.

[0064] The method for selecting N groups of target monitoring data from M groups of initial monitoring data according to the deviation distance includes:

[0065] The real-time fit value corresponding to each set of initial monitoring data is calculated according to the deviation distance, and the absolute value of the difference between the real-time fit value and the standard fit value is calculated. The initial monitoring data whose absolute value of the difference is greater than the preset difference threshold is eliminated, so as to screen out N groups of target monitoring data.

[0066] It can be understood that the fit value represents the degree of deviation of each set of initial monitoring data, and the degree of deviation here refers to the degree of deviation from the user's ideal state.

[0067] Methods for calculating real-time fit values ​​include:

[0068]

[0069] In the formula, C i is the real-time fit value corresponding to the i-th group of initial monitoring data;

[0070] It should be noted that in this embodiment, C i The larger the value, the greater the deviation from the user's ideal state. i The smaller it is, the less the degree of deviation from the user's ideal state, and the standard fit value can be 0.5, then the absolute value of the above difference represents the degree of abnormality of the initial monitoring data, because the closer the initial monitoring data is to the user's ideal state and the more it deviates from the user's ideal state, the more it indicates that the initial monitoring data is abnormal, and the purpose of this embodiment is to reduce data noise and improve the accuracy of subsequent evaluation results. Therefore, whether the initial monitoring data is close to the user's ideal state (the real-time fit value is close to 1) or deviates from the user's ideal state (the real-time fit value is close to 0), it indicates that this set of data is abnormal. Such initial monitoring data may not meet the expected standard state due to interference from various external or internal factors, and should therefore be eliminated.

[0071] In this embodiment, M groups of initial monitoring data are obtained within a preset time period to ensure the integrity of data collection. Then, the deviation distance between each group of data and the fuzzy ideal solution is calculated, and potential abnormal data is preliminarily screened out through deviation analysis. On this basis, the real-time fit value is further calculated to screen out initial monitoring data with large deviations. Through such a multi-level screening and correction process, noise data can be gradually removed to ensure that the N groups of target monitoring data finally retained are more in line with the actual situation, thereby improving the accuracy of subsequent data analysis and evaluation.

[0072] S20: acquiring external influence data within a preset time period, calculating an external influence degree value based on the external influence data, and correcting N groups of target monitoring data based on the external influence degree value to obtain N groups of accurate monitoring data;

[0073] Among them, the external impact data includes but is not limited to the wearing pressure value, skin humidity value and signal-to-noise ratio value. The wearing pressure value represents the degree of contact between the smart bracelet and the user's skin. The skin humidity value represents the degree of wetness on the user's skin surface (such as sweating). Excessive skin humidity may affect the optical and conductivity readings of the sensor, resulting in distorted or inaccurate data. The signal-to-noise ratio value represents the degree of stability of the smart bracelet during the detection process. The larger the signal-to-noise ratio value, the worse the stability of the smart bracelet during the detection process.

[0074] Methods for calculating external influence values ​​based on external influence data include:

[0075]

[0076] Where EXV is the external influence value, WEP is rt is the wearing pressure value, WEP sd is the standard pressure value, SKM is the skin moisture value, SNR is the signal-to-noise ratio value, cosh[·] is the hyperbolic cosine function, cot- 1 [·] is the inverse cotangent function, EXP(·) is the exponential function with base e, ln(·) is the logarithmic function with base e, and e is a natural constant.

[0077] In this embodiment, taking the wearing pressure value as an example, a larger or smaller wearing pressure value will cause inaccurate detection by the smart bracelet. This is because when the wearing pressure is too large, the bracelet will press the user's skin tightly, which may affect local blood flow, which will interfere with the bracelet sensor (such as heart rate, blood oxygen sensor) to detect real physiological signals. If the wearing pressure is too small, the contact between the sensor and the skin may not be close enough, resulting in the sensor being unable to stably receive the body's physiological signals, resulting in weak signals, large fluctuations, or even data loss. Therefore, it can be seen from the above content that in this embodiment, the larger the external influence value, the more inaccurate the smart bracelet detection.

[0078] The method for correcting N groups of target monitoring data based on external influence values ​​includes:

[0079] The external influence value and N groups of target monitoring data are input into the pre-built monitoring correction model to obtain N groups of accurate monitoring data.

[0080] The construction method of the monitoring and correction model includes:

[0081] Acquire a sample data set, wherein the sample data set includes historical external impact values, historical target monitoring data, and historical precise monitoring data;

[0082] Divide the sample data set into a sample training set and a sample test set, and build a regression network;

[0083] The historical external influence values ​​and historical target monitoring data in the sample training set are used as the input data of the regression network, and the historical accurate monitoring data in the sample training set is used as the output data of the regression network. The regression network is trained to obtain an initial regression network for predicting accurate monitoring data.

[0084] The initial regression network is tested using a sample test set, and the initial regression network with an output smaller than a preset error value is used as a monitoring and correction model. The initial regression network is a deep neural network model.

[0085] It is not difficult to understand that the above-mentioned historical precise monitoring data are obtained through experiments by technical personnel in this field, and the historical precise monitoring data are obtained under the ideal experimental conditions. The ideal experimental state refers to the monitoring state under the best conditions created for obtaining the historical precise monitoring data. In this state, all external influencing factors (such as wearing pressure, skin humidity, signal noise, etc.) are strictly controlled within the optimal range, thereby ensuring that the sensor can stably and accurately obtain the user's physiological signals.

[0086] In this embodiment, through the layer-by-layer correction and optimization of S10 and S20, not only can the initial monitoring data with large deviations be removed, but also the target monitoring data can be finely adjusted according to more subtle external influences. This step-by-step correction method can effectively reduce the cumulative error in the data processing process, making the final N groups of precise monitoring data more accurate, and improving the reliability of subsequent data analysis and evaluation results.

[0087] S30: generating a membership judgment matrix based on N groups of precise monitoring data, inputting the N groups of precise monitoring data and the membership judgment matrix into a pre-built prediction evaluation model to obtain an evaluation result;

[0088] like Figure 4 As shown, the method for generating a membership judgment matrix based on N groups of accurate monitoring data includes:

[0089] Traverse N groups of precise monitoring data, normalize each element indicator in each group of precise monitoring data, determine the corresponding membership based on the preset state range value and the normalized element indicator, and construct a membership judgment matrix based on the membership corresponding to the N groups of precise monitoring data.

[0090] It should be noted that the purpose of normalizing each element indicator is to eliminate the dimensional differences between different element indicators so that each element indicator can be reasonably compared and calculated on the same scale. The membership degree refers to the degree of compliance or closeness of each element indicator to the preset health state. In fuzzy mathematics, the membership degree usually takes a value between 0 and 1, which is used to indicate the degree to which an element indicator belongs to a certain state. The closer the membership value is to 1, the more the indicator complies with the preset health state; the closer the membership value is to 0, the more serious the deviation of the indicator from the health state.

[0091] Methods for determining the corresponding membership include:

[0092] The preset state range value is normalized, and the membership between the element index and the preset state range value is calculated according to the membership function.

[0093] It can be understood that the membership function refers to a mathematical function used to calculate the membership value of an element index. Common membership functions include linear membership function, triangular membership function, trapezoidal membership function, Gaussian membership function, etc. For example:

[0094] One of the elements in a set of accurate monitoring data is a heart rate value of 72bpm, and the corresponding preset state range value is 60-100bpm. Then the membership function is a linear membership function, and the membership calculation includes:

[0095]

[0096] Among them, MD is the degree of membership.

[0097] In this embodiment, constructing a membership judgment matrix based on the membership corresponding to N groups of precise monitoring data means arranging the membership values ​​of each element indicator of each group of data in the N groups of precise monitoring data into a matrix, each group of data corresponds to a row of the matrix, and the membership value of each element indicator corresponds to a column of the matrix. The membership judgment matrix constructed in this way can intuitively represent the degree of deviation of each group of data under different health status indicators, thereby being used for subsequent analysis and judgment.

[0098] It can be understood that the construction method of the predictive evaluation model is basically the same as the construction method of the monitoring and correction model, except that the input data is adjusted to N groups of historical precise monitoring data and a historical affiliation judgment matrix, and the output is changed to a historical evaluation result. In this embodiment, the evaluation result can be a health risk warning, such as "low risk", "medium risk" or "high risk", so as to promptly remind users or medical personnel to intervene, or it can be an abnormal warning to identify which specific element indicators (such as heart rate values, blood pressure values, etc.) have experienced abnormal fluctuations within a certain period of time.

[0099] In this embodiment, the fuzzy ideal solution corresponding to the M groups of initial monitoring data is first calculated, the deviation distance between each group of monitoring data and the fuzzy ideal solution is calculated, N groups of target monitoring data are screened out from the M groups of initial monitoring data according to the deviation distance, the external influence value is calculated based on the external influence data, and the N groups of target monitoring data are corrected based on the external influence value to obtain N groups of precise monitoring data. Finally, a membership judgment matrix is ​​generated based on the N groups of precise monitoring data, and the N groups of precise monitoring data and the membership judgment matrix are input into a pre-built predictive evaluation model to obtain an evaluation result. This embodiment can effectively filter out inaccurate initial monitoring data through the above-mentioned multi-level data processing and optimization method, and perform intelligent correction in combination with external influencing factors, thereby significantly improving the accuracy of smart bracelet data acquisition and reducing the impact of noise interference on subsequent evaluation results, thereby providing more accurate and reliable health assessments.

[0100] Example 2

[0101] Based on Example 1, the medical intelligent monitoring method based on the smart bracelet also includes:

[0102] The real-time location information and the real-time movement trajectory are obtained, and the real-time location information and the real-time movement trajectory are cascaded and fused to generate a fusion result, and the fusion result is input into a pre-built monitoring and early warning model to obtain a monitoring result.

[0103] It should be noted that the user's real-time location information and real-time movement trajectory can be obtained according to the high-precision locator in the smart bracelet, and whether the user is in an emergency situation can be judged based on the real-time location information and real-time movement trajectory. For example, when the user is a patient, when the patient falls out of the bed and gets out of bed, the user's real-time location information will change significantly, such as the patient's height information, etc., but the real-time movement trajectory may not change significantly. Therefore, the generated monitoring result is to immediately alarm the central control system, notify medical staff and handle it, to avoid accidents.

[0104] It should be added that the above-mentioned cascade fusion refers to converting the real-time location information and the real-time movement trajectory into feature vectors, and then splicing them end to end. Cascade fusion is an existing technology and this embodiment will not go into details.

[0105] The construction methods of monitoring and early warning models include:

[0106] Obtain Q groups of training data, where Q is a positive integer greater than 1, and the training data includes historical fusion results and historical monitoring results;

[0107] The historical fusion results and historical monitoring results are used as sample sets, and the sample sets are divided into training sets and test sets to build a classifier;

[0108] The historical fusion results in the training set are used as input data, and the historical monitoring results in the training set are used as output data to train the classifier and obtain the initial classifier;

[0109] The initial classifier is tested using the test set, and a classifier that meets the preset accuracy is output as a monitoring and early warning model. The classifier is preferably a naive Bayes model or a support vector machine model.

[0110] Example 3

[0111] See also Figure 2 As shown, based on the same inventive concept, this embodiment discloses a medical intelligent monitoring system based on a smart bracelet. For details not provided in this embodiment, please refer to the description of the relevant parts in Embodiment 1. The system includes:

[0112] Filtering module: used to obtain M groups of initial monitoring data within a preset time period, calculate the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculate the deviation distance between each group of initial monitoring data and the fuzzy ideal solution, and filter out N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1;

[0113] In this embodiment, the preset time period can be set according to prior knowledge or expert experience. For example, the preset time period can be 10 minutes, then a set of initial monitoring data is obtained every half minute. The initial monitoring data represents the physiological characteristics of the user. For example, the initial monitoring data may include heart rate value, blood pressure value, blood oxygen saturation value, and respiratory rate value.

[0114] The fuzzy ideal solution includes a positive fuzzy ideal solution and a negative fuzzy ideal solution. The method for calculating the fuzzy ideal solution corresponding to the M groups of initial monitoring data includes:

[0115] Traverse M groups of initial monitoring data, combine the corresponding element indicators in the M groups of initial monitoring data to construct S fuzzy sets, determine the optimal element indicators and the worst element indicators in the S fuzzy sets based on the preset state range values, combine the S optimal element indicators to construct a positive fuzzy ideal solution, and combine the S worst element indicators to construct a negative fuzzy ideal solution. The initial monitoring data includes S element indicators.

[0116] The deviation distance includes positive deviation distance and negative deviation distance. The method for calculating the positive deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes:

[0117]

[0118] In the formula, is the positive deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, x ij is the jth element index in the i-th group of initial monitoring data, is the optimal element index corresponding to the positive fuzzy ideal solution.

[0119] Similarly, the method for calculating the negative deviation distance between each set of monitoring data and the fuzzy ideal solution includes:

[0120]

[0121] In the formula, is the negative deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, is the worst element index corresponding to the negative fuzzy ideal solution.

[0122] The method for selecting N groups of target monitoring data from M groups of initial monitoring data according to the deviation distance includes:

[0123] The real-time fit value corresponding to each set of initial monitoring data is calculated according to the deviation distance, and the absolute value of the difference between the real-time fit value and the standard fit value is calculated. The initial monitoring data whose absolute value of the difference is greater than the preset difference threshold is eliminated, so as to screen out N groups of target monitoring data.

[0124] It can be understood that the fit value represents the degree of deviation of each set of initial monitoring data, and the degree of deviation here refers to the degree of deviation from the user's ideal state.

[0125] Methods for calculating real-time fit values ​​include:

[0126]

[0127] In the formula, C i is the real-time fit value corresponding to the i-th group of initial monitoring data.

[0128] Correction module: used to obtain external impact data within a preset time period, calculate the external impact degree value based on the external impact data, and correct N groups of target monitoring data based on the external impact degree value to obtain N groups of accurate monitoring data;

[0129] Among them, the external impact data includes but is not limited to the wearing pressure value, skin humidity value and signal-to-noise ratio value. The wearing pressure value represents the degree of contact between the smart bracelet and the user's skin. The skin humidity value represents the degree of wetness on the user's skin surface (such as sweating). Excessive skin humidity may affect the optical and conductivity readings of the sensor, resulting in distorted or inaccurate data. The signal-to-noise ratio value represents the degree of stability of the smart bracelet during the detection process. The larger the signal-to-noise ratio value, the worse the stability of the smart bracelet during the detection process.

[0130] Methods for calculating external influence values ​​based on external influence data include:

[0131]

[0132] Where EXV is the external influence value, WEP is rt is the wearing pressure value, WEP sd is the standard pressure value, SKM is the skin moisture value, SNR is the signal-to-noise ratio value, cosh[·] is the hyperbolic cosine function, cot -1 [·] is the inverse cotangent function, EXP(·) is the exponential function with base e, ln(·) is the logarithmic function with base e, and e is a natural constant.

[0133] The method for correcting N groups of target monitoring data based on external influence values ​​includes:

[0134] The external influence value and N groups of target monitoring data are input into the pre-built monitoring correction model to obtain N groups of accurate monitoring data.

[0135] The construction method of the monitoring and correction model includes:

[0136] Acquire a sample data set, wherein the sample data set includes historical external impact values, historical target monitoring data, and historical precise monitoring data;

[0137] Divide the sample data set into a sample training set and a sample test set, and build a regression network;

[0138] The historical external influence values ​​and historical target monitoring data in the sample training set are used as the input data of the regression network, and the historical accurate monitoring data in the sample training set is used as the output data of the regression network. The regression network is trained to obtain an initial regression network for predicting accurate monitoring data.

[0139] The initial regression network is tested using a sample test set, and the initial regression network with an output smaller than a preset error value is used as a monitoring and correction model. The initial regression network is a deep neural network model.

[0140] Evaluation module: Generate a membership judgment matrix based on N groups of precise monitoring data, input the N groups of precise monitoring data and the membership judgment matrix into the pre-built prediction evaluation model to obtain the evaluation results;

[0141] The method of generating a membership judgment matrix based on N groups of accurate monitoring data includes:

[0142] Traverse N groups of precise monitoring data, normalize each element indicator in each group of precise monitoring data, determine the corresponding membership based on the preset state range value and the normalized element indicator, and construct a membership judgment matrix based on the membership corresponding to the N groups of precise monitoring data.

[0143] Methods for determining the corresponding membership include:

[0144] The preset state range value is normalized, and the membership between the element index and the preset state range value is calculated according to the membership function.

[0145] It can be understood that the membership function refers to a mathematical function used to calculate the membership value of an element index. Common membership functions include linear membership function, triangular membership function, trapezoidal membership function, Gaussian membership function, etc.

[0146] Example 4

[0147] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the medical intelligent monitoring method based on the smart bracelet provided by the above methods is implemented.

[0148] Since the electronic device introduced in this embodiment is an electronic device used to implement the medical intelligent monitoring method based on a smart bracelet in the embodiment of the present application, based on the medical intelligent monitoring method based on a smart bracelet introduced in the embodiment of the present application, technical personnel in this field can understand the specific implementation method of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is no longer introduced in detail. As long as technical personnel in this field implement the electronic device used by the medical intelligent monitoring method based on a smart bracelet in the embodiment of the present application, it belongs to the scope of protection of this application.

[0149] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0150] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD) or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0151] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0152] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0153] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0154] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0155] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0156] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0157] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A medical intelligent monitoring method based on a smart bracelet, characterized in that: include: Obtain M groups of initial monitoring data within a preset time period, calculate the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculate the deviation distance between each group of initial monitoring data and the fuzzy ideal solution, and select N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1; Acquire external impact data within a preset time period, calculate external impact degree values ​​based on the external impact data, and correct N groups of target monitoring data based on the external impact degree values ​​to obtain N groups of accurate monitoring data; A membership judgment matrix is ​​generated based on N groups of precise monitoring data, and the N groups of precise monitoring data and the membership judgment matrix are input into a pre-built prediction and evaluation model to obtain an evaluation result.

2. The medical intelligent monitoring method based on the smart bracelet according to claim 1 is characterized in that: The fuzzy ideal solution includes a positive fuzzy ideal solution and a negative fuzzy ideal solution. The method for calculating the fuzzy ideal solution corresponding to the M groups of initial monitoring data includes: Traverse M groups of initial monitoring data, combine the corresponding element indicators in the M groups of initial monitoring data to construct S fuzzy sets, determine the optimal element indicators and the worst element indicators in the S fuzzy sets based on the preset state range values, combine the S optimal element indicators to construct a positive fuzzy ideal solution, and combine the S worst element indicators to construct a negative fuzzy ideal solution. The initial monitoring data includes S element indicators.

3. The medical intelligent monitoring method based on the smart bracelet according to claim 2 is characterized in that: The deviation distance includes a positive deviation distance and a negative deviation distance. The method for calculating the positive deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes: In the formula, is the positive deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, x ij is the jth element index in the i-th group of initial monitoring data, is the optimal element index corresponding to the positive fuzzy ideal solution.

4. The medical intelligent monitoring method based on the smart bracelet according to claim 3 is characterized in that: The method for calculating the negative deviation distance between each set of initial monitoring data and the fuzzy ideal solution includes: In the formula, is the negative deviation distance between the i-th group of initial monitoring data and the fuzzy ideal solution, is the worst element index corresponding to the negative fuzzy ideal solution.

5. The medical intelligent monitoring method based on the smart bracelet according to claim 4 is characterized in that: The method for selecting N groups of target monitoring data from M groups of initial monitoring data according to the deviation distance includes: The real-time fit value corresponding to each set of initial monitoring data is calculated according to the deviation distance, and the absolute value of the difference between the real-time fit value and the standard fit value is calculated. The initial monitoring data whose absolute value of the difference is greater than the preset difference threshold is eliminated, so as to screen out N groups of target monitoring data.

6. The medical intelligent monitoring method based on the smart bracelet according to claim 5 is characterized in that: The method for calculating the real-time fit value includes: In the formula, C i is the real-time fit value corresponding to the i-th group of initial monitoring data.

7. The medical intelligent monitoring method based on the smart bracelet according to claim 5 is characterized in that: The method for correcting N groups of target monitoring data based on external influence values ​​includes: Input the external impact value and N groups of target monitoring data into the pre-built monitoring correction model to obtain N groups of accurate monitoring data; The construction method of the monitoring and correction model includes: Acquire a sample data set, wherein the sample data set includes historical external impact values, historical target monitoring data, and historical precise monitoring data; Divide the sample data set into a sample training set and a sample test set, and build a regression network; The historical external influence values ​​and historical target monitoring data in the sample training set are used as the input data of the regression network, and the historical accurate monitoring data in the sample training set is used as the output data of the regression network. The regression network is trained to obtain an initial regression network for predicting accurate monitoring data. The initial regression network is tested using a sample test set, and the initial regression network with an output smaller than a preset error value is used as a monitoring and correction model.

8. The medical intelligent monitoring method based on the smart bracelet according to claim 7 is characterized in that: The method for generating a membership judgment matrix based on N groups of accurate monitoring data includes: Traverse N groups of precise monitoring data, normalize each element indicator in each group of precise monitoring data, determine the corresponding membership based on the preset state range value and the normalized element indicator, and construct a membership judgment matrix based on the membership corresponding to the N groups of precise monitoring data.

9. A medical intelligent monitoring system based on a smart bracelet, which is used to implement the medical intelligent monitoring method based on a smart bracelet as described in any one of claims 1 to 8, characterized in that: include: Filtering module: used to obtain M groups of initial monitoring data within a preset time period, calculate the fuzzy ideal solution corresponding to the M groups of initial monitoring data, calculate the deviation distance between each group of monitoring data and the fuzzy ideal solution, and filter out N groups of target monitoring data from the M groups of initial monitoring data according to the deviation distance, M>N>1; Correction module: used to obtain external impact data within a preset time period, calculate the external impact degree value based on the external impact data, and correct N groups of target monitoring data based on the external impact degree value to obtain N groups of accurate monitoring data; Evaluation module: Generate a membership judgment matrix based on N groups of precise monitoring data, input the N groups of precise monitoring data and the membership judgment matrix into the pre-built prediction evaluation model to obtain the evaluation results.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the medical intelligent monitoring method based on the smart bracelet is implemented as described in any one of claims 1 to 8.

Images