Disability level analysis method for the elderly based on millimeter wave radar and related products

Abstract

The application relates to the technical field of intelligent decision-making, and discloses a disability level analysis method based on a millimeter wave radar and related products, which comprises the following steps: collecting a radar echo signal, converting the radar echo signal into point cloud data; performing vital sign analysis on a target user according to the point cloud data to obtain vital sign parameters; extracting in-bed state features, activity ability features and out-of-bed behavior features of the target user according to the vital sign parameters; performing initial disability level analysis on the target user according to the features to obtain an initial disability level; and performing rule correction on the initial disability level according to preset out-of-bed assistance information to obtain a target disability level. The in-bed state features, the activity ability features and the out-of-bed behavior features are used to analyze the initial disability level of the target user, so that the normativeness and consistency of disability level evaluation are improved; meanwhile, the initial disability level is corrected according to rules, so that the precision and reliability of the disability level analysis are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent decision-making technology, and in particular to a method for analyzing the disability level of elderly people based on millimeter-wave radar and related products. Background Technology

[0002] To promote the high-quality development of the elderly care industry, meet the diverse living and health security needs of the elderly, and improve the digitalization and intelligence of elderly care services, accurate, objective, and quantitative assessment of disability levels has become a key technical requirement for the standardized implementation of intelligent elderly care and long-term care insurance.

[0003] Currently, traditional methods for assessing the level of disability in elderly people employ video surveillance and behavioral observation techniques. This involves installing cameras in the home environment to record and analyze the daily activities of disabled individuals. However, this method still relies on manual interpretation during the actual assessment process, making it highly subjective and difficult to establish unified, quantifiable evaluation standards, resulting in inaccurate level analysis results. On the other hand, some traditional methods utilize wearable devices for continuous monitoring, collecting data such as heart rate, activity levels, and sleep patterns to track and analyze the status of disabled individuals over multiple days. However, wearable devices are easily affected by human removal, misuse, or environmental interference, making it difficult to guarantee data stability and authenticity, thus leading to insufficient accuracy in the final level analysis results.

[0004] Therefore, in the face of the growing demand for disability grading analysis of elderly people, the current methods for disability grading analysis urgently need to be improved in order to solve the problem of insufficient accuracy of the results of existing methods. Summary of the Invention

[0005] This invention provides a method and related products for analyzing the disability level of elderly people based on millimeter-wave radar, with the main purpose of solving the problem of insufficient accuracy in the analysis results of disability level of elderly people.

[0006] To achieve the above objectives, this invention provides a method for analyzing the disability level of elderly people based on millimeter-wave radar, comprising: Collect the radar echo signal returned after the millimeter-wave radar transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user. Vital signs analysis of the target user is performed based on point cloud data to obtain vital sign parameters; Extract the target user's in-bed status characteristics, activity level characteristics, and out-of-bed behavior characteristics based on vital sign parameters; Based on the characteristics of bedridden status, activity level, and out-of-bed behavior, the initial disability level of the target user is analyzed to obtain the initial disability level. The initial disability level is corrected according to the preset assisted bed exit information to obtain the target disability level.

[0007] The present invention also provides a device for analyzing the disability level of elderly people based on millimeter-wave radar, the device comprising: The point cloud data conversion module is used to collect the radar echo signal returned by the millimeter-wave radar after it transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user. The vital signs parameter analysis module is used to analyze the vital signs of the target user based on point cloud data and obtain vital signs parameters. The in-bed activity and out-of-bed feature extraction module is used to extract the in-bed status features, activity ability features, and out-of-bed behavior features of the target user based on vital sign parameters. The initial disability level analysis module is used to analyze the initial disability level of the target user based on the characteristics of bed status, activity ability, and out-of-bed behavior, and to obtain the initial disability level. The target disability level correction module is used to correct the initial disability level according to preset assisted bed leaving information, so as to obtain the target disability level.

[0008] The present invention also provides an electronic device comprising: At least one processor; and, A memory that is communicatively connected to at least one processor; wherein, The memory stores a computer program that can be executed by at least one processor, which enables the at least one processor to perform the above-described method for analyzing the disability level of elderly people based on millimeter-wave radar.

[0009] The present invention also provides a computer-readable medium storing at least one computer program, which is executed by a processor in an electronic device to implement the above-described method for analyzing the disability level of elderly people based on millimeter-wave radar.

[0010] In this embodiment of the invention, radar echo signals are collected and converted into point cloud data, enabling non-contact data acquisition without the need for wearable devices, thus avoiding data distortion caused by manual removal or misuse. Vital signs analysis of the target user is performed based on the point cloud data, replacing manual interpretation, reducing subjective bias, and improving the accuracy of assessing the target's disability level. By extracting the target user's in-bed status characteristics, activity ability characteristics, and out-of-bed behavior characteristics, the problem of inconsistent evaluation standards is avoided, improving the completeness of the assessment basis. Initial disability level analysis of the target user is performed based on these in-bed status characteristics, activity ability characteristics, and out-of-bed behavior characteristics, eliminating the subjectivity of manual interpretation and improving the standardization and consistency of disability level assessment. The initial disability level is corrected according to preset assisted out-of-bed information, and the judgment results are further optimized through rule constraints, significantly improving the accuracy and reliability of disability level analysis. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a method for analyzing the disability level of elderly people based on millimeter-wave radar, provided in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the process of analyzing the vital signs of a target user based on point cloud data to obtain vital sign parameters, as provided in an embodiment of the present invention. Figure 3 A functional block diagram of a disability level analysis device based on millimeter-wave radar provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of an electronic device for implementing a method for analyzing the disability level of elderly people based on millimeter-wave radar, according to an embodiment of the present invention.

[0012] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0013] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0014] To address the problem that existing millimeter-wave radar-based methods for analyzing the disability levels of elderly people are susceptible to damage from human removal, misuse of equipment, or environmental interference, making it difficult to guarantee data stability and authenticity, and thus resulting in insufficient accuracy of the final level analysis results, this application provides a millimeter-wave radar-based method for analyzing the disability levels of elderly people. This method improves the standardization and consistency of disability level assessment by performing initial disability level analysis on the target user based on bed status characteristics, activity ability characteristics, and out-of-bed behavior characteristics; at the same time, it performs rule correction on the initial disability level, significantly improving the accuracy and reliability of disability level analysis.

[0015] Reference Figure 1The diagram shown is a flowchart illustrating a method for analyzing the disability level of elderly people based on millimeter-wave radar, according to an embodiment of this application. In this embodiment, the method for analyzing the disability level of elderly people based on millimeter-wave radar includes: S1. Collect the radar echo signal returned after the millimeter-wave radar transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user.

[0016] In this embodiment of the invention, radar echo signal refers to the original signal collected by the receiving antenna after the electromagnetic waves emitted by the millimeter-wave radar are reflected by the target user's chest, abdomen, torso, or other parts; point cloud data refers to a discrete point data set consisting of three-dimensional spatial coordinates (and attributes such as velocity and energy) of multiple reflection points on the surface of the target user's body parts, used to characterize the user's position, posture, and motion state, and the point cloud data contains multiple point data.

[0017] In this embodiment of the invention, the millimeter-wave radar continuously acquires radar echo signals according to a fixed frame period. Each time frame corresponds to a complete acquisition process, that is, one frame of radar echo signal is acquired in each frame period.

[0018] In this embodiment of the invention, converting radar echo signals into point cloud data characterizing the state features of target users includes: performing a fast time-dimensional Fast Fourier Transform on the radar echo signals to obtain a first transformed signal; performing range-dimensional coherent accumulation processing on the first transformed signal to obtain a second transformed signal; performing a slow time-dimensional Fast Fourier Transform on the second transformed signal to obtain a third transformed signal; extracting the range index and Doppler index corresponding to the target reflection points in the third transformed signal; performing beamforming processing on the target reflection points based on the range index and Doppler index to obtain reflection point angle information; and calculating the spatial coordinates of the target reflection points based on the range index and reflection point angle information to generate point cloud data composed of multiple target reflection points.

[0019] In this embodiment of the invention, the target reflection point is a reflection location point from various local areas (such as chest cavity, arms, torso, legs, etc.) on the surface of the target user's body, and there may be one or more target reflection points.

[0020] In this embodiment of the invention, during the fast time-dimensional fast Fourier transform of the radar echo signal, the millimeter-wave radar samples the radar echo signal at a preset time interval within each linear frequency modulation cycle, thereby obtaining a set of one-dimensional time-domain signals arranged in time order within a single linear frequency modulation cycle; then, the one-dimensional time-domain signal corresponding to each linear frequency modulation cycle and each receiving antenna is converted into a one-dimensional frequency-domain signal, and all one-dimensional frequency-domain signals are stacked in a fixed dimensional order to obtain a three-dimensional complex matrix containing the range dimension (range cell), the frequency modulation cycle dimension (i.e., the nth linear frequency modulation cycle), and the antenna dimension, and the three-dimensional complex matrix is ​​used as the first transform signal.

[0021] Furthermore, the signals belonging to the same linear frequency modulation period and adjacent different range cells in the range dimension of the first transformed signal are coherently superimposed to obtain the second transformed signal; the signals in the second transformed signal that are in the same range cell and the same antenna are transformed to the frequency domain. After the transformation, each frequency component in the frequency domain corresponds to a Doppler frequency; thus, the second transformed signal is converted into a three-dimensional matrix containing the range dimension, the Doppler dimension, and the antenna dimension, which is the third transformed signal; in the third transformed signal, the range dimension represents the distance between the target reflection point and the radar, and the Doppler dimension represents the radial velocity of the target reflection point.

[0022] Furthermore, the signal power of each range-Doppler dimension in the third transformed signal is calculated, and a signal power spectrum is generated based on the signal power. Signals whose signal power exceeds a preset power threshold in the signal power spectrum are identified as target reflection points. The range index and Doppler index corresponding to the target reflection point are extracted from the three-dimensional matrix. The signals under all antennas at the range index and the Doppler index are read out. The source direction of the read signal is calculated using the phase difference between these different antennas. The azimuth and elevation angles of the target reflection point are generated based on the source direction. The azimuth and elevation angles are combined to form the reflection point angle information.

[0023] Furthermore, for each target reflection point, the actual distance from the target reflection point to the millimeter-wave radar is calculated based on the distance index. Taking the millimeter-wave radar as the origin, the three-dimensional spatial coordinates of the target reflection point are calculated using trigonometric functions based on the reflection point angle information and the actual distance. The three-dimensional spatial coordinates of all target reflection points are combined into point cloud data.

[0024] In this embodiment of the invention, by performing a fast time-dimensional fast Fourier transform on the radar echo signal, the time-domain echo is converted into a frequency-domain signal, enabling the separation and measurement of target distance. By performing range-dimensional coherent accumulation processing on the first transformed signal, the signal-to-noise ratio is significantly improved and noise is suppressed, making the target reflection more prominent and facilitating subsequent detection. By performing a slow time-dimensional fast Fourier transform on the second transformed signal, the Doppler frequency is extracted from the time-series changes to obtain the target motion / micro-motion velocity information, achieving joint resolution of distance and velocity. By calculating the spatial coordinates of the target reflection point based on the distance index and reflection point angle information, the radar signal is converted into an intuitive spatial point cloud, fully representing the human posture and spatial distribution.

[0025] S2. Analyze the vital signs of the target user based on the point cloud data to obtain vital sign parameters.

[0026] In this embodiment of the invention, vital signs parameters refer to basic quantitative indicators extracted from point cloud data to characterize the physical activities and state changes of the target user, including physical movement status, bed occupancy status, and physical movement energy value (i.e., physical movement intensity). Physical movement status includes having obvious physical movement and not having obvious physical movement; bed occupancy status includes being in bed and being out of bed.

[0027] like Figure 2 As shown, in this embodiment of the invention, vital sign analysis of a target user is performed based on point cloud data to obtain vital sign parameters, including: S21, filtering the point cloud data to obtain filtered target point cloud data; S22, extracting the main direction features and spatial dispersion features of the target point cloud data, and combining the main direction features and spatial dispersion features into a point cloud feature vector; S23, calculating the body motion energy value of the target user based on the point cloud feature vector; S24, identifying the body motion state of the target user based on the body motion energy value and a preset body motion threshold; S25, identifying the bed occupancy status of the target user, and combining the body motion energy value, body motion state, and bed occupancy status into vital sign parameters.

[0028] In this embodiment of the invention, the point cloud density of each point data in the point cloud data is calculated within a preset neighborhood. Point data with excessively low density are removed based on the point cloud density, and the remaining point data are used as the filtered target point cloud data. Principal component analysis is performed on the target point cloud data to obtain the three-dimensional covariance matrix of the point cloud data. Eigenvalue decomposition is performed on the three-dimensional covariance matrix to obtain three eigenvalues ​​and their corresponding eigenvectors. The direction of the eigenvector corresponding to the largest eigenvalue is selected as the principal direction (principal direction feature). The variance of the point cloud data along the principal axis of PCA, the minimum bounding box size of the point cloud data, and the overall density of the point cloud data are calculated. The variance along the principal axis, the minimum bounding box size, and the overall density are combined into a spatial scatter feature. The principal direction feature is encoded into a principal direction encoded value, and then concatenated with the variance along the principal axis, the minimum bounding box size, and the overall density to form a point cloud feature vector.

[0029] In this context, the PCA principal axis refers to the main direction vector obtained in principal component analysis; the minimum bounding box size refers to the length, width, and height of the smallest cuboid that completely contains all point cloud data; eigenvalue decomposition of the three-dimensional covariance matrix yields three orthogonal principal axes: the first principal axis corresponds to the largest eigenvalue, representing the longest extension direction of the point cloud distribution; the second principal axis corresponds to the second largest eigenvalue, and its direction is orthogonal to the first principal axis; the third principal axis corresponds to the smallest eigenvalue, and its direction is orthogonal to the first two.

[0030] Furthermore, the point cloud feature vector of the previous time frame is obtained, and the difference between the two is calculated using the Euclidean distance method. The difference is used as the body motion energy value to obtain the body motion energy value (i.e., body motion intensity) of each time frame.

[0031] For example, if the feature vector of the point cloud in the current time frame is [1, 2, 3] and the feature vector of the point cloud in the previous time frame is [4, 6, 8], and the difference between the two is 7.07, then the kinetic energy value corresponding to the feature vector of the point cloud in the current frame is 7.07.

[0032] Furthermore, the body movement energy value is compared with a preset body movement threshold. If the body movement energy value is greater than or equal to the body movement threshold, it is determined that the target user has obvious body movement (turning over, getting up, raising hands, etc.) in the current time frame; if the body movement energy value is less than the body movement threshold, it is determined that the target user has no obvious body movement (resting state) in the current time frame.

[0033] In this embodiment of the invention, identifying the bed occupancy status of a target user includes: determining the rectangular area where the target user's bed is located based on the installation location of the millimeter-wave radar; generating a bed area mask containing boundary coordinate values ​​based on the rectangular area; counting the number of reflection points of the target reflection points located within the bed area mask; calculating the percentage of points inside the bed based on the number of reflection points; comparing the percentage of points inside the bed with a preset bed occupancy threshold; and determining the bed occupancy status of the target user based on the comparison result.

[0034] In this embodiment of the invention, based on the actual installation method of the millimeter-wave radar and the preset spatial coordinate system, the four corner points of the target user's bed are determined by automatic detection, and the four corner points are connected to form a rectangular area; then, based on the three-dimensional coordinates of the four corner points in the rectangular area, the minimum and maximum values ​​of x, y, and z are extracted respectively, and a bed area mask is generated based on the extracted minimum and maximum values.

[0035] For example, the three-dimensional coordinates of the four corner points of the bed are as follows: corner point 1: (1.0, 2.0, 0.0), corner point 2: (1.0, 3.0, 0.0), corner point 3: (2.5, 3.0, 0.0), and corner point 4: (2.5, 2.0, 0.0). The minimum and maximum values ​​of x, y, and z are extracted respectively, and the bed area mask is {x∈[1.0, 2.5], y∈[2.0, 3.0], z∈[0.0, 0.0]}.

[0036] Furthermore, it is determined whether the three-dimensional spatial coordinates of each target reflection point fall inside the bed area mask. The number of target reflection points that fall inside the bed area mask is counted. The number of target reflection points that fall inside the bed area mask is divided by the total number of target reflection points to obtain the percentage of points inside the bed.

[0037] Furthermore, the percentage of points inside the bed is compared with a preset in-bed threshold. If the percentage of points inside the bed is greater than or equal to the preset in-bed threshold, the bed occupancy status in the current time frame is determined to be in bed. If the percentage of points inside the bed is less than the preset in-bed threshold, the bed occupancy status in the current time frame is determined to be out of bed.

[0038] In this embodiment of the invention, point cloud data is filtered to remove noise points and abnormal reflection points, thereby improving the purity of the point cloud and ensuring the accuracy of subsequent feature extraction and body motion calculation. Body motion energy values ​​are calculated based on point cloud feature vectors, which intuitively reflect the intensity of micro-movements or large-amplitude movements. By identifying the body motion state of the target user in each time frame, effective interference markers are provided for subsequent vital sign extraction. The vital sign parameters of the target user are determined based on the body motion state, and parameters such as respiration and heart rate are extracted during periods without stable body motion, eliminating body motion interference and improving the reliability of vital sign results.

[0039] S3. Extract the target user's in-bed status characteristics, activity level characteristics, and out-of-bed behavior characteristics based on vital sign parameters.

[0040] In this embodiment of the invention, the following steps are taken to extract the bed-occupancy characteristics, activity capacity characteristics, and bed-leaving behavior characteristics of the target user based on vital sign parameters: determining bed-occupancy segments and their durations, as well as bed-leaving segments, based on the bed occupancy status in the vital sign parameters; calculating the target user's total bed-occupancy time and the number of short-term bed-occupancy sessions during the day based on the duration of the bed-occupancy segments, and combining the total bed-occupancy time and the number of short-term bed-occupancy sessions during the day into bed-occupancy characteristics; calculating the target user's average physical activity intensity and the ratio of the diurnal average physical activity intensity based on the physical activity energy value in the vital sign parameters, and combining the average physical activity intensity and the diurnal average physical activity ratio into activity capacity characteristics; and calculating the target user's total number of bed-leaving sessions and the number of bed-leaving sessions during the day based on the duration of the bed-occupancy segments and the bed-leaving segments, and combining the total number of bed-leaving sessions and the number of bed-leaving sessions during the day into bed-leaving behavior characteristics.

[0041] In this embodiment of the invention, the bed occupancy status of vital sign parameters in all time frames within a preset time period (such as one day) is read. When multiple consecutive frames have the same status and there is no interruption, they are merged into a continuous segment. If the bed occupancy status is "in bed", it is called the "in bed segment". The start timestamp and end timestamp of the "in bed segment" are obtained, and the end timestamp is subtracted from the start timestamp to calculate the duration of the "in bed segment". Similarly, the "out bed segment" and the duration of the "out bed segment" can be obtained.

[0042] Furthermore, the durations of all in-bed segments within the preset time period are summed to obtain the total in-bed time of the target user within the preset time period; based on the preset daytime time period (e.g., 8 am to 8 pm) and short-time threshold (e.g., continuous in-bed time not exceeding 30 minutes), in-bed segments that are within the daytime time period and whose duration does not exceed the short-time threshold are selected from the in-bed segments, and the number of selected in-bed segments is taken as the number of short-time in-bed times during the day.

[0043] Furthermore, the average value of body movement energy values ​​of all time frames within the preset time period is analyzed to obtain the average body movement intensity. Then, the daytime period and nighttime period are preset (e.g., 6:00-18:00 during the day and 18:00-6:00 the next day at night). The average value of body movement energy values ​​during the daytime period and the average value of body movement energy values ​​during the nighttime period are calculated respectively. Finally, the average value of body movement energy values ​​during the daytime period is divided by the average value of body movement energy values ​​during the nighttime period to obtain the ratio of the average daytime and nighttime body movement values.

[0044] Optionally, the body motion energy value of each time frame can be compared with a preset intensity threshold to determine the number of times the body motion energy value exceeds the intensity threshold, thus obtaining the number of large movements. The variance of the body motion energy value during the daytime period can be calculated based on the average value of the body motion energy value during the nighttime period, and the variance of the body motion energy value during the nighttime period can be calculated based on the average value of the body motion energy value during the nighttime period. The daytime body motion energy value variance ratio can be obtained by dividing the variance of the body motion energy value during the daytime period by the variance of the body motion energy value during the nighttime period.

[0045] In this embodiment of the invention, the total number of bed-leaving segments is counted as the total number of bed-leavings, and the number of bed-leaving segments whose starting time falls within the daytime period is counted to obtain the daytime bed-leaving count.

[0046] Optionally, the system can also count the first number of times the duration of the bed-staying segment exceeds a preset first time threshold, and use the counted first number as the number of times the bed-staying segment lasts for a long time and then the second number of times the duration of the bed-staying segment does not exceed a preset second time threshold, and use the second number as the number of times the bed-staying segment lasts for a short time and then the number of times the bed-staying segment lasts for a short time.

[0047] In this embodiment of the invention, by determining the duration of in-bed segments and out-of-bed segments, continuous in-bed and out-of-bed periods are accurately divided, providing a reliable temporal basis for subsequent duration and frequency statistics and avoiding confusion in feature calculations. By combining the total in-bed time and the number of short-term in-bed visits within a single day for the target user, and then combining these into in-bed status features, the user's bed rest patterns and stability are quantified, objectively reflecting their bed rest habits and improving the accuracy of in-bed status assessment. The average physical activity intensity and the ratio of the diurnal average physical activity intensity of the target users are calculated, and these two ratios are combined to form an activity capacity characteristic. This characterizes the activity level from both overall intensity and diurnal rhythm perspectives, objectively reflecting the user's mobility and daily routine, and reducing the bias of a single indicator. By statistically analyzing the total number of times the target users get out of bed and the number of times they get out of bed during the day, and combining these two numbers to form an out-of-bed behavior characteristic, the frequency of users getting out of bed and their daytime activity level are quantified, providing intuitive behavioral evidence for disability level assessment and making the evaluation more comprehensive.

[0048] S4. Based on the characteristics of bedridden status, activity level, and out-of-bed behavior, conduct an initial disability level analysis on the target user to obtain the initial disability level.

[0049] In this embodiment of the invention, the initial disability level includes the viability level, the semi-disability level, and the severe disability level. The initial disability level refers to the initial disability level within a preset time period.

[0050] In this embodiment of the invention, an initial disability level analysis is performed on the target user based on the characteristics of bed status, activity ability, and out-of-bed behavior to obtain the initial disability level. This includes: normalizing the characteristics of bed status, activity ability, and out-of-bed behavior to generate normalized feature vectors; and calculating the predicted probability value of the target user belonging to each preset disability level based on the normalized feature vectors. The initial disability level is determined based on the predicted probability values ​​of each preset disability level obtained from the calculation.

[0051] In this embodiment of the invention, a normalization method is used to normalize each feature in the bed-state feature, activity ability feature, and out-of-bed behavior feature, and each normalized feature is concatenated into a normalized feature vector.

[0052] Furthermore, the normalized feature vector is input into the pre-trained classification model. The classification model will output a probability distribution vector corresponding to each feature vector in the normalized feature vector. That is, each vector in the probability distribution vector represents the predicted probability value of the target user belonging to each disability level. The highest predicted probability value is selected from the predicted probability values, and the disability level corresponding to the highest predicted probability value is determined. The determined disability level is used as the initial disability level.

[0053] The pre-trained classification model has learned the mapping relationship between features and disability levels (such as viable, semi-disabled, and severely disabled); the classification model can be a decision tree, random forest, or neural network, etc.

[0054] For example, a feature vector of [0.1, 0.7, 0.2] represents a 10% probability of viability, a 70% probability of partial disability, and a 20% probability of severe disability.

[0055] In this embodiment of the invention, the predicted probability value of the target user belonging to each preset disability level is calculated based on the normalized feature vector, which intuitively reflects the degree of matching between the target user and each disability level; by selecting the disability level corresponding to the maximum predicted probability value as the initial disability level, the objectivity and rationality of disability level identification are ensured.

[0056] S5. Based on the preset assisted bed exit information, the initial disability level is modified according to the rules to obtain the target disability level.

[0057] In this embodiment of the invention, the preset assisted bed-leaving information can also be the number of times the target user is assisted to leave the bed within a preset time period.

[0058] In this embodiment of the invention, the initial disability level is corrected according to preset assisted bed leaving information to obtain a target disability level. This includes: calculating the assisted bed leaving ratio of the target user based on the preset assisted bed leaving information; comparing the assisted bed leaving ratio with a preset level threshold range; if the assisted bed leaving ratio is lower than the lower limit of the level threshold range and the initial disability level is semi-disabled, the initial disability level is corrected to a first corrected level, and the first corrected level is used as the target disability level; if the assisted bed leaving ratio is within the level threshold range, the initial disability level is used as the target disability level; if the assisted bed leaving ratio is higher than the upper limit of the level threshold range and the initial disability level is semi-disabled, the initial disability level is corrected to a second corrected level, and the second corrected level is used as the target disability level.

[0059] In this embodiment of the invention, during the process of calculating the assisted bed leaving ratio of the target user based on the preset assisted bed leaving information, the assisted bed leaving information (i.e., the number of assisted bed leavings) is divided by the total number of bed leavings to obtain the assisted bed leaving ratio.

[0060] Furthermore, the proportion of assisted bed alighting is compared with a preset level threshold range. If the proportion of assisted bed alighting is lower than the lower limit of the level threshold range and the initial disability level is semi-disabled, it indicates that the target user has a strong ability to get out of bed independently, and the initial disability level is corrected to the active level, i.e., the first correction level. If the proportion of assisted bed alighting is lower than the lower limit of the level threshold range and the initial disability level is active, the initial disability level is kept unchanged and is used as the target disability level. If the proportion of assisted bed alighting is lower than the lower limit of the level threshold range and the initial disability level is severe disability, the initial disability level is corrected to the semi-disabled level, i.e., the first correction level.

[0061] Furthermore, if the proportion of assisted bed alighting exceeds the upper limit of the grade threshold range, and the initial disability grade is semi-disabled, the initial disability grade is revised to severe disability, i.e., the second revised grade; if the proportion of assisted bed alighting exceeds the upper limit of the grade threshold range, and the initial disability grade is active, the initial disability grade is revised to semi-disabled, i.e., the second revised grade; if the proportion of assisted bed alighting exceeds the upper limit of the grade threshold range, and the initial disability grade is severe disability, the initial disability grade remains unchanged, and the initial disability grade is used as the second revised grade.

[0062] In one embodiment, a time window of length T (e.g., T=7 or T=30) can be set, and the target disability level for T consecutive days within the set time window can be used as input. The frequency of occurrence of each target disability level within the set time window can be counted, and the disability level with the most occurrences can be selected as the final target disability level.

[0063] In this embodiment of the invention, by calculating the proportion of assisted bed leaving, the degree of assistance received by the user when leaving the bed is quantified, providing an objective behavioral basis for the correction of disability level and improving the accuracy of the target disability level; by comparing the proportion of assisted bed leaving with a preset level threshold range, a standardized judgment rule is established to realize the graded judgment of the degree of assisted bed leaving.

[0064] like Figure 3 The diagram shown is a functional block diagram of a disability level analysis device based on millimeter-wave radar provided in an embodiment of the present invention.

[0065] The present invention provides a millimeter-wave radar-based device 300 for analyzing the disability level of elderly individuals, which can be installed in an electronic device. Depending on the functions implemented, the millimeter-wave radar-based device 300 may include a point cloud data conversion module 301, a vital signs parameter analysis module 302, a bedridden / out-of-bed feature extraction module 303, an initial disability level analysis module 304, and a target disability level correction module 305. The modules of this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0066] In this embodiment, the functions of each module / unit are as follows: The point cloud data conversion module 301 is used to collect the radar echo signal returned by the millimeter-wave radar after it transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user. The vital signs parameter analysis module 302 is used to analyze the vital signs of the target user based on point cloud data to obtain vital signs parameters; The in-bed activity and out-of-bed feature extraction module 303 is used to extract the in-bed status features, activity ability features, and out-of-bed behavior features of the target user based on vital sign parameters. The initial disability level analysis module 304 is used to perform initial disability level analysis on the target user based on the characteristics of bed status, activity ability, and out-of-bed behavior, and to obtain the initial disability level. The target disability level correction module 305 is used to correct the initial disability level according to preset assisted bed leaving information to obtain the target disability level.

[0067] In one embodiment, the point cloud data conversion module 301 is specifically used to perform a fast time-dimensional fast Fourier transform on the radar echo signal to obtain a first transformed signal; perform range-dimensional coherent accumulation processing on the first transformed signal to obtain a second transformed signal; perform a slow time-dimensional fast Fourier transform on the second transformed signal to obtain a third transformed signal; extract the range index and Doppler index corresponding to the target reflection point in the third transformed signal; perform beamforming processing on the target reflection point based on the range index and Doppler index to obtain the reflection point angle information; and calculate the spatial coordinates of the target reflection point based on the range index and the reflection point angle information to generate point cloud data composed of multiple target reflection points.

[0068] In one embodiment, the vital signs parameter analysis module 302 is specifically used to filter point cloud data to obtain filtered target point cloud data; extract the main direction features and spatial dispersion features of the target point cloud data, and combine the main direction features and spatial dispersion features into a point cloud feature vector; calculate the body motion energy value of the target user based on the point cloud feature vector; identify the body motion state of the target user based on the body motion energy value and a preset body motion threshold; identify the bed occupancy status of the target user, and combine the body motion energy value, body motion state, and bed occupancy status into vital signs parameters.

[0069] In one embodiment, the vital signs parameter analysis module 302 is specifically used to determine the rectangular area where the target user's bed is located based on the installation location of the millimeter-wave radar, generate a bed area mask containing boundary coordinate values ​​based on the rectangular area, count the number of reflection points of the target reflection points located within the bed area mask, calculate the bed-in-point ratio based on the number of reflection points, compare the bed-in-point ratio with a preset bed-occupancy threshold, and determine the bed occupancy status of the target user based on the comparison result.

[0070] In one embodiment, the in-bed activity and out-of-bed feature extraction module 303 is specifically used to determine in-bed segments and their durations, and out-of-bed segments based on the bed occupancy status in the vital signs parameters; to calculate the total in-bed time and the number of short-term in-bed visits during the day based on the duration of the in-bed segments, and to combine the total in-bed time and the number of short-term in-bed visits during the day into in-bed status features; to calculate the average physical activity intensity and the ratio of the average physical activity intensity to the average physical activity intensity during the day and night based on the physical activity energy value in the vital signs parameters, and to combine the average physical activity intensity and the ratio of the average physical activity intensity to the average physical activity intensity during the day into activity ability features; and to calculate the total number of times the target user leaves the bed and the number of times the user leaves the bed during the day based on the duration of the in-bed segments and the out-of-bed segments, and to combine the total number of times the user leaves the bed and the number of times the user leaves the bed during the day into out-of-bed behavior features.

[0071] In one embodiment, the initial disability level analysis module 304 is specifically used to normalize the bed status characteristics, activity ability characteristics, and out-of-bed behavior characteristics respectively to generate normalized feature vectors; calculate the predicted probability value of the target user belonging to each preset disability level based on the normalized feature vectors; and determine the initial disability level based on the calculated predicted probability value of each preset disability level.

[0072] In one embodiment, the target disability level correction module 305 is specifically used to calculate the assisted bed leaving ratio of the target user based on preset assisted bed leaving information; compare the assisted bed leaving ratio with a preset level threshold range; if the assisted bed leaving ratio is lower than the lower limit of the level threshold range and the initial disability level is semi-disabled, the initial disability level is corrected to a first correction level and the first correction level is used as the target disability level; if the assisted bed leaving ratio is within the level threshold range, the initial disability level is used as the target disability level; if the assisted bed leaving ratio is higher than the upper limit of the level threshold range and the initial disability level is semi-disabled, the initial disability level is corrected to a second correction level and the second correction level is used as the target disability level.

[0073] In detail, each module in the millimeter-wave radar-based disability level analysis device 300 of the present invention uses the same technical means as the millimeter-wave radar-based disability level analysis method in the accompanying drawings, and can produce the same technical effect, which will not be repeated here.

[0074] like Figure 4 The diagram shown is a schematic representation of an electronic device for implementing a method for analyzing the disability level of elderly people based on millimeter-wave radar, according to an embodiment of the present invention.

[0075] Electronic device 4 may include processor 40, memory 41, communication bus 42 and communication interface 43, and may also include computer programs stored in memory 41 and run on processor 40, such as a disability level analysis program based on millimeter-wave radar.

[0076] In some embodiments, the processor 40 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The memory 41 includes at least one type of readable medium, including flash memory, portable hard drives, multimedia cards, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disks, optical disks, etc. The communication bus 42 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The communication interface 43 is used for communication between the above-mentioned electronic device and other devices, including network interfaces and user interfaces.

[0077] Figure 4 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 4 The structure shown does not constitute a limitation on the electronic device 4, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0078] For example, although not shown, the electronic device may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to at least one processor 40 via a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power sources, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be elaborated further here.

[0079] It should be understood that the embodiments are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0080] Specifically, the specific implementation method of the processor 40 for the above instructions can be found in the description of the relevant steps in the corresponding embodiment of the accompanying drawings, and will not be repeated here.

[0081] Furthermore, if the modules / units integrated in the electronic device 4 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable medium. The computer-readable medium can be volatile or non-volatile. For example, a computer-readable medium may include: any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0082] The present invention also provides a computer-readable medium storing a computer program, which, when executed by a processor, can implement a method for analyzing the disability level of elderly people based on millimeter-wave radar according to any of the above embodiments.

[0083] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0084] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0085] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0086] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0087] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices recited in a system claim may also be implemented by a single unit or device through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for analyzing the disability level of elderly people based on millimeter-wave radar, characterized in that, The method includes: Collect the radar echo signal returned after the millimeter-wave radar transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user. Based on the point cloud data, the target user's vital signs are analyzed to obtain vital sign parameters; Based on the vital signs parameters, extract the target user's in-bed status characteristics, activity ability characteristics, and out-of-bed behavior characteristics; Based on the in-bed status characteristics, the activity ability characteristics, and the out-of-bed behavior characteristics, an initial disability level analysis is performed on the target user to obtain the initial disability level; The initial disability level is corrected according to preset assisted bed exit information to obtain the target disability level; The step of extracting the target user's in-bed status characteristics, activity level characteristics, and out-of-bed behavior characteristics based on the vital sign parameters includes: The bed occupancy status in the vital signs parameters is used to determine the in-bed segment, the duration of the in-bed segment, and the out-of-bed segment; The total time spent in bed and the number of short-term stays in bed during the day are calculated based on the duration of the in-bed segment, and the total time spent in bed and the number of short-term stays in bed during the day are combined to form an in-bed status feature; The average physical activity intensity and the ratio of the day-night average physical activity intensity of the target user are calculated based on the physical activity energy value in the vital signs parameters, and the average physical activity intensity and the ratio of the day-night average physical activity intensity are combined to form an activity ability characteristic; Based on the duration of the in-bed segment and the out-of-bed segment, the total number of times the target user leaves the bed and the number of times they leave the bed during the day are counted, and the total number of times they leave the bed and the number of times they leave the bed during the day are combined to form an out-of-bed behavior feature.

2. The method for analyzing the disability level of elderly people based on millimeter-wave radar as described in claim 1, characterized in that, The step of performing vital sign analysis on the target user based on the point cloud data to obtain vital sign parameters includes: The point cloud data is filtered to obtain filtered target point cloud data; Extract the main direction features and spatial distribution features of the target point cloud data, and combine the main direction features and spatial distribution features into a point cloud feature vector; Calculate the body kinetic energy value of the target user based on the point cloud feature vector; The body movement state of the target user is identified based on the body movement energy value and the preset body movement threshold; Identify the bed occupancy status of the target user, and combine the body energy value, body movement status, and bed occupancy status into vital sign parameters.

3. The method for analyzing the disability level of elderly people based on millimeter-wave radar as described in claim 2, characterized in that, The process of identifying the bed occupancy status of the target user includes: The rectangular area where the target user's bed is located is determined based on the installation location of the millimeter-wave radar, and a bed area mask containing boundary coordinate values ​​is generated based on the rectangular area. The number of reflection points of the target reflection points located within the mask of the bed area is counted, and the percentage of points within the bed is calculated based on the number of reflection points. The percentage of points within the bed is compared with a preset bed occupancy threshold, and the bed occupancy status of the target user is determined based on the comparison result.

4. The method for analyzing the disability level of elderly people based on millimeter-wave radar as described in claim 1, characterized in that, The initial disability level analysis of the target user based on the in-bed status characteristics, the activity ability characteristics, and the out-of-bed behavior characteristics, to obtain the initial disability level, includes: The in-bed status characteristics, the activity ability characteristics, and the out-of-bed behavior characteristics are normalized respectively to generate normalized feature vectors; The predicted probability value of the target user belonging to each preset disability level is calculated based on the normalized feature vector; The initial disability level is determined based on the predicted probability values ​​of each preset disability level obtained from the calculation.

5. The method for analyzing the disability level of elderly people based on millimeter-wave radar as described in claim 1, characterized in that, The step of correcting the initial disability level according to preset assisted bed exit information to obtain the target disability level includes: The assisted bed exit ratio of the target user is calculated based on the preset assisted bed exit information; The assisted bed exit ratio is compared with a preset level threshold range; If the assisted bed exit ratio is lower than the lower limit of the threshold range of the disability level, and the initial disability level is semi-disabled, the initial disability level is corrected to the first corrected level, and the first corrected level is taken as the target disability level. If the assisted bed exit ratio is within the threshold range of the level, then the initial disability level is taken as the target disability level; If the assisted bed exit ratio is higher than the upper limit of the threshold range of the disability level, and the initial disability level is semi-disability, the initial disability level is corrected to the second corrected level, and the second corrected level is taken as the target disability level.

6. The method for analyzing the disability level of elderly people based on millimeter-wave radar as described in claim 1, characterized in that, The step of converting the radar echo signal into point cloud data characterizing the target user's state features includes: The radar echo signal is subjected to a fast time-dimensional fast Fourier transform to obtain the first transformed signal; The first transformed signal is subjected to range-dimensional coherent accumulation processing to obtain the second transformed signal; The second transformed signal is subjected to a slow-time dimension fast Fourier transform to obtain the third transformed signal; Extract the range index and Doppler index corresponding to the target reflection point from the third transformation signal, and perform beamforming processing on the target reflection point based on the range index and Doppler index to obtain the reflection point angle information; Based on the distance index and the reflection point angle information, the spatial coordinates of the target reflection point are calculated to generate point cloud data composed of multiple target reflection points.

7. A device for analyzing the disability level of elderly people based on millimeter-wave radar, characterized in that, The device includes: The point cloud data conversion module is used to collect the radar echo signal returned by the millimeter-wave radar after it transmits electromagnetic waves to the chest area of ​​the target user, and convert the radar echo signal into point cloud data that characterizes the state features of the target user. The vital signs parameter analysis module is used to perform vital signs analysis on the target user based on the point cloud data to obtain vital signs parameters; The in-bed activity and out-of-bed feature extraction module is used to extract the in-bed status features, activity ability features, and out-of-bed behavior features of the target user based on the vital sign parameters. The initial disability level analysis module is used to perform an initial disability level analysis on the target user based on the bedside status characteristics, the activity ability characteristics, and the out-of-bed behavior characteristics, and to obtain the initial disability level. The target disability level correction module is used to correct the initial disability level according to preset assisted bed exit information to obtain the target disability level. The step of extracting the target user's in-bed status characteristics, activity level characteristics, and out-of-bed behavior characteristics based on the vital sign parameters includes: The bed occupancy status in the vital signs parameters is used to determine the in-bed segment, the duration of the in-bed segment, and the out-of-bed segment; The total time spent in bed and the number of short-term stays in bed during the day are calculated based on the duration of the in-bed segment, and the total time spent in bed and the number of short-term stays in bed during the day are combined to form an in-bed status feature; The average physical activity intensity and the ratio of the day-night average physical activity intensity of the target user are calculated based on the physical activity energy value in the vital signs parameters, and the average physical activity intensity and the ratio of the day-night average physical activity intensity are combined to form an activity ability characteristic; Based on the duration of the in-bed segment and the out-of-bed segment, the total number of times the target user leaves the bed and the number of times they leave the bed during the day are counted, and the total number of times they leave the bed and the number of times they leave the bed during the day are combined to form an out-of-bed behavior feature.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the millimeter-wave radar-based method for analyzing the disability level of elderly people as described in any one of claims 1 to 6.

9. A computer-readable medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for analyzing the disability level of elderly people based on millimeter-wave radar as described in any one of claims 1 to 6.

Images