Cerebral stroke diagnosis aided decision-making method and device and storage medium

Abstract

The invention provides a cerebral apoplexy diagnosis aided decision-making method and device and a storage medium, and relates to the technical field of cerebral apoplexy data processing, and the method comprises the steps: importing physiological parameter values of a patient according to a time sequence, and carrying out the preprocessing, so as to enable the data to be regular and complete; constructing a time sequence model based on the preprocessed data, and generating a fluctuation graph to visually display the change trend of the physiological parameters along with time; key values are obtained from the fluctuation diagram, an auxiliary decision-making model is constructed, and auxiliary decision-making data are output after the key values are input. Preprocessing provides a high-quality data basis for subsequent analysis modeling, a fluctuation diagram assists in clearly knowing the state of illness development, an auxiliary decision-making model provides powerful support reference for cerebral apoplexy diagnosis, diagnosis accuracy and efficiency are improved, high-risk patients are assisted in being recognized in time, and corresponding measures are taken.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of stroke data processing, and specifically to a stroke diagnosis auxiliary decision-making method, device and storage medium. Background Art

[0002] Stroke, a major disease that poses a serious threat to human health, is characterized by high morbidity, disability, and mortality. Globally, a large number of new stroke cases occur each year, placing a heavy burden on society and families. Effective prevention and control of stroke is crucial.

[0003] The current field of stroke diagnosis and treatment faces numerous challenges. From a data acquisition perspective, while wearable devices, home medical monitoring instruments, and various hospital monitoring devices can collect a wealth of patient physiological parameter data, such as blood pressure, blood sugar, heart rate, and oxygen saturation, these data often vary in format and quality, and are subject to missing data, outliers, and other issues. Furthermore, the data's time series recording frequency is inconsistent, making subsequent data integration and analysis extremely difficult.

[0004] When it comes to clinical assessment, medical professionals primarily rely on experience to assess patients' conditions, a method that lacks precise quantitative evidence. For example, when determining the severity of a patient's condition, doctors often rely on their own experience and limited clinical indicators, making it difficult to provide a comprehensive and accurate assessment of the patient's overall condition.

[0005] Medical imaging tests, such as CT and MRI, are important tools in stroke diagnosis and treatment. However, these tests have limitations. Firstly, their high cost increases the financial burden on patients. In areas with limited medical resources, some patients may not be able to undergo timely examinations. Secondly, these tests are not suitable for all patients, and some patients cannot undergo such tests due to physical conditions or other reasons. Furthermore, imaging tests can usually only be performed after symptoms appear, making it difficult to provide early warning.

[0006] In terms of risk prediction, most existing risk prediction models are built based on a single factor or a small number of factors, and are unable to comprehensively consider the patient's multi-dimensional information, resulting in poor prediction accuracy. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method, device and storage medium for assisting decision-making in diagnosing stroke in view of the deficiencies in the existing technology.

[0008] The present invention solves the above-mentioned technical problem with the following technical solution: A method for assisting decision-making in diagnosing stroke, comprising:

[0009] Importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values;

[0010] constructing a time series model based on the preprocessed physiological parameter values, and generating a fluctuation graph based on an output result of the time series model;

[0011] Key values ​​are obtained from the fluctuation graph, an auxiliary decision model is constructed, the key values ​​are input into the auxiliary decision model, and auxiliary decision data is output.

[0012] Another technical solution of the present invention to solve the above technical problem is as follows: a stroke diagnosis and decision-making auxiliary device, comprising:

[0013] A preprocessing module, used for importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values;

[0014] a fluctuation graph generating module, configured to construct a time series model based on the preprocessed physiological parameter values, and to generate a fluctuation graph based on an output result of the time series model;

[0015] The decision output module is used to obtain key values ​​from the fluctuation diagram, construct an auxiliary decision model, input the key values ​​into the auxiliary decision model, and output auxiliary decision data.

[0016] Another technical solution of the present invention to solve the above-mentioned technical problem is as follows: A stroke diagnosis and decision-making support device includes 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 stroke diagnosis and decision-making support method described above is implemented.

[0017] Another technical solution of the present invention to solve the above technical problem is as follows: a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned stroke diagnosis and decision-making auxiliary method is implemented.

[0018] The beneficial effects of the present invention are: by importing the physiological parameter values ​​of stroke patients in time series and performing preprocessing, the data can be made more regular and complete, providing a high-quality data basis for subsequent analysis and modeling, and constructing a time series model based on the preprocessed physiological parameter values, and generating a fluctuation graph, which can intuitively display the changing trend of physiological parameters over time, helping doctors or researchers to understand the development dynamics of the patient's condition more clearly, and extracting key values ​​from the fluctuation graph, constructing an auxiliary decision-making model, and outputting auxiliary decision-making data, which can provide strong support and reference for the diagnosis of stroke, improve the accuracy and efficiency of diagnosis, and help to timely identify high-risk patients and take corresponding treatment measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1A schematic diagram of a flow chart of a stroke diagnosis and decision-making support method provided by an embodiment of the present invention;

[0020] Figure 2 This is a functional module block diagram of the stroke diagnosis and decision support device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0022] Example 1: Figure 1 As shown, an embodiment of the present invention provides a method for assisting decision-making in stroke diagnosis, comprising:

[0023] Importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values;

[0024] constructing a time series model based on the preprocessed physiological parameter values, and generating a fluctuation graph based on an output result of the time series model;

[0025] Key values ​​are obtained from the fluctuation graph, an auxiliary decision model is constructed, the key values ​​are input into the auxiliary decision model, and auxiliary decision data is output.

[0026] In the above embodiment, by importing the physiological parameter values ​​of stroke patients in time series and performing preprocessing, the data can be made more regular and complete, providing a high-quality data basis for subsequent analysis and modeling, and constructing a time series model based on the preprocessed physiological parameter values, and generating a fluctuation graph, which can intuitively display the changing trend of physiological parameters over time, helping doctors or researchers to understand the development dynamics of the patient's condition more clearly, and extracting key values ​​from the fluctuation graph, constructing a decision-making support model, and outputting decision-making support data, which can provide strong support and reference for the diagnosis of stroke, improve the accuracy and efficiency of diagnosis, and help to timely identify high-risk patients and take corresponding treatment measures.

[0027] Preferably, the preprocessing of the physiological parameter value includes:

[0028] The missing values ​​in the physiological parameter values ​​are supplemented by a linear interpolation formula, and the linear interpolation formula is:

[0029]

[0030] Where t1 and t2 are time points, and t1<t<t2, x(t1) is the value corresponding to time point t1, x(t2) is the value corresponding to time point t2, and x(t) is the missing value at time point t.

[0031] Suppose you want to analyze the blood pressure data of a stroke patient. The blood pressure data was collected hourly and monitored continuously for 72 hours.

[0032] Among these 72 data, it is found that the data of the 10th hour and the 35th hour are missing. Taking the data of the 10th hour as an example, the blood pressure value of the previous time point (the 9th hour) is Y9=120mmHg, and the blood pressure value of the next time point (the 11th hour) is Y 11 =125mmHg. According to the linear interpolation formula, where t1=9, t2=11, and t=10, the missing blood pressure value at the 10th hour is calculated. Similarly, the missing value at the 35th hour can be calculated to complete the data preprocessing.

[0033] In the above embodiment, when linear interpolation fills missing values, it will be calculated based on the values ​​of adjacent known data points, which can better maintain the original trends and rules of the data without introducing too many abnormalities or unreasonable changes, thereby more realistically reflecting the patient's physiological state.

[0034] Preferably, the step of constructing a time series model based on the preprocessed physiological parameter values ​​and generating a fluctuation graph based on the output results of the time series model comprises:

[0035] Setting autoregressive parameters The autoregressive parameters Represents the influence weight of the physiological parameter value of the past i time intervals on the current value, i∈p, and sets the moving average parameter θ j , the moving average parameter θ j represents the error term ∈ of the past j time intervals t-j Correction of the current physiological parameter value, j∈q, setting the physiological parameter value at time t to Y t , based on the autoregressive parameters Moving average parameter θ j And the physiological parameter value is Y t Construct a time series model, the time series model is:

[0036]

[0037] Among them, ∈ t is a negligible physiological fluctuation value, Y t-i is the physiological parameter value of the previous i day;

[0038] The horizontal axis is time t and the physiological parameter value Y t The vertical axis is used, and a fluctuation chart is generated using data visualization tools.

[0039] Specifically, set the autoregressive parameter p = 2 and the moving average parameter q = 1. Assume θ1=0.4,∈t is a negligible physiological fluctuation value (assumed to be 0). Taking the 50th hour as an example, the blood pressure value Y in the previous hour (49th hour) 49 =130 mmHg, blood pressure value Y in the first 2 hours (48th hour) 48 =128 mmHg, the error term at 49 hours∈ 49 = 0. According to the time series model formula, calculate the blood pressure prediction value Y at the 50th hour 50 = 0.6 × 130 + 0.3 × 128 + 0.4 × 0 + 0 = 128.4 mmHg. Similarly, calculate the predicted value for each hour in the entire time period.

[0040] Next, using data visualization tools (such as Python's Matplotlib library), the original and predicted blood pressure data are plotted into a fluctuation chart, with time (hours) as the horizontal axis and blood pressure values ​​(mmHg) as the vertical axis. On this fluctuation chart, doctors or researchers can visually visualize blood pressure trends over time, such as increases, decreases, or relative stability during certain time periods. For example, the fluctuation chart shows that between the 20th and 30th hours, blood pressure shows a gradual upward trend, while between the 50th and 60th hours, blood pressure remains relatively stable.

[0041] In the above embodiment, by setting autoregressive parameters and moving average parameters and constructing a time series model, it is possible to predict future changes in physiological parameter values, provide the possibility for early warning and intervention of stroke, and help take measures before the condition worsens; generating a fluctuation graph can display complex physiological parameter data in an intuitive graphical form, allowing doctors or researchers to quickly and intuitively observe key information such as the changing trends and abnormal fluctuations of physiological parameters, facilitate further analysis and diagnosis, and improve work efficiency.

[0042] Preferably, obtaining key values ​​from the fluctuation graph, constructing an auxiliary decision model, inputting the key values ​​into the auxiliary decision model, and outputting auxiliary decision data include:

[0043] The change rate of adjacent time points is calculated based on the physiological parameter values ​​in the fluctuation graph using a change rate formula to obtain multiple key physiological parameter values ​​as key values. The change rate formula is:

[0044]

[0045] Among them, ChangeRate t is the rate of change of physiological parameters at time t, Y t is the physiological parameter value at time t, Y t-1 is the physiological parameter value at time t-1, Δt is the time interval,

[0046] Set the change rate threshold CR , when |ChangeRate t |≥Threshold CR When the physiological parameter value Y t As the key value X, X=[X1,X2,...,X n ],

[0047] Constructing auxiliary decision-making models based on logistic regression algorithm:

[0048]

[0049] Among them, P(Y=1) represents the probability of stroke, β0 is the intercept, β1, β2,…, β n is the regression coefficient;

[0050] Substitute the key value X into the auxiliary decision model to output the probability of stroke occurrence,

[0051] Set the probability threshold P , when P(Y=1)≥Threshold p When P(Y=1)<Threshold p When , low-risk decision data is output.

[0052] Specifically, suppose the systolic blood pressure of a stroke patient is monitored, and data is recorded every 30 minutes to obtain a series of systolic blood pressure values. In the fluctuation graph, from hour 1 to hour 1.5, the systolic blood pressure increases from 120 mmHg to 135 mmHg. According to the rate of change formula, here Y t =135mmHg, Y t-1 =120mmHg, Δt=0.5 hours (30 minutes is converted to 0.5 hours), then the systolic blood pressure change rate during this period is Set the change rate threshold CR =20 mmHg / h. Since 30 mmHg / h>20 mmHg / h., the systolic blood pressure value corresponding to 1.5 hours, 135 mmHg, is used as the key value X.

[0053] Next, the clinical data set is imported to train the decision support model and obtain the values ​​corresponding to each regression coefficient.

[0054] Substitute the key value X = 135 mmHg into the trained decision support model to calculate the probability of stroke, for example Set the probability threshold to Threshold P=0.6. Since 0.735 > 0.6, a high-risk decision is output. This means that based on the currently acquired key values ​​and the constructed model, the patient has a high risk of stroke. Doctors can use this result to further diagnose and take appropriate preventive and treatment measures, such as arranging further examinations and adjusting treatment plans.

[0055] In the above embodiment, by calculating the rate of change of adjacent time points and setting a change rate threshold, key values ​​can be screened out from a large amount of physiological parameter data. These key values ​​can better reflect the important changes in the condition and provide a core basis for diagnosis; an auxiliary decision model is constructed based on a logistic regression algorithm, and the probability of stroke occurrence is output after the key values ​​are input into the model. High-risk or low-risk decision data is output based on the probability threshold, providing a quantitative risk assessment for the doctor's diagnostic decision, helping to make more accurate and scientific judgments and reduce the possibility of misdiagnosis and missed diagnosis.

[0056] Example 2: Figure 2 As shown, an embodiment of the present invention further provides a stroke diagnosis decision-making auxiliary device, comprising:

[0057] A preprocessing module, used for importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values;

[0058] a fluctuation graph generating module, configured to construct a time series model based on the preprocessed physiological parameter values, and to generate a fluctuation graph based on an output result of the time series model;

[0059] The decision output module is used to obtain key values ​​from the fluctuation diagram, construct an auxiliary decision model, input the key values ​​into the auxiliary decision model, and output auxiliary decision data.

[0060] Preferably, the preprocessing of the physiological parameter value includes:

[0061] The missing values ​​in the physiological parameter values ​​are supplemented by a linear interpolation formula, and the linear interpolation formula is:

[0062]

[0063] Where t1 and t2 are time points, and t1<t<t2, x(t1) is the value corresponding to time point t1, x(t2) is the value corresponding to time point t2, and x(t) is the missing value at time point t.

[0064] Preferably, the step of constructing a time series model based on the preprocessed physiological parameter values ​​and generating a fluctuation graph based on the output results of the time series model comprises:

[0065] Setting autoregressive parameters The autoregressive parameters Represents the influence weight of the physiological parameter value of the past i time intervals on the current value, i∈p, and sets the moving average parameter θ j , the moving average parameter θ j represents the error term ∈ of the past j time intervals t-j Correction of the current physiological parameter value, j∈q, setting the physiological parameter value at time t to Y t , based on the autoregressive parameters Moving average parameter θ j And the physiological parameter value is Y t Construct a time series model, the time series model is:

[0066]

[0067] Among them, ∈ t is a negligible physiological fluctuation value, Y t-i is the physiological parameter value of the previous i day;

[0068] The horizontal axis is time t and the physiological parameter value Y t The vertical axis is used, and a fluctuation chart is generated using data visualization tools.

[0069] Preferably, obtaining key values ​​from the fluctuation graph, constructing an auxiliary decision model, inputting the key values ​​into the auxiliary decision model, and outputting auxiliary decision data include:

[0070] The change rate of adjacent time points is calculated based on the physiological parameter values ​​in the fluctuation graph using a change rate formula to obtain multiple key physiological parameter values ​​as key values. The change rate formula is:

[0071]

[0072] Among them, ChangeRate t is the rate of change of physiological parameters at time t, Y t is the physiological parameter value at time t, Y t-1 is the physiological parameter value at time t-1, Δt is the time interval,

[0073] Set the change rate threshold CR , when |ChangeRate t |≥Threshold CR When the physiological parameter value Y t As the key value X, X=[X1,X2,...,X n ],

[0074] Constructing auxiliary decision-making models based on logistic regression algorithm:

[0075]

[0076] Among them, P(Y=1) represents the probability of stroke, β0 is the intercept, β1, β2,…, β n is the regression coefficient;

[0077] Substitute the key value X into the auxiliary decision model to output the probability of stroke occurrence,

[0078] Set the probability threshold P , when P(Y=1)≥Threshold p When P(Y=1)<Threshold p When , low-risk decision data is output.

[0079] Example 3: Another embodiment of the present invention provides a stroke diagnosis and decision-making support 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 stroke diagnosis and decision-making support method described above is implemented.

[0080] Embodiment 4: Another embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the above-mentioned stroke diagnosis decision-making support method is implemented.

[0081] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0082] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of units is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple units or components into another system, or ignoring or not implementing certain features.

[0083] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the objectives of the embodiments of the present invention.

[0084] 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 scope of protection of the present invention.

Claims

1. A stroke diagnosis decision-making support method, characterized in that: include: Importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values; constructing a time series model based on the preprocessed physiological parameter values, and generating a fluctuation graph based on an output result of the time series model; Key values ​​are obtained from the fluctuation graph, an auxiliary decision model is constructed, the key values ​​are input into the auxiliary decision model, and auxiliary decision data is output.

2. The stroke diagnosis and decision-making support method according to claim 1, characterized in that: The preprocessing of the physiological parameter value includes: The missing values ​​in the physiological parameter values ​​are supplemented by a linear interpolation formula, and the linear interpolation formula is: Where t1 and t2 are time points, and t1<t<t2, x(t1) is the value corresponding to time point t1, x(t2) is the value corresponding to time point t2, and x(t) is the missing value at time point t.

3. The stroke diagnosis and decision-making support method according to claim 1, characterized in that: The step of constructing a time series model based on the preprocessed physiological parameter values ​​and generating a fluctuation graph based on the output results of the time series model includes: Setting autoregressive parameters The autoregressive parameters Represents the influence weight of the physiological parameter value of the past i time intervals on the current value, i∈p, and sets the moving average parameter θ j , the moving average parameter θ j represents the error term ∈ of the past j time intervals t-j Correction of the current physiological parameter value, j∈q, setting the physiological parameter value at time t to Y t , based on the autoregressive parameters Moving average parameter θ j And the physiological parameter value is Y t Construct a time series model, the time series model is: Among them, ∈ t is a negligible physiological fluctuation value, Y t-i is the physiological parameter value of the previous i day; The horizontal axis is time t and the physiological parameter value Y t The vertical axis is used, and a fluctuation chart is generated using data visualization tools.

4. The stroke diagnosis and decision-making support method according to claim 1, characterized in that: The step of obtaining key values ​​from the fluctuation graph, constructing an auxiliary decision model, inputting the key values ​​into the auxiliary decision model, and outputting auxiliary decision data includes: The change rate of adjacent time points is calculated based on the physiological parameter values ​​in the fluctuation graph using a change rate formula to obtain multiple key physiological parameter values ​​as key values. The change rate formula is: Among them, ChangeRate t is the rate of change of physiological parameters at time t, Y t is the physiological parameter value at time t, Y t-1 is the physiological parameter value at time t-1, Δt is the time interval, Set the change rate threshold CR , when |ChangeRate t |≥Threshold CR When the physiological parameter value Y t As the key value X, X=[X1,X2,...,X n ], Constructing auxiliary decision-making models based on logistic regression algorithm: Among them, P(Y=1) represents the probability of stroke, β0 is the intercept, β1, β2,…, β n is the regression coefficient; Substitute the key value X into the auxiliary decision model to output the probability of stroke occurrence, Set the probability threshold P When P(Y=1)≥Thresholdp, high-risk decision data is output. When P(Y=1)<Threshold p When , low-risk decision data is output.

5. A stroke diagnosis and decision-making support device, characterized in that: include: A preprocessing module, used for importing physiological parameter values ​​of stroke patients in time series and preprocessing the physiological parameter values; a fluctuation graph generating module, configured to construct a time series model based on the preprocessed physiological parameter values, and to generate a fluctuation graph based on an output result of the time series model; The decision output module is used to obtain key values ​​from the fluctuation diagram, construct an auxiliary decision model, input the key values ​​into the auxiliary decision model, and output auxiliary decision data.

6. The stroke diagnosis and decision-making support device according to claim 5, characterized in that: The preprocessing of the physiological parameter value includes: The missing values ​​in the physiological parameter values ​​are supplemented by a linear interpolation formula, and the linear interpolation formula is: Where t1 and t2 are time points, and t1<t<t2, x(t1) is the value corresponding to time point t1, x(t2) is the value corresponding to time point t2, and x(t) is the missing value at time point t.

7. The stroke diagnosis and decision-making support device according to claim 5, characterized in that: The step of constructing a time series model based on the preprocessed physiological parameter values ​​and generating a fluctuation graph based on the output results of the time series model includes: Setting autoregressive parameters The autoregressive parameters Represents the influence weight of the physiological parameter value of the past i time intervals on the current value, i∈p, and sets the moving average parameter θ j , the moving average parameter θ j represents the error term ∈ of the past j time intervals t-j Correction of the current physiological parameter value, j∈q, setting the physiological parameter value at time t to Y t , based on the autoregressive parameters Moving average parameter θ j And the physiological parameter value is Y t Construct a time series model, the time series model is: Among them, ∈ t is a negligible physiological fluctuation value, Y t-i is the physiological parameter value of the previous i day; The horizontal axis is time t and the physiological parameter value Y t The vertical axis is used, and a fluctuation chart is generated using data visualization tools.

8. The stroke diagnosis and decision-making support device according to claim 5, characterized in that: The step of obtaining key values ​​from the fluctuation graph, constructing an auxiliary decision model, inputting the key values ​​into the auxiliary decision model, and outputting auxiliary decision data includes: The change rate of adjacent time points is calculated based on the physiological parameter values ​​in the fluctuation graph using a change rate formula to obtain multiple key physiological parameter values ​​as key values. The change rate formula is: Among them, ChangeRate t is the rate of change of physiological parameters at time t, Y t is the physiological parameter value at time t, Y t-1 is the physiological parameter value at time t-1, Δt is the time interval, Set the change rate threshold CR , when |ChangeRate t |≥Threshold CR When the physiological parameter value Y t As the key value X, X=[X1,X2,...,X n ], Constructing auxiliary decision-making models based on logistic regression algorithm: Among them, P(Y=1) represents the probability of stroke, β0 is the intercept, β1, β2,…, β n is the regression coefficient; Substitute the key value X into the auxiliary decision model to output the probability of stroke occurrence, Set the probability threshold P , when P(Y=1)≥Threshold p When P(Y=1)<Threshold p When , low-risk decision data is output.

9. A stroke diagnosis and decision-making support device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the stroke diagnosis and decision-making support method as described in any one of claims 1 to 4 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the stroke diagnosis and decision-making auxiliary method according to any one of claims 1 to 4 is implemented.

Images