Cognitive impairment behavior monitoring and early warning method for alzheimer disease

Abstract

The application relates to the technical field of cognitive disorder behavior monitoring, in particular to a cognitive disorder behavior monitoring and early warning method for Alzheimer's disease, which acquires gait parameters of each gait of an Alzheimer's disease patient, screens abnormal gaits, clusters the abnormal gaits based on distance measurement between any two abnormal gaits, obtains a plurality of class clusters, acquires gait disorder features of each class cluster, obtains gait turning features of each class cluster based on differences in disorder conditions between adjacent gaits in each class cluster, fuses the gait disorder features and the gait turning features of each class cluster, determines cognitive disorder behavior indexes of the Alzheimer's disease patient in combination with time intervals between adjacent class clusters, and performs early warning. Compared with a manual observation mode, the application comprehensively analyzes gaits of the Alzheimer's disease patient from multiple aspects, so that accurate cognitive disorder behavior monitoring results can be obtained.

Description

Technical Field

[0001] This invention relates to the field of cognitive impairment behavior monitoring technology, specifically to a method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease. Background Technology

[0002] Alzheimer's disease (AD) is a common neurodegenerative disease, primarily characterized by cognitive decline, memory loss, and behavioral changes. The cognitive impairment and emotional and behavioral problems experienced by Alzheimer's patients cause significant distress to both the patients and their families. Monitoring and early warning of cognitive impairment and behavioral changes in Alzheimer's patients are therefore crucial.

[0003] Existing methods for monitoring cognitive impairment behaviors in Alzheimer's disease typically involve analyzing the gait of Alzheimer's patients through human observation to obtain the monitoring results. This method has a large margin of error and it is difficult to ensure the accuracy of the cognitive impairment monitoring results. Summary of the Invention

[0004] To address the technical problem of low accuracy in existing methods for monitoring cognitive impairment in Alzheimer's disease, the present invention aims to provide a method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease. The specific technical solution adopted is as follows:

[0005] In a first aspect of the present invention, a method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease is provided, comprising:

[0006] Gait parameters of various gait patterns in Alzheimer's patients were obtained, and abnormal gait patterns were identified based on these parameters.

[0007] Based on the distance metric between any two abnormal gaits, the abnormal gaits are clustered to obtain multiple clusters;

[0008] Gait disorder features are obtained for various clusters, and these features are used to characterize the disorder of gait within a cluster.

[0009] Based on the differences in disorder between adjacent gaits in various clusters, gait turning characteristics of various clusters are obtained;

[0010] By integrating gait disorder and gait turning characteristics of various clusters and combining the time intervals between adjacent clusters, cognitive impairment behavioral indicators of Alzheimer's disease patients can be determined and early warnings can be issued.

[0011] In one exemplary embodiment, gait parameters include: gait speed, maximum pitch angle when the toes leave the ground, and maximum pitch angle when the heels strike the ground.

[0012] In one exemplary embodiment, abnormal gait is identified based on gait parameters, including:

[0013] The gait movement disorder factor is obtained based on the difference in the maximum pitch angle when the toes of both feet leave the ground and the difference in the maximum pitch angle when the heels of both feet land.

[0014] Based on the gait disorder factor and gait speed, abnormal characteristic indicators of gait are obtained; the abnormal characteristic indicators are directly proportional to the gait disorder factor and inversely proportional to the gait speed.

[0015] Abnormal gait patterns are identified by screening based on abnormal gait characteristic indicators.

[0016] In an exemplary embodiment, the process of obtaining the distance metric between any two aberrant gaits includes:

[0017] Obtain the differences in abnormal feature indicators and time intervals between any two abnormal gaits;

[0018] The distance metric is obtained based on the differences in anomaly indicator values ​​and the time interval; the distance metric is directly proportional to the differences in anomaly indicator values ​​and the time interval.

[0019] In one exemplary embodiment, gait disorder features of various clusters are obtained, including:

[0020] The abnormal gait patterns in the target cluster are sorted chronologically, and the fitting curves of the motion disorder factors of each abnormal gait pattern are obtained; the target cluster can be any cluster.

[0021] Obtain the variation amplitude of the motion disorder factor for each gait in the fitted curve;

[0022] Based on the variation amplitude of the motion disorder factor of each gait in the fitted curve, the gait disorder characteristics of the target cluster are obtained.

[0023] In an exemplary embodiment, the gait disorder features of the target cluster are obtained, including:

[0024] Slide the window through the fitted curve according to the preset window size to obtain the data window corresponding to each gait in the fitted curve;

[0025] The window disorder coefficient of the data window is obtained based on the sub-disorder coefficients of each gait in the data window; the sub-disorder coefficient of the gait is obtained by the change amplitude of the gait and the frequency proportion of the gait motion disorder factor in the fitted curve.

[0026] By fusing the window disorder coefficients of all data windows in the target cluster, the gait disorder characteristics of the target cluster are obtained.

[0027] In an exemplary embodiment, the process of obtaining gait turning features includes:

[0028] Obtain the first difference between the motion disorder factors of the first and second gait in the fitted curve of the target cluster; the first gait is any gait other than the first and last gait in the fitted curve, and the second gait is the gait following the first gait.

[0029] Obtain the second difference in the motor disorder factors between the second and third gait; the third gait is the gait preceding the first gait.

[0030] Based on the first and second differences, the gait turning characteristics of the first step are obtained; the gait turning characteristics are directly proportional to the first difference and inversely proportional to the second difference.

[0031] In one exemplary embodiment, gait disorder features and gait turning features of various clusters are fused, and the time intervals between adjacent clusters are combined to determine cognitive impairment behavioral indicators for Alzheimer's disease patients, including:

[0032] Based on the gait disorder characteristics and gait turning characteristics of various clusters, the comprehensive gait disorder characteristics of various clusters are obtained;

[0033] Based on the gait disturbance characteristics and temporal distribution characteristics of each cluster, cognitive impairment behavioral index components are obtained for each cluster. The cognitive impairment behavioral index components are directly proportional to the gait disturbance characteristics and inversely proportional to the temporal distribution characteristics. The temporal distribution characteristics are obtained from the time intervals between adjacent clusters.

[0034] By integrating the cognitive impairment behavioral index components of various clusters, cognitive impairment behavioral indicators for Alzheimer's disease patients are obtained.

[0035] In an exemplary embodiment, the process of obtaining time distribution features includes:

[0036] Obtain the intermediate time for each type of cluster;

[0037] The time interval between the intermediate time of each cluster and the previous intermediate time is obtained as the temporal distribution feature of each cluster.

[0038] In one exemplary embodiment, the warning process includes: comparing the cognitive impairment behavior indicators of Alzheimer's patients with a preset warning threshold; if the indicators are greater than or equal to the preset warning threshold, then outputting a warning signal.

[0039] The present invention has the following beneficial effects: By acquiring a large number of gait parameters from Alzheimer's patients, the present invention first filters abnormal gait based on gait parameters, then clusters the abnormal gait to obtain different abnormal gait patterns of Alzheimer's patients. Next, by considering the gait disorder characteristics and gait turning characteristics of the clusters, and combining the time intervals between adjacent clusters, the present invention finally determines the cognitive impairment behavior indicators of Alzheimer's patients to achieve early warning. Compared with the method of manual observation, the present invention comprehensively analyzes the gait of Alzheimer's patients from multiple aspects, thereby obtaining accurate cognitive impairment behavior monitoring results. Attached Figure Description

[0040] Figure 1 This is a flowchart of a method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease, provided in one embodiment of the present invention;

[0041] Figure 2 This is a flowchart of the abnormal gait screening process provided in one embodiment of the present invention;

[0042] Figure 3 This is a flowchart of the process for obtaining the distance metric between any two abnormal gaits, provided in one embodiment of the present invention.

[0043] Figure 4 This is a flowchart of the process for obtaining gait disorder features of various clusters according to an embodiment of the present invention;

[0044] Figure 5 This is a flowchart of the process for obtaining gait disorder features of a target cluster according to an embodiment of the present invention;

[0045] Figure 6 This is a flowchart of the process for obtaining gait turning features according to an embodiment of the present invention;

[0046] Figure 7 This is a flowchart illustrating the acquisition of cognitive impairment behavioral indicators according to an embodiment of the present invention;

[0047] Figure 8 This is a flowchart illustrating the acquisition of time distribution features according to an embodiment of the present invention. Detailed Implementation

[0048] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. All data and information collected in this application have been obtained with full consent, and the collection, use, and processing of such information must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0050] like Figure 1 As shown, this embodiment provides a method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease, including:

[0051] Step 1: Obtain gait parameters for each gait of Alzheimer's patients, and filter out abnormal gaits based on the gait parameters;

[0052] Step 2: Based on the distance metric between any two abnormal gaits, cluster the abnormal gaits to obtain multiple clusters;

[0053] Step 3: Obtain gait disorder features for each type of cluster. Gait disorder features are used to characterize the disorder of gait in each cluster.

[0054] Step 4: Based on the differences in disorder between adjacent gaits in various clusters, obtain the gait turning characteristics of each cluster;

[0055] Step 5: Integrate the gait disorder and gait turning characteristics of various clusters, and combine them with the time intervals between adjacent clusters to determine cognitive impairment behavioral indicators for Alzheimer's patients and provide early warnings.

[0056] The following detailed description, in conjunction with the accompanying drawings, outlines each step of a cognitive impairment behavior monitoring and early warning method for Alzheimer's disease provided in this embodiment.

[0057] Step 1: Obtain gait parameters for each gait of Alzheimer's patients, and based on the gait parameters, filter out abnormal gaits.

[0058] In one specific application, smart insoles are worn by Alzheimer's patients to ensure they are securely and correctly fitted. The smart insoles integrate accelerometers and gyroscopes, and their built-in sensors capture real-time motion data during walking. After wearing the smart insoles, the Alzheimer's patient engages in daily walking. This embodiment pre-sets a monitoring period for monitoring and analyzing the patient's gait patterns within that period. The specific duration of this monitoring period is set according to actual conditions, for example, one hour. Furthermore, to ensure reliable monitoring of Alzheimer's patients, it can be set for multiple consecutive days, such as 10 consecutive days, with a cumulative walking time exceeding 30 minutes each day for gait parameter collection, ensuring data representativeness and long-term monitoring effectiveness. Based on the smart insoles, gait parameters for each gait pattern of the Alzheimer's patient are obtained.

[0059] A single gait refers to a complete walking process, or gait activity, which is a gait behavior from the moment one foot touches the ground until the same foot touches the ground again. It includes the alternating stepping motion of the left and right feet and covers the entire movement process within a single gait cycle.

[0060] In this embodiment, the gait parameters include: gait speed, maximum pitch angle when the toes leave the ground, and maximum pitch angle when the heels land.

[0061] To obtain walking speed, stride length and gait cycle time are needed. Stride length refers to the horizontal distance traveled during a single gait activity from the moment one foot lands until the same foot lands again. Gait cycle time refers to the time taken during a single gait activity from the moment one foot lands until the same foot lands again. Walking speed is equal to the ratio of stride length to gait cycle time.

[0062] Therefore, for each gait of an Alzheimer's patient, the gait speed, maximum toe-off pitch angle, and maximum heel-off pitch angle were obtained. Gait parameters for each gait of the Alzheimer's patient over the past 10 days were recorded for subsequent analysis.

[0063] Alzheimer's disease is closely associated with gait changes, primarily due to damage to higher cognitive brain regions such as the prefrontal cortex and hippocampus. This damage affects motor coordination and flexibility, leading to a lack of coordination in Alzheimer's patients. Changes in gait patterns effectively reflect the motor coordination and cognitive status of Alzheimer's patients at different stages of the disease. The reduced volume of the prefrontal cortex in Alzheimer's patients impairs the region's ability to perform executive functions, attention, and motor planning. Degeneration of the prefrontal cortex weakens the brain's processing and reaction speed to environmental information, thus affecting motor coordination and causing gait instability.

[0064] Then, based on gait parameters, abnormal gaits are filtered out from each gait. In an exemplary embodiment, such as... Figure 2 As shown, the following is a specific screening process for abnormal gait:

[0065] Step 1-1: Based on the difference in the maximum pitch angle when the toes of both feet leave the ground and the difference in the maximum pitch angle when the heels of both feet land, obtain the gait movement disorder factor.

[0066] For ease of explanation, the target gait is set to any gait.

[0067] The difference in maximum pitch angles when the toes of the two feet leave the ground is obtained for the target gait. In an exemplary embodiment, the maximum pitch angles when the toes of the left and right feet leave the ground are set as follows: , Calculate the absolute value of the difference between the two: ,in, The difference in the maximum pitch angle when the toes of both feet leave the ground is defined as the driving asymmetry coefficient of the target gait. The larger the value of this driving asymmetry coefficient, the more asymmetrical and disordered the driving behavior of the left and right feet in the target gait of Alzheimer's patients is, that is, the driving behavior of the gait of Alzheimer's patients is unstable.

[0068] The difference in maximum heel-to-toe pitch angles of the two feet when striking the target gait is obtained. In an exemplary embodiment, the maximum heel-to-toe pitch angles of the left and right feet are set as follows: , Calculate the absolute value of the difference between the two: ,in, The difference in the maximum heel-to-toe pitch angle when both feet strike the ground is defined as the braking asymmetry coefficient of the target gait. The larger the value of this braking asymmetry coefficient, the more asymmetrical and disordered the braking behavior of the left and right feet in the target gait of Alzheimer's patients is, indicating that the braking behavior of the Alzheimer's patient's gait is unstable.

[0069] By combining the difference in maximum pitch angle when the toes leave the ground and the difference in maximum pitch angle when the heels strike the ground, the motion disorder factor of the target gait is obtained. In an exemplary embodiment, the calculation formula for the motion disorder factor of the target gait is as follows:

[0070] ;

[0071] Where R is the movement disorder factor of the target gait. This represents the normalization function. The normalization here can be achieved by obtaining the maximum and minimum values ​​of the sum of the driving and braking asymmetric coefficients for each gait, and then using a maximum / minimum value normalization method. Normalize.

[0072] Step 1-2: Obtain abnormal gait characteristic indicators based on gait movement disorder factors and gait speed.

[0073] Due to the complexity of gait and subtle changes at different stages, there are slight differences and asymmetries in the angle of the left and right feet off the ground. Therefore, relying solely on motor disturbance factors in the target gait to measure abnormal characteristics may be erroneous. Healthy individuals exhibit slight differences between the left and right feet, but this does not affect the overall stability of the gait. Conversely, individuals with neurological impairments (such as Alzheimer's patients) exhibit significant asymmetrical gait characteristics. This asymmetry leads to overall gait instability, accompanied by symptoms of slow walking speed.

[0074] Therefore, by combining the movement disorder factor and gait speed of the target gait, an abnormal characteristic index of the target gait is obtained. The abnormal characteristic index is directly proportional to the movement disorder factor and inversely proportional to the gait speed.

[0075] In an exemplary embodiment, a specific method for calculating the abnormal feature index of the target gait is given below:

[0076] ;

[0077] Where F is the abnormal feature index of the target gait, and v is the gait speed of the target gait.

[0078] This represents the negative correlation normalization of v. The negative correlation normalization method here can be: obtain the maximum and minimum values ​​of the gait speed in each gait, then normalize v using the maximum and minimum value normalization method, and then subtract the value 1 from the normalized v. The result is the negative correlation normalization of v.

[0079] The larger the value of the abnormal characteristic index F of the target gait, the more obvious the gait abnormality of the Alzheimer's disease patient. That is, the gait of Alzheimer's disease patients not only has a strong asymmetry between the left and right feet, but also a slower walking speed and a more unstable gait, which is related to more severe neurological damage or motor dysfunction.

[0080] Using the above process, abnormal feature indices for each gait are obtained.

[0081] Steps 1-3: Based on the abnormal gait characteristic indicators, abnormal gaits are screened out.

[0082] Since a larger value of the abnormal feature index indicates a more abnormal gait, abnormal gaits are selected from each gait based on its abnormal feature index. In an exemplary embodiment, an abnormal feature threshold is preset, with a value ranging from 0 to 1. The specific value of this preset threshold is set according to the actual situation. If a more secure monitoring logic is required, the preset abnormal feature threshold can be set slightly smaller, making it easier to identify abnormal feature indices that meet the monitoring requirements. In this embodiment, a preset abnormal feature threshold of 0.5 is used as an example. Then, the abnormal feature indices of each gait are compared with the preset abnormal feature threshold, and the gait corresponding to the abnormal feature index greater than the preset abnormal feature threshold is determined to be an abnormal gait.

[0083] Step 2: Based on the distance metric between any two abnormal gaits, cluster the abnormal gaits to obtain multiple clusters.

[0084] Alzheimer's disease causes abnormal movement patterns and gait by extensively damaging brain regions, particularly those related to motor control and cognitive function. Due to these neural injuries, Alzheimer's patients exhibit consistent abnormal gait characteristics when performing similar behaviors at different times and in different situations. Therefore, clustering of abnormal gait patterns is necessary.

[0085] In one exemplary embodiment, the k-means clustering algorithm is used for clustering. First, K initial seed points are selected, the specific value of K being set by the desired number of clusters. Then, the distance metric between any two abnormal gaits is obtained. In one exemplary embodiment, as... Figure 3 As shown, the process of obtaining the distance metric between any two abnormal gaits includes:

[0086] Step 2-1: Obtain the difference in abnormal feature indicators and time interval between any two abnormal gaits.

[0087] The abnormal gaits are sorted chronologically. For ease of explanation, let the 1st... The first abnormal gait and the first An abnormal gait is any two distinct abnormal gaits.

[0088] Get the The first abnormal gait and the first The differences in abnormal feature indicators of the first abnormal gait, specifically the first... The first abnormal gait and the first The absolute value of the difference in the abnormal feature indicators of each abnormal gait, in order to To express.

[0089] Get the The first abnormal gait and the first The time interval of the abnormal gait, specifically: obtaining the time interval of the first abnormal gait. The start time of the first abnormal gait, and the first Get the start time of the first abnormal gait. The start time of the first abnormal gait and the first The time interval between the start times of the first abnormal gait is taken as the time interval between the first abnormal gait. The first abnormal gait and the first The time interval of each abnormal gait.

[0090] To improve clustering accuracy, the first... The first abnormal gait and the first After determining the time interval of the first abnormal gait, the time interval is normalized. The normalization method can be as follows: obtain the maximum and minimum values ​​of the time intervals between any two abnormal gaits, and then use the maximum / minimum value normalization method to normalize the time interval of the first abnormal gait. The first abnormal gait and the first The time intervals of each abnormal gait are normalized, and the normalized result is defined as follows: .

[0091] Step 2-2: Obtain the distance metric based on the differences in abnormal feature indicators and time intervals.

[0092] According to the The first abnormal gait and the first Differences in abnormal feature indicators of the first abnormal gait, and the first The first abnormal gait and the first The time interval of the first abnormal gait is obtained. The first abnormal gait and the first Distance metric for the first abnormal gait. The distance metric characterizes the first abnormal gait. The first abnormal gait and the first The greater the time interval between abnormal gaits, the greater the distance between them, and the greater the difference in abnormal characteristic indicators. Therefore, the distance metric is directly proportional to the difference in abnormal characteristic indicators and the time interval.

[0093] In one exemplary embodiment, the first The first abnormal gait and the first The formula for calculating the distance metric of an abnormal gait is as follows:

[0094] ;

[0095] in, Indicates the first The abnormal gait and the first The distance metric for each anomalous gait. The normalization method here could be: obtaining the distance between any two anomalous gaits. Find the maximum and minimum values ​​in the range, and then normalize them using the maximum and minimum value normalization method.

[0096] Clustering is performed using the k-means clustering algorithm and distance metrics. The cluster centroids are continuously updated through iterative clustering until the clusters no longer change, resulting in multiple clusters. Each cluster represents an abnormal gait behavior pattern in Alzheimer's patients, corresponding to a complete abnormal activity.

[0097] Step 3: Obtain gait disorder features for each type of cluster. Gait disorder features are used to characterize the disorder of gait in each cluster.

[0098] Although Alzheimer's disease patients may exhibit significant gait abnormalities at certain stages, cognitive impairment symptoms may not yet appear due to individual differences. This is closely related to the individual preferences and behavioral characteristics of Alzheimer's patients. Therefore, a more refined analysis of various abnormal gait patterns (i.e., clusters) is needed to accurately assess the relationship between gait and cognitive impairment. In healthy individuals, gait is characterized by coordinated and purposeful movement patterns, with steps and movements flexibly adjusted according to needs. In Alzheimer's patients, due to the degeneration of the cerebral cortex and hippocampus, the coordination and stability of gait are impaired, often manifesting as irregular steps and pauses in movement, leading to disordered gait characteristics within the same behavioral activity.

[0099] Therefore, gait disorder features of various clusters are obtained, and these features are used to characterize the disorder of gait within a cluster. In an exemplary embodiment, such as... Figure 4 As shown, the process of obtaining gait disorder features for various clusters includes:

[0100] Step 3-1: Sort each abnormal gait in the target cluster according to time sequence, and obtain the fitting curve of the motion disorder factor of each abnormal gait after sorting.

[0101] For ease of explanation, the target cluster is set to any cluster.

[0102] Since the target cluster includes multiple abnormal gaits, the abnormal gaits in the target cluster are sorted in chronological order, so that the abnormal gaits in the target cluster have a sequential order.

[0103] Based on the motion disorder factors of each abnormal gait after sorting in the target cluster, a polynomial fitting method is used to perform curve fitting on each abnormal gait after sorting in the target cluster, and the fitting curve of the motion disorder factor of the target cluster is obtained.

[0104] Step 3-2: Obtain the variation amplitude of the motion disorder factor for each gait in the fitted curve.

[0105] By performing curve fitting on the motion disorder factors of each abnormal gait in the target cluster, the motion disorder factors corresponding to each data point in the motion disorder factor fitting curve are obtained. Each data point in the motion disorder factor fitting curve is essentially a gait, thus obtaining the motion disorder factors of each gait in the motion disorder factor fitting curve of the target cluster.

[0106] Obtain the variation amplitude of the motion disorder factor for each gait in the motion disorder factor fitting curve of the target cluster. In an exemplary embodiment, obtain the slope of each gait at the position of the fitting curve, take the absolute value of the slope, and use the magnitude of the slope after taking the absolute value as the variation amplitude of the motion disorder factor for each gait in the fitting curve. The larger the slope after taking the absolute value, the larger the variation amplitude of the motion disorder factor for the corresponding gait.

[0107] Step 3-3: Based on the variation amplitude of the motion disorder factor of each gait in the fitted curve, obtain the gait disorder characteristics of the target cluster.

[0108] Based on the variation range of the motion disorder factor of each gait in the motion disorder factor fitting curve of the target cluster, the gait disorder characteristics of the target cluster are obtained. The greater the variation range of the motion disorder factor of each gait, the more disordered the gait of the target cluster.

[0109] In an exemplary embodiment, to improve the accuracy of obtaining gait disorder features of the target cluster, such as Figure 5 As shown, a specific process for obtaining gait disorder features of the target cluster is presented:

[0110] Step 3-3-1: Slide the window across the fitted curve according to the preset window size to obtain the data window corresponding to each gait in the fitted curve.

[0111] A preset window size is used for sliding windowing of the fitting curves of the motion disorder factor of the target cluster. The size of this preset window is set according to the actual situation. If the window is too small, it may not be possible to ensure that all relevant data are included in the analysis; if the window is too large, it may include data with weak correlations in the analysis. In this embodiment, the preset window uses 11 data points as an example.

[0112] A sliding window is applied to the fitted curve of the motion disorder factor of the target cluster according to a preset window size to obtain the data window corresponding to each gait in the target cluster. The sliding window step size is 1. It should be understood that for any gait, defined as the target gait, the target gait is used as the center point of its corresponding sliding window. This allows the five gaits adjacent to the left and the five adjacent gaits to the right of the target gait to be included in the sliding window of the target gait, thus obtaining the data window of the target gait, and consequently, the data window corresponding to each gait in the target cluster.

[0113] It should be understood that for several gaits at the beginning and end of the target cluster, a complete data window may not be possible. For example, taking a preset window of 11 data points as an example, the first 5 gaits and the last 5 gaits in the target cluster may not have 5 gaits before or after them. Therefore, all remaining gaits that meet the conditions are included in the data window. For example, taking the 3rd gait in the target cluster as an example, since there are only 2 gaits before it, the data window for the 3rd gait consists of the 2 gaits before the 3rd gait, the 3rd gait itself, and the 5 gaits after it.

[0114] Step 3-3-2: Obtain the window disorder coefficient of the data window based on the sub-disorder coefficients of each gait in the data window.

[0115] Since each gait corresponds to a motion disorder factor in the motion disorder factor fitting curve of the target cluster, obtaining the number of gaits in the motion disorder factor fitting curve of the target cluster, i.e., the number of data points, is equivalent to obtaining the number of motion disorder factors in the motion disorder factor fitting curve of the target cluster.

[0116] Then, the frequency of the target gait's movement disorder factor appearing in the movement disorder factor fitting curve of the target cluster is obtained. Based on the number of movement disorder factors in the target cluster's movement disorder factor fitting curve, the frequency percentage of the target gait's movement disorder factor appearing in the target cluster's movement disorder factor fitting curve is obtained (the frequency percentage is equal to the ratio of the number of times the target gait's movement disorder factor appears in the target cluster's movement disorder factor fitting curve to the number of movement disorder factors in the target cluster's movement disorder factor fitting curve). This yields the frequency percentage of each gait's movement disorder factor appearing in the target cluster's movement disorder factor fitting curve.

[0117] The sub-disorder coefficient of the target gait is obtained based on the amplitude of change in the target gait and the frequency proportion of the target gait's disorder factors appearing in the disorder factor fitting curve of the target cluster. It should be understood that the larger the amplitude of change in the target gait, the more drastic the gait characteristic changes and the more disordered the gait; simultaneously, the more randomly the disorder factors of the target gait appear in the disorder factor fitting curve of the target cluster, i.e., the lower the frequency proportion, the more disordered the gait. Therefore, the sub-disorder coefficient is directly proportional to the amplitude of change and inversely proportional to the frequency proportion.

[0118] In one exemplary embodiment, a specific formula for calculating the sub-disorder coefficient of the target gait is given below:

[0119] ;

[0120] Where Q represents the sub-disorder coefficient of the target gait, K represents the variation range of the target gait, and P represents the frequency proportion of the motion disorder factor of the target gait in the motion disorder factor fitting curve of the target cluster.

[0121] This indicates the normalization of K. The normalization method here can be: obtain the maximum and minimum values ​​of the variation amplitude of each gait within the data window of the target gait, and normalize K using the maximum and minimum value normalization method.

[0122] Using the above method, the sub-disorder coefficients of each gait are obtained. This yields the sub-disorder coefficients of each gait within the data window of the target gait. Based on the sub-disorder coefficients of each gait within the data window of the target gait, the window disorder coefficient of the target gait's data window is obtained. In this embodiment, the average value of the sub-disorder coefficients of each gait within the data window of the target gait is calculated and used as the window disorder coefficient of the target gait's data window. This results in the window disorder coefficients of each data window within the target cluster.

[0123] Step 3-3-3: Fuse the window disorder coefficients of all data windows in the target cluster to obtain the gait disorder characteristics of the target cluster.

[0124] The average value of the window disorder coefficients of all data windows in the target cluster is calculated and used as the gait disorder feature of the target cluster. This yields the gait disorder features for each cluster.

[0125] Step 4: Based on the differences in disorder between adjacent gaits in various clusters, the gait turning characteristics of each cluster are obtained.

[0126] Patients with cognitive impairment, especially those with Alzheimer's disease, often experience frequent turning during walking due to spatial cognitive impairment, manifesting as a loss of orientation, which further exacerbates gait instability and incoordination. Further in-depth analysis of the abnormal gait patterns in Alzheimer's patients, particularly regarding whether they are accompanied by repetitive movements or disorientation, is crucial to effectively rule out the influence of individual preferences on gait disturbances and cognitive behavior monitoring.

[0127] Typically, when Alzheimer's patients turn, they usually slow down their pace to ensure stability and control their balance, avoiding falls or loss of balance due to turning too quickly. Once they stabilize, their gait gradually returns to its original rhythm.

[0128] Based on the differences in disorder between adjacent gaits within different clusters, gait turning characteristics for each cluster are obtained. In an exemplary embodiment, such as... Figure 6 As shown, the process of obtaining gait turning features includes:

[0129] Step 4-1: Obtain the first difference between the first and second gait motion disorder factors in the fitted curve of the target cluster.

[0130] For ease of explanation, let's define the first gait as any gait other than the first and last gait in the fitted curve of the target cluster, the second gait as the gait following the first gait, and the third gait as the gait preceding the first gait. Therefore, the first, second, and third gaits are three adjacent gaits, and their order in the fitted curve of the target cluster is as follows: third gait, first gait, and second gait.

[0131] Since it is necessary to obtain the gait turning features of a certain gait based on a certain gait and its preceding and following gait, and since the first gait has no preceding gait and the last gait has no following gait, the gait turning features of the first and last gait are no longer obtained. Only the gait turning features of all other gaits are obtained.

[0132] The difference between the first-step and second-step motion disorder factors in the fitted curve of the target cluster is defined as the first difference. Specifically, the first difference is the absolute value of the difference between the motion disorder factors of the first-step and the second-step motion disorder factors.

[0133] set up, The first line in the fitted curve representing the target cluster Gait, The first line in the fitted curve representing the target cluster Gait, The first line in the fitted curve representing the target cluster Gait, Indicates the first A gait movement disorder factor, Representing the A gait movement disorder factor, Representing the The first gait disturbance factor. Then, the second... The gait and the first The first difference between the gaits is: .

[0134] Step 4-2: Obtain the second difference in the motor disorder factors between the second and third gait.

[0135] The difference between the motion disorder factors of the second and third gait in the fitted curve of the target cluster is defined as the second difference. Specifically, the second difference is the absolute value of the difference between the motion disorder factors of the second and third gait. Then, the... The gait and the first The second difference in gait is: .

[0136] Step 4-3: Based on the first and second differences, obtain the gait turning characteristics of the first gait.

[0137] The greater the first difference, the more likely Alzheimer's patients are to engage in activities corresponding to the target cluster. The greater the difference in motor disorder factors between the first gait and the next gait, and the greater the difference in the first gait... The smaller the difference in the dysregulation factor between the previous gait and the next gait, the better the dysregulation factor. The greater the likelihood of a gait turning, the more pronounced the gait turning characteristic. This typically means that Alzheimer's patients slow their pace to ensure stability during turning movements, thereby effectively controlling their balance and reducing the risk of falls. As Alzheimer's patients stabilize, their gait gradually returns to a normal rhythm. Therefore, the gait turning characteristic is directly proportional to the first difference and inversely proportional to the second difference.

[0138] In an exemplary embodiment, the following is given: A specific method for calculating the gait turning characteristics of a gait:

[0139] ;

[0140] in, Indicates the first Gait turning features for each gait. The normalization method here is as follows: obtain the difference between the first and second differences for each gait in the fitted curve of the target cluster, then obtain the maximum and minimum differences, and finally normalize according to the maximum and minimum values. Normalize.

[0141] The above method is used to obtain the gait turning characteristics of each gait in the fitted curve of the target cluster.

[0142] Then, the average value of the gait turning features in the fitted curve of the target cluster (here referring to all gaits except the first and last gait in the fitted curve of the target cluster) is calculated as the gait turning feature of the target cluster. The larger the value of the gait turning feature of the target cluster, the higher the frequency of gait turning in Alzheimer's patients under the abnormal behavior pattern corresponding to the target cluster, meaning that Alzheimer's patients frequently exhibit turning movements when performing the abnormal behavior pattern corresponding to the target cluster.

[0143] Using the above method, the gait turning characteristics of each cluster are obtained.

[0144] Step 5: Integrate the gait disorder and gait turning characteristics of various clusters, and combine them with the time intervals between adjacent clusters to determine cognitive impairment behavioral indicators for Alzheimer's patients and provide early warnings.

[0145] Based on steps 1 to 4, gait disorder and gait turning characteristics of each cluster are obtained. By fusing the gait disorder and gait turning characteristics of each cluster and combining them with the time intervals between adjacent clusters, cognitive impairment behavioral indicators for Alzheimer's disease patients are determined, and early warnings are issued.

[0146] In one exemplary embodiment, such as Figure 7 As shown, the process of obtaining cognitive impairment behavioral indicators for Alzheimer's disease patients includes:

[0147] Step 5-1: Based on the gait disorder characteristics and gait turning characteristics of each type of cluster, obtain the comprehensive gait disorder characteristics of each type of cluster.

[0148] For each target cluster, the product of its gait disorder feature and its gait turning feature is calculated; this product serves as the comprehensive gait disorder feature for that cluster. This method is then used to obtain the comprehensive gait disorder features for each cluster. The stronger the comprehensive gait disorder feature for each cluster, the more pronounced the cognitive impairment behaviors in Alzheimer's patients. Therefore, the cognitive impairment behavior component is directly proportional to the comprehensive gait disorder feature.

[0149] Step 5-2: Based on the gait disorder characteristics and temporal distribution characteristics of various clusters, obtain the cognitive impairment behavioral index components of various clusters.

[0150] First, the temporal distribution characteristics of each cluster are obtained based on the time intervals between adjacent clusters. In an exemplary embodiment, such as... Figure 8 As shown, a specific process for obtaining time distribution characteristics includes:

[0151] Step 5-2-1: Obtain the intermediate time for each type of cluster.

[0152] In an exemplary embodiment, the midpoint of the target cluster's fitting curve is obtained. Each gait in the target cluster's fitting curve has a corresponding sampling time. Therefore, the midpoint of the target cluster's fitting curve, i.e., the time corresponding to the midpoint of the horizontal axis of the target cluster's fitting curve, is obtained as the midpoint of the target cluster. This yields the midpoint of each cluster.

[0153] Step 5-2-2: Obtain the time interval between the intermediate time of each cluster and the previous intermediate time, as the time distribution feature of each cluster.

[0154] The intermediate moments of each cluster are sorted chronologically to obtain an intermediate moment time series. The time interval between the intermediate moment of the target cluster and the previous intermediate moment is obtained within this intermediate moment time series, and this time interval is used as the temporal distribution feature of the target cluster. It should be understood that for the first intermediate moment in this intermediate moment time series, the time interval between the first intermediate moment and the moment when the gait parameters of the first gait of the Alzheimer's patient are obtained (i.e., the start time of the monitoring period mentioned above) is used as the temporal distribution feature of the cluster corresponding to the first intermediate moment. Using the above method, the temporal distribution features of each cluster are obtained.

[0155] The shorter the time interval, the more concentrated the temporal distribution of cognitive impairment in Alzheimer's patients, meaning the more obvious the cognitive impairment in Alzheimer's patients. Therefore, cognitive impairment behavioral indicators are inversely proportional to temporal distribution characteristics.

[0156] Based on the gait disturbance characteristics and temporal distribution characteristics of the target cluster, the cognitive impairment behavioral index components of the target cluster are obtained. In an exemplary embodiment, the calculation formula for the cognitive impairment behavioral index components is as follows:

[0157] ;

[0158] in, This represents the cognitive impairment behavioral index component of the h-th cluster. This represents the gait disorder characteristics of the h-th cluster. This represents the temporal distribution characteristics of the h-th cluster.

[0159] Indicates to Negative correlation normalization. The negative correlation normalization method here can be: obtain the maximum and minimum values ​​in the time distribution features of each cluster, and use the maximum and minimum value normalization method to... Normalization is performed, and finally the value 1 is compared with the normalized value. Subtracting them, the result is the pair. The negative correlation normalization.

[0160] Using the above process, the cognitive impairment behavior index components of each cluster are obtained.

[0161] Step 5-3: Integrate the cognitive impairment behavior index components of various clusters to obtain the cognitive impairment behavior index of Alzheimer's disease patients.

[0162] The average value of the cognitive impairment behavioral index components for each cluster was calculated as a cognitive impairment behavioral index for Alzheimer's disease patients. A higher value for the cognitive impairment behavioral index indicates a smaller degree and time span of cognitive impairment in Alzheimer's patients across all clusters, reflecting more severe cognitive impairment and a more concentrated temporal distribution. This suggests that the cognitive and motor functions of Alzheimer's patients may have experienced significant decline in a short period, posing a greater risk.

[0163] Finally, an early warning is issued based on cognitive impairment behavioral indicators of Alzheimer's patients. In an exemplary embodiment, a preset early warning threshold is established, with a value ranging from 0 to 1. The specific value is set according to actual judgment needs. If a safer early warning logic is required, the preset early warning threshold can be set slightly lower, making it easier for cognitive impairment behavioral indicators that meet the early warning requirements to appear. In this embodiment, a preset early warning threshold of 0.6 is used as an example. Then, the cognitive impairment behavioral indicators of Alzheimer's patients are compared with the preset early warning threshold. If the indicator is greater than or equal to the preset early warning threshold, an early warning signal is output, notifying relevant personnel to pay close attention to the physical condition of the Alzheimer's patient.

[0164] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0165] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease, characterized in that, include: Gait parameters of various gait patterns in Alzheimer's patients were obtained, and abnormal gait patterns were identified based on these parameters. Based on the distance metric between any two abnormal gaits, the abnormal gaits are clustered to obtain multiple clusters; Gait disorder features are obtained for various clusters, and these features are used to characterize the disorder of gait within a cluster. Based on the differences in disorder between adjacent gaits in various clusters, gait turning characteristics of various clusters are obtained; By integrating gait disorder and gait turning characteristics of various clusters and combining the time intervals between adjacent clusters, cognitive impairment behavioral indicators of Alzheimer's disease patients can be determined and early warnings can be issued. The process of obtaining the distance metric between any two aberrant gaits includes: Obtain the differences in abnormal feature indicators and time intervals between any two abnormal gaits; The distance metric is obtained based on the differences in anomaly indicator values ​​and the time interval; the distance metric is directly proportional to the differences in anomaly indicator values ​​and the time interval. Obtain gait disorder features for various clusters, including: The abnormal gait patterns in the target cluster are sorted chronologically, and the fitting curves of the motion disorder factors of each abnormal gait pattern are obtained; the target cluster can be any cluster. Obtain the variation amplitude of the motion disorder factor for each gait in the fitted curve; Based on the variation amplitude of the motion disorder factor of each gait in the fitted curve, the gait disorder characteristics of the target cluster are obtained; The process of obtaining gait turning features includes: Obtain the first difference between the motion disorder factors of the first and second gait in the fitted curve of the target cluster; the first gait is any gait other than the first and last gait in the fitted curve, and the second gait is the gait following the first gait. Obtain the second difference in the motor disorder factors between the second and third gait; the third gait is the gait preceding the first gait. Based on the first and second differences, the gait turning characteristics of the first step are obtained; the gait turning characteristics are directly proportional to the first difference and inversely proportional to the second difference.

2. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 1, characterized in that, Gait parameters include: gait speed, maximum pitch angle when the toes leave the ground, and maximum pitch angle when the heels strike the ground.

3. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 2, characterized in that, Based on gait parameters, abnormal gaits are identified, including: The gait movement disorder factor is obtained based on the difference in the maximum pitch angle when the toes of both feet leave the ground and the difference in the maximum pitch angle when the heels of both feet land. Based on the gait disorder factor and gait speed, abnormal characteristic indicators of gait are obtained; the abnormal characteristic indicators are directly proportional to the gait disorder factor and inversely proportional to the gait speed. Abnormal gait patterns are identified by screening based on abnormal gait characteristic indicators.

4. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 1, characterized in that, The gait disorder features of the target cluster are obtained, including: Slide the window through the fitted curve according to the preset window size to obtain the data window corresponding to each gait in the fitted curve; The window disorder coefficient of the data window is obtained based on the sub-disorder coefficients of each gait in the data window; the sub-disorder coefficient of the gait is obtained by the change amplitude of the gait and the frequency proportion of the gait motion disorder factor in the fitted curve. By fusing the window disorder coefficients of all data windows in the target cluster, the gait disorder characteristics of the target cluster are obtained.

5. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 1, characterized in that, By integrating gait disturbance and gait turning characteristics of various clusters and combining the time intervals between adjacent clusters, cognitive impairment behavioral indicators for Alzheimer's disease patients were determined, including: Based on the gait disorder characteristics and gait turning characteristics of various clusters, the comprehensive gait disorder characteristics of various clusters are obtained; Based on the gait disturbance characteristics and temporal distribution characteristics of each cluster, cognitive impairment behavioral index components are obtained for each cluster. The cognitive impairment behavioral index components are directly proportional to the gait disturbance characteristics and inversely proportional to the temporal distribution characteristics. The temporal distribution characteristics are obtained from the time intervals between adjacent clusters. By integrating the cognitive impairment behavioral index components of various clusters, cognitive impairment behavioral indicators for Alzheimer's disease patients are obtained.

6. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 5, characterized in that, The process of obtaining time distribution characteristics includes: Obtain the intermediate time for each type of cluster; The time interval between the intermediate time of each cluster and the previous intermediate time is obtained as the temporal distribution feature of each cluster.

7. The method for monitoring and early warning of cognitive impairment behaviors in Alzheimer's disease as described in claim 1, characterized in that, The early warning process includes comparing the cognitive impairment behavior indicators of Alzheimer's patients with preset early warning thresholds. If the indicators are greater than or equal to the preset early warning thresholds, an early warning signal is output.

Images