Prediction model construction method, system and method for hypertension accompanied with cognitive impairment
By constructing a predictive model of hypertension-associated cognitive impairment, combining brain imaging and physiological indicators, the early identification problem of hypertension-associated cognitive impairment is solved, effective evaluation and intervention of cognitive function is achieved, and biology is provided Marker.
Patent Information
- Application Number
- CN202410863037.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2024-11-19
AI Technical Summary
The prior art is difficult to effectively identify and predict cognitive impairment associated with hypertension, and the lack of objective imaging indicators as biomarkers, resulting in difficulty in early identification and intervention.
By obtaining brain imaging data and cognitive assessment data of hypertensive patients, using multiple regression analysis and mediation models, we screened out both predictive and differentiated brain imaging indicators, and constructed a predictive model of hypertension-related cognitive impairment, combined with Structural gray matter, white matter fiber bundles and resting brain functions reveal neural mechanisms and obtain biological markers.
It realizes early identification and targeted intervention in cognitive impairment in patients with hypertension, reveals the impact of neuroimaging indicators on cognitive function, provides biological markers, reveals the interaction between hypertension severity and physiological indicators on cognitive function effect.
Smart Images

Figure CN118983074A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a prediction model construction method for hypertension-associated cognitive impairment, and also relates to a corresponding prediction model construction system and a prediction method for hypertension-associated cognitive impairment, belonging to the technical field of cognitive assessment. Background Art
[0002] Cognitive impairment is a state of decreased cognitive function, including mild cognitive impairment (MCI) and dementia. This type of cognitive impairment may be caused by a variety of factors, among which modifiable risk factors for cardiovascular diseases (CVDs), such as hypertension, diabetes, smoking and dyslipidemia, have been shown to be associated with an increased risk of cognitive impairment.
[0003] Cardiovascular diseases themselves, especially abnormal fluctuations in blood pressure, coronary heart disease, atrial fibrillation and heart failure, have been proven to accelerate the decline of cognitive functions in the brain, including processing speed, executive function, language ability, learning and memory. It is worth noting that the risk of cardiovascular disease patients with cognitive impairment is significantly higher than that of the general population, and the presence of cognitive impairment may also increase the risk of cardiovascular disease or lead to poor prognosis.
[0004] At present, the clinical diagnosis of cognitive impairment mainly relies on neuropsychological assessment, imaging examinations such as cranial MRI, and genetic testing. However, these examinations are difficult to implement. For example, neuropsychological assessments such as the Montreal Cognitive Assessment (MoCA) and the Mini-Mental State Examination (MMSE) require professional doctors to perform, which is time-consuming and may lead to misdiagnosis due to improper assessment. In addition, for patients with severe depression, dementia, or illiteracy and aphasia, these assessment methods may not be effective even in the best of circumstances.
[0005] Although cognitive impairment is associated with damage to brain function and structure, there is currently no method to assess and screen cognitive impairment using objective imaging indicators as biomarkers. This means that early identification and intervention of cognitive impairment in patient groups such as those with hypertension remain challenging. Summary of the invention
[0006] The primary technical problem to be solved by the present invention is to provide a method for constructing a prediction model for hypertension-associated cognitive impairment.
[0007] Another technical problem to be solved by the present invention is to provide a prediction model construction system for hypertension-associated cognitive impairment.
[0008] Another technical problem to be solved by the present invention is to provide a method for predicting cognitive impairment associated with hypertension.
[0009] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:
[0010] According to a first aspect of an embodiment of the present invention, there is provided a method for constructing a prediction model for hypertension associated with cognitive impairment, comprising the following steps:
[0011] Obtain brain imaging data and cognitive assessment data from a group of hypertensive patients;
[0012] Using the cognitive assessment data of the hypertension patient group as a predictor variable, a predictive brain imaging index capable of predicting hypertension-associated cognitive impairment was selected through multivariate regression analysis;
[0013] Dividing the high-pressure patient group into a mild cognitive impairment group and a severe cognitive impairment group according to the degree of cognitive impairment, using the cognitive assessment data of the high-pressure patient group as a predictor variable, and obtaining differential brain imaging indicators with significant differences between the groups through comparison;
[0014] Taking the intersection of the predictive brain imaging index and the differential brain imaging index to obtain a set of brain imaging indexes capable of screening cognitive impairment in hypertensive patients; wherein the set of brain imaging indexes includes a plurality of brain imaging indexes that are both predictive and differential;
[0015] A mediation model was constructed by taking one of the multiple physiological indicators of the hypertensive patient group as an independent variable, the cognitive assessment data of the hypertensive patient group as a dependent variable, and a concurrent brain imaging indicator as a mediating variable;
[0016] Repeat the above process until each physiological index takes each concurrent brain imaging index as the mediating variable, constructs all mediation models and performs significance tests respectively;
[0017] Based on the significance test results of each mediation model, all established mediation models are obtained to jointly form a prediction model for hypertension-associated cognitive impairment.
[0018] Preferably, the brain image data at least includes:
[0019] Resting state scan data, network partitioning of the supracortical data of hypertensive patients and calculation of the functional connection matrix between each network partition, based on the network partitioning of the supracortical data and the functional connection strength of each network, extract the signal of each voxel of the subcortical structure and establish a connection with the signal of the supracortical network, thereby calculating the functional network of the subcortical structure and the network connection matrix;
[0020] Structural imaging data were used to calculate the white matter volume, gray matter volume, and cortical thickness of each hypertensive patient. ROI analysis was combined to extract key brain regions related to cognitive function, and the cortical thickness and gray matter volume of key brain regions were calculated respectively.
[0021] Diffusion tensor imaging data are used to calculate the direction of white matter fiber bundles and the anisotropy indicators of different fiber bundles for each hypertensive patient. The connection strength between specific fiber bundles and the connection network of different fiber bundles are calculated through probabilistic fiber tracking methods.
[0022] Preferably, the predictive brain imaging index is obtained by the following steps:
[0023] For each brain imaging index, the data information of the brain imaging index was used as the independent variable, the demographic information was used as the control variable, and the total score of the cognitive assessment was used as the predictor variable for multiple regression analysis;
[0024] Based on the results of multiple regression analysis, determine whether this brain imaging index can reveal the impact of hypertension cognitive impairment on brain function and brain structure;
[0025] If it can be revealed, then the brain imaging indicator is a predictive brain imaging indicator; otherwise, the brain imaging indicator is not a predictive brain imaging indicator.
[0026] Preferably, the differential brain imaging index is obtained by the following steps:
[0027] Based on the preset criteria, the hypertensive patients were divided into mild cognitive impairment group and severe cognitive impairment group according to the MoCA scale scores;
[0028] For each brain imaging indicator, data information of the mild cognitive impairment group and the severe cognitive impairment group were obtained and compared;
[0029] Determine whether the data comparison result of the brain imaging index has significant difference. If so, the brain imaging index is a differential brain imaging index. Otherwise, the brain imaging index is not a differential brain imaging index.
[0030] Preferably, a significance test is performed on the mediation model. If the test result is significant, the mediation model is established; otherwise, the mediation model is not established.
[0031] Preferably, the physiological indicators of the hypertensive patient include at least:
[0032] Blood pressure data measured by a blood pressure measuring device, the blood pressure data at least including systolic pressure, diastolic pressure, mean arterial pressure and pulse pressure difference;
[0033] echocardiographic data obtained by echocardiography;
[0034] The blood data is obtained by blood measurement, and the blood data at least includes data information of blood sugar, cholesterol, liver function and kidney function.
[0035] Preferably, the significance test of the mediation model includes the following sub-steps:
[0036] The existing samples in the mediation model are used to represent the overall samples, and sampling with replacement is performed from the existing samples, and n samples are drawn each time to form a sample set, where n is a positive integer;
[0037] Calculate and fit the mediation effect value for the sample set;
[0038] Repeat the above process until a preset number of mediation effect values are obtained;
[0039] Based on the preset number of mediation effect values, calculating the 95% confidence interval of the mediation effect value to form a bootstrap confidence interval with the 2.5th percentile and the 97.5th percentile as boundaries;
[0040] Check whether the bootstrap confidence interval contains 0 to verify whether the mediation effect is significant; if the bootstrap confidence interval contains 0, it means that the mediation model is not significant, otherwise, it means that the mediation model is significant.
[0041] According to a second aspect of an embodiment of the present invention, there is provided a prediction model construction system for hypertension associated with cognitive impairment, comprising a processor and a memory, wherein the processor reads a computer program in the memory to execute the above method.
[0042] According to a third aspect of an embodiment of the present invention, there is provided a method for predicting hypertension associated with cognitive impairment, comprising:
[0043] Obtaining physiological indicators of the patient to be predicted;
[0044] Inputting the physiological index into a prediction model to output the cognitive assessment score of the patient to be predicted, and a plurality of concurrent brain imaging indexes that affect the cognitive prediction score;
[0045] Wherein, the prediction model is constructed by the above method, and different cognitive prediction scores correspond to different degrees of cognitive impairment.
[0046] Preferably, if there are multiple physiological indicators, all the intermediary models in the prediction model are traversed for each physiological indicator to output multiple cognitive assessment scores; wherein the total number of cognitive assessment scores output by the prediction model is less than or equal to the total number of the physiological indicators;
[0047] An average of the multiple cognitive assessment scores is calculated to take the average cognitive assessment score as the final cognitive assessment score of the patient to be predicted.
[0048] Compared with the prior art, the present invention has the following technical effects:
[0049] 1. For the first time, the neural mechanism of cognitive impairment in hypertensive patients was systematically revealed by combining structural gray matter, white matter fiber bundles and resting-state brain function. Based on this, the relationship between neuroimaging indicators and cognitive scores can be obtained, and the corresponding biological markers can be obtained.
[0050] 2. For the first time, all blood indicators such as blood sugar, cholesterol, liver function, kidney function, and ultrasound cardiology indicators are integrated to reveal how the severity of hypertension and various physiological indicators interact with each other in hypertensive patients and ultimately affect the patient's cognitive function. In addition, the mediation model reveals how various physiological indicators of hypertension ultimately lead to cognitive impairment in patients by affecting brain function.
[0051] 3. By dividing cognitive functions into episodic memory, processing speed, working memory, and visuospatial ability, the different dimensions of cognitive impairment in hypertensive patients were explored, and the neural and physiological mechanisms specific to impairments in different dimensions were revealed in a more targeted manner. Based on this, the relationship between physiological indicators, brain structure and function, and specific cognitive scores can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] Figure 1 A schematic diagram of the overall process of a method for constructing a prediction model for hypertension associated with cognitive impairment provided in the first embodiment of the present invention;
[0053] Figure 2 A schematic diagram of the principle of a method for constructing a prediction model for hypertension associated with cognitive impairment provided in the first embodiment of the present invention;
[0054] Figure 3 This is a schematic diagram of a simple mediation model in the first embodiment of the present invention;
[0055] Figure 4 This is a schematic diagram of a moderated mediation model in the first embodiment of the present invention;
[0056] Figure 5 A structural diagram of a prediction model building system for hypertension associated with cognitive impairment provided in a second embodiment of the present invention;
[0057] Figure 6 This is a flow chart of a method for predicting hypertension-associated cognitive impairment provided in the third embodiment of the present invention. DETAILED DESCRIPTION
[0058] The technical content of the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.
[0059] The embodiment of the present invention extracts brain imaging and structural indicators from existing hypertensive patients, and selects brain imaging indicators that can predict patients with hypertension accompanied by cognitive impairment. Then, based on the physiological indicators of hypertensive patients, the obtained brain imaging indicators are jointly modeled to establish a mediation model between hypertension-brain imaging indicators-cognitive scores. In the modeling process, the patient's overall cognitive function score is first used as the dependent variable, the brain impact index is used as the mediating variable, and the physiological index is used as the independent variable. Through a simple mediation model and a mediation model with adjustment, the physiological and brain mechanisms of overall cognitive function damage are revealed. Finally, an association prediction model of physiological indicators-brain structural imaging indicators-cognitive scores of patients with hypertension and cognitive impairment is obtained.
[0060] In addition, in other embodiments of the present invention, sub-cognitive functions can also be used as dependent variables to explore the specific neuro-physiological mechanisms of sub-cognitive function impairment.
[0061] First embodiment
[0062] like Figure 1 and Figure 2 As shown, a prediction model construction method for hypertension associated with cognitive impairment provided by the first embodiment of the present invention specifically includes steps S1 to S8:
[0063] S1: Obtain brain imaging data and cognitive assessment data of hypertensive patients.
[0064] Specifically, in this embodiment, the brain imaging data includes resting state scanning data, structural imaging data and diffusion tensor imaging data. Different data processing is required for different data types. The specific data processing process is as follows:
[0065] (1) For resting-state scan data: perform network partitioning on the supracortical data of hypertensive patients and calculate the functional connectivity matrix between each network partition. Based on the network partitioning of supracortical data and the functional connectivity strength of each network, extract the signal of each voxel in the subcortical structure and establish a connection with the signal of the supracortical network, thereby calculating the functional network of the subcortical structure and the network connectivity matrix.
[0066] (2) For structural imaging data: calculate the white matter volume, gray matter volume, and cortical thickness of each hypertensive patient, extract key brain regions related to cognitive function through ROI analysis, and calculate the cortical thickness and gray matter volume of key brain regions respectively;
[0067] (3) For diffusion tensor imaging data: calculate the orientation of white matter fiber bundles and the anisotropy index of different fiber bundles for each hypertensive patient, and calculate the connection strength between specific fiber bundles and the connection network of different fiber bundles through probabilistic fiber tracking method.
[0068] In addition, it is also necessary to obtain cognitive assessment data for each hypertensive patient, specifically the total score of the cognitive assessment. It is understandable that the above brain imaging data and cognitive assessment data can be directly obtained based on the existing database; or each hypertensive patient can be re-performed with MRI and a corresponding cognitive assessment can be performed using a preset scale.
[0069] S2: Screening predictive brain imaging indicators.
[0070] After completing the data preprocessing based on step S1, the cognitive assessment data of the hypertension patient group is used as a predictive variable, and predictive brain imaging indicators that can predict hypertension-associated cognitive impairment are selected through multivariate regression analysis.
[0071] Specifically, it includes steps S21 to S23:
[0072] S21: For each brain imaging index, the data information of the brain imaging index was used as the independent variable, the demographic information was used as the control variable, and the total score of the cognitive assessment was used as the predictor variable for multiple regression analysis;
[0073] S22: Based on the results of multiple regression analysis, determine whether this brain imaging index can reveal the impact of hypertension cognitive impairment on brain function and structure;
[0074] S23: If it can be revealed, then the brain imaging indicator is a predictive brain imaging indicator; otherwise, the brain imaging indicator is not a predictive brain imaging indicator.
[0075] It is understandable that in step S2, a multivariate regression analysis needs to be performed on each brain imaging index to screen out predictive brain imaging indexes from all brain imaging indexes.
[0076] S3: Obtain differential brain imaging indicators.
[0077] After completing the data preprocessing based on step S1, the high-pressure patient group is divided into a mild cognitive impairment group and a severe cognitive impairment group according to the degree of cognitive impairment. The cognitive assessment data of the high-pressure patient group is used as a predictor variable, and differential brain imaging indicators with significant differences between the groups are obtained through comparison.
[0078] Specifically, it includes steps S31 to S33:
[0079] S31: Based on the preset criteria, the hypertensive patients were divided into mild cognitive impairment group and severe cognitive impairment group according to the MoCA scale scores;
[0080] S32: For each brain imaging index, obtain data information of the mild cognitive impairment group and the severe cognitive impairment group respectively, and compare the data;
[0081] S33: Determine whether the data comparison result of the brain imaging index has a significant difference. If so, the brain imaging index is a differential brain imaging index. Otherwise, the brain imaging index is not a differential brain imaging index.
[0082] It is understandable that in step S3, it is necessary to perform comparative analysis on each brain imaging index to screen out differential brain imaging indexes from all brain imaging indexes. Moreover, all brain imaging indexes in step S3 are exactly the same as all brain imaging indexes in step S2, except that brain imaging indexes of different natures are found from all brain imaging indexes in different ways. In addition, it is understandable that there is no order between step S3 and step S2, and the two can be performed simultaneously or separately.
[0083] S4: Take the intersection and obtain the comprehensive brain imaging indicators.
[0084] After obtaining the predictive brain imaging index based on step S2 and obtaining the differential brain imaging index through step S3, the intersection of the two is taken to obtain a set of brain imaging indexes that can screen cognitive impairment in hypertensive patients. The set of brain imaging indexes includes a plurality of brain imaging indexes that are both predictive and differential.
[0085] S5: Construct a mediation model.
[0086] Specifically, one of the multiple physiological indicators of the hypertensive patient group was used as the independent variable X, the cognitive assessment data of the hypertensive patient group was used as the dependent variable Y, and a concurrent brain imaging indicator was used as the mediating variable M. A mapping relationship from the physiological indicator X to the degree of cognitive impairment Y was constructed to form a mediation model in which the physiological indicator corresponded to the concurrent brain imaging indicator.
[0087] In this example, in order to test the relationship between basic hypertension indicators, brain imaging and cognitive function, we first used a simple mediation model for verification. A mediation model refers to a mediation effect in which the effect of an independent variable on a dependent variable is affected by a third variable (e.g. Figure 3 The mediation effect is defined as the path from the independent variable X to the mediator variable M and the path from the mediator variable M to the dependent variable Y (i.e., ab), ab = c–c'. c represents the slope of the overall effect of X on Y without controlling M, while c' represents the slope of the overall effect of X on Y after M is added to the model.
[0088] First, without controlling M, the overall effect of X on Y is expressed as follows:
[0089] Y=i Y +cX (1)
[0090] Among them, i Y represents the intercept and c represents the slope of the overall effect.
[0091] Secondly, the effect of independent variable X on mediating variable M is expressed as follows:
[0092] M=i M +aX (2)
[0093] where i M represents the intercept and a represents the effect slope.
[0094] Finally, the mediation model expression formula of independent variable X and mediating variable M on dependent variable Y is as follows:
[0095] Y=i Y +c'X+bM (3)
[0096] Substituting the above formula (2) into formula (3), we can get:
[0097] Y=i Y +c'X+bi M +abX (4)
[0098] S6: Conduct significance test on the mediation model.
[0099] Specifically, it includes steps S61 to S65:
[0100] S61: Using the existing samples in the mediation model to represent the overall samples, sampling with replacement is performed from the existing samples, and n samples are drawn each time to form a sample set, where n is a positive integer;
[0101] S62: Calculate and fit the mediation effect value for the sample set;
[0102] S63: Repeat the above process until a preset number (e.g., 5000) of mediation effect values are obtained;
[0103] S64: Based on the preset number of mediation effect values, calculate a 95% confidence interval of the mediation effect value to form a bootstrap confidence interval with the 2.5th percentile and the 97.5th percentile as boundaries;
[0104] S65: Check whether the bootstrap confidence interval contains 0 to verify whether the mediation effect is significant; if the bootstrap confidence interval contains 0, it means that the mediation model is not significant, otherwise, it means that the mediation model is significant.
[0105] It can be understood that, for the mediation model, if the test result is significant, the mediation model is established, indicating that the mediation model is applicable. On the contrary, if the test result is not significant, the mediation model is not established, indicating that the mediation model is not applicable.
[0106] S7: Repeat the above steps S5 to S6 until each physiological index is used as a mediating variable with each concurrent brain imaging index, all mediation models are constructed and significance tests are performed respectively.
[0107] It is understandable that for the same physiological indicator, different mediation models can be constructed with different concurrent brain imaging indicators as mediating variables. For example: assuming that there are 10 physiological indicators and the brain imaging indicator set includes 10 concurrent brain imaging indicators, 10 mediation models can be constructed for each physiological indicator, and a total of 10*10=100 mediation models can be constructed. However, not all of these 100 mediation models are available. Only mediation models that pass the significance test will be retained, and other mediation models can be directly deleted.
[0108] In this embodiment, the physiological indicators of hypertensive patients include at least:
[0109] (1) Blood pressure data measured by a blood pressure measuring device, which includes at least systolic pressure, diastolic pressure, mean arterial pressure and pulse pressure difference;
[0110] (2) echocardiographic data obtained by echocardiography;
[0111] (3) Blood data obtained through blood measurement, the blood data at least includes data information on blood sugar, cholesterol, liver function and kidney function.
[0112] S8: Based on the significance test results of each mediation model, all established mediation models are obtained to jointly form a prediction model for hypertension-associated cognitive impairment.
[0113] It can be understood that the prediction model constructed in this embodiment is equivalent to a collection of mediation models, which includes which physiological indicators are affected by which brain imaging indicators, thereby affecting the patient's final total cognitive assessment score (ie, the degree of cognitive impairment).
[0114] In addition, in other embodiments, in order to reveal how different hypertension indicators interact with each other to affect brain function and ultimately affect cognitive function. In this embodiment, a moderated mediation model (such as Figure 4 (as shown) to reveal how different physiological indicators interact to affect brain function, and whether the mediation effect exists and the mediation model is established after controlling different demographic indicators.
[0115] It is understandable that in the moderated mediation model, the independent variable X is the evaluation index of the severity of hypertension, and the moderating variable W is the demographic information index such as age, gender, and education level; XW is the possible interaction between physiological indicators such as blood, liver and kidney function and demographic information. Similarly, the cognitive function is divided into episodic memory, processing speed, working memory, and visual-spatial ability, and the sub-cognitive function score is used as the dependent variable. Based on this, the relationship between physiological indicators-brain structure function-specific cognitive score can be obtained, forming a prediction model of the specific physiological-neural mechanism of sub-cognitive function damage.
[0116] Second embodiment
[0117] like Figure 5 As shown, based on the above-mentioned prediction model construction method for hypertension-associated cognitive impairment, the present invention further provides a prediction model construction system for hypertension-associated cognitive impairment. The prediction model construction system includes one or more processors 21 and a memory 22. Among them, the memory 22 is coupled to the processor 21 and is used to store one or more programs. When the one or more programs are executed by the one or more processors 21, the one or more processors 21 implement the prediction model construction method for hypertension-associated cognitive impairment in the above-mentioned embodiment.
[0118] Among them, the processor 21 is used to control the overall operation of the prediction model construction system to complete all or part of the steps of the prediction model construction method for hypertension associated with cognitive impairment. The processor 21 can be a central processing unit (CPU), a graphics processing unit (GPU), a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a digital signal processing (DSP) chip, etc. The memory 22 is used to store various types of data to support the operation of the prediction model construction system, and these data may include instructions for any application or method for operating on the prediction model construction system, and application-related data. The memory 22 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, etc.
[0119] In an exemplary embodiment, the prediction model construction system can be implemented by a computer chip or entity, or by a product with a certain function, for executing the above-mentioned prediction model construction method for hypertension associated with cognitive impairment, and achieving the technical effect consistent with the above-mentioned method. A typical embodiment is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a vehicle-mounted human-computer interaction device, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0120] In another exemplary embodiment, the present invention further provides a computer-readable storage medium including program instructions, which, when executed by a processor, implements the steps of the method for constructing a prediction model of hypertension associated with cognitive impairment in any of the above embodiments. For example, the computer-readable storage medium may be the above-mentioned memory including program instructions, and the above-mentioned program instructions may be executed by a processor of a construction system to complete the above-mentioned method for constructing a prediction model of hypertension associated with cognitive impairment, and achieve the technical effect consistent with the above-mentioned method.
[0121] Third embodiment
[0122] like Figure 6 As shown, a method for predicting hypertension associated with cognitive impairment provided by the third embodiment of the present invention specifically includes steps S10 to S20:
[0123] S10: Obtain physiological indicators of the patient to be predicted.
[0124] S20: Input the physiological indicators into the prediction model to output the cognitive assessment score of the patient to be predicted, as well as a plurality of combined brain imaging indicators that affect the cognitive prediction score.
[0125] The prediction model is constructed by the method of the first embodiment, and different cognitive prediction scores correspond to different degrees of cognitive impairment. Moreover, the combined brain imaging index can provide guidance for subsequent cognitive intervention, so as to improve the effect of cognitive intervention.
[0126] In addition, in the above step S10, if there are multiple physiological indicators, all the mediation models in the prediction model are traversed for each physiological indicator to output multiple cognitive assessment scores. For example: if 5 physiological indicators are input in step S10, all the mediation models in the prediction model are traversed for each physiological indicator. If a corresponding mediation model exists, the corresponding cognitive assessment score is output. If not, the output result is empty. Therefore, the total number of cognitive assessment scores finally output by the prediction model is less than or equal to 5. Then, the average of the multiple cognitive assessment scores is calculated to take the average cognitive assessment score as the final cognitive assessment score of the patient to be predicted.
[0127] In summary, the prediction model construction method, construction system and prediction method for hypertension associated with cognitive impairment provided by the embodiments of the present invention have the following beneficial effects:
[0128] 1. For the first time, the neural mechanism of cognitive impairment in hypertensive patients was systematically revealed by combining structural gray matter, white matter fiber bundles and resting-state brain function. Based on this, the relationship between neuroimaging indicators and cognitive scores can be obtained, and the corresponding biological markers can be obtained.
[0129] 2. For the first time, all blood indicators such as blood sugar, cholesterol, liver function, kidney function, and ultrasound cardiology indicators are integrated to reveal how the severity of hypertension and various physiological indicators interact with each other in hypertensive patients and ultimately affect the patient's cognitive function. In addition, the mediation model reveals how various physiological indicators of hypertension ultimately lead to cognitive impairment in patients by affecting brain function.
[0130] 3. The present invention explores the different dimensions of cognitive impairment in hypertensive patients by dividing cognitive functions into episodic memory, processing speed, working memory, and visual-spatial ability, and more specifically reveals the neural and physiological mechanisms of impairment-specificity in different dimensions, thereby obtaining the relationship between physiological indicators-brain structure and function-specific cognitive scores.
[0131] It should be noted that the above-mentioned multiple embodiments are only examples, and the technical solutions of various embodiments can be combined, all of which are within the protection scope of the present invention.
[0132] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0133] The above is a detailed description of the prediction model construction method, system and method for hypertension-associated cognitive impairment provided by the present invention. For those skilled in the art, any obvious changes made to it without departing from the essence of the present invention will constitute an infringement of the patent right of the present invention and will bear the corresponding legal liability.
Claims
1. A method for constructing a prediction model for hypertension associated with cognitive impairment, characterized in that The steps include: Obtain brain imaging data and cognitive assessment data from a group of hypertensive patients; Using the cognitive assessment data of the hypertension patient group as a predictor variable, a predictive brain imaging index capable of predicting hypertension-associated cognitive impairment was selected through multivariate regression analysis; Dividing the high-pressure patient group into a mild cognitive impairment group and a severe cognitive impairment group according to the degree of cognitive impairment, using the cognitive assessment data of the high-pressure patient group as a predictor variable, and obtaining differential brain imaging indicators with significant differences between the groups through comparison; Taking the intersection of the predictive brain imaging index and the differential brain imaging index to obtain a set of brain imaging indexes capable of screening cognitive impairment in hypertensive patients; wherein the set of brain imaging indexes includes a plurality of brain imaging indexes that are both predictive and differential; An intermediate model is constructed by taking one of the multiple physiological indicators of the hypertensive patient group as an independent variable, taking the cognitive assessment data of the hypertensive patient group as a dependent variable, and taking a concurrent brain imaging indicator as a mediating variable; Repeat the above process until each physiological index takes each concurrent brain imaging index as the mediating variable, constructs all mediation models and performs significance tests respectively; Based on the significance test results of each mediation model, all established mediation models are obtained to jointly form a prediction model for hypertension-associated cognitive impairment.
2. The prediction model construction method according to claim 1, characterized in that The brain image data at least includes: Resting state scan data, network partitioning of the supracortical data of hypertensive patients and calculation of the functional connection matrix between each network partition, based on the network partitioning of the supracortical data and the functional connection strength of each network, extract the signal of each voxel of the subcortical structure and establish a connection with the signal of the supracortical network, thereby calculating the functional network of the subcortical structure and the network connection matrix; Structural imaging data were used to calculate the white matter volume, gray matter volume, and cortical thickness of each hypertensive patient. ROI analysis was combined to extract key brain regions related to cognitive function, and the cortical thickness and gray matter volume of key brain regions were calculated respectively. Diffusion tensor imaging data are used to calculate the direction of white matter fiber bundles and the anisotropy indicators of different fiber bundles for each hypertensive patient. The connection strength between specific fiber bundles and the connection network of different fiber bundles are calculated through probabilistic fiber tracking methods.
3. The prediction model construction method according to claim 1, characterized in that The predictive brain imaging index is obtained by the following steps: For each brain imaging index, the data information of the brain imaging index was used as the independent variable, the demographic information was used as the control variable, and the total score of the cognitive assessment was used as the predictor variable for multiple regression analysis; Based on the results of multiple regression analysis, determine whether this brain imaging index can reveal the impact of hypertension cognitive impairment on brain function and structure; If it can be revealed, then the brain imaging indicator is a predictive brain imaging indicator; otherwise, the brain imaging indicator is not a predictive brain imaging indicator.
4. The prediction model construction method according to claim 1, characterized in that The differential brain imaging index is obtained by the following steps: Based on the preset criteria, the hypertensive patients were divided into mild cognitive impairment group and severe cognitive impairment group according to the MoCA scale scores; For each brain imaging indicator, data information of the mild cognitive impairment group and the severe cognitive impairment group were obtained and compared; Determine whether the data comparison result of the brain imaging index has significant difference. If so, the brain imaging index is a differential brain imaging index. Otherwise, the brain imaging index is not a differential brain imaging index.
5. The prediction model construction method according to claim 1, characterized in that After constructing the intermediate model, a significance test is performed on the mediation model. If the test result is significant, the mediation model is established; otherwise, the mediation model is not established.
6. The prediction model construction method according to claim 1, characterized in that The physiological indicators of the hypertensive patient include at least: Blood pressure data measured by a blood pressure measuring device, the blood pressure data at least including systolic pressure, diastolic pressure, mean arterial pressure and pulse pressure difference; echocardiographic data obtained by echocardiography; The blood data is obtained by blood measurement, and the blood data at least includes data information of blood sugar, cholesterol, liver function and kidney function.
7. The prediction model construction method according to claim 1, characterized in that The significance test of the mediation model includes the following sub-steps: The existing samples in the mediation model are used to represent the overall samples, and sampling with replacement is performed from the existing samples, and n samples are drawn each time to form a sample set, where n is a positive integer; Calculate and fit the mediation effect value for the sample set; Repeat the above process until a preset number of mediation effect values are obtained; Based on the preset number of mediation effect values, calculating the 95% confidence interval of the mediation effect value to form a bootstrap confidence interval with the 2.5th percentile and the 97.5th percentile as boundaries; Check whether the bootstrap confidence interval contains 0 to verify whether the mediation effect is significant; if the bootstrap confidence interval contains 0, it means that the mediation model is not significant, otherwise, it means that the mediation model is significant.
8. A prediction model construction system for hypertension associated with cognitive impairment, characterized in that It comprises a processor and a memory, wherein the processor reads a computer program in the memory to execute the prediction model building method according to any one of claims 1 to 7.
9. A method for predicting hypertension-associated cognitive impairment, characterized in that include: Obtaining physiological indicators of the patient to be predicted; Inputting the physiological index into a prediction model to output the cognitive assessment score of the patient to be predicted, and a plurality of concurrent brain imaging indexes that affect the cognitive prediction score; Wherein, the prediction model is constructed by the prediction model construction method described in any one of claims 1 to 7, and different cognitive prediction scores correspond to different degrees of cognitive impairment.
10. The prediction method according to claim 9, characterized in that: If there are multiple physiological indicators, all the intermediary models in the prediction model are traversed for each physiological indicator to output multiple cognitive assessment scores; wherein the total number of cognitive assessment scores output by the prediction model is less than or equal to the total number of the physiological indicators; An average of the multiple cognitive assessment scores is calculated to take the average cognitive assessment score as the final cognitive assessment score of the patient to be predicted.
