A method and system for evaluating the turning-over nursing behavior of disabled old people based on multi-data fusion

Abstract

The present application relates to a kind of based on the disabled old man turning over nursing behavior evaluation method, system of multi-data fusion, comprising the following steps: S1, obtain the best turning over track for target object pre-stored;S2, before nursing turning over operation, extract initial best life measurement point position;S3, in the process of nursing turning over operation, control radar collects the echo signal of each preset space field and generates corresponding local vital sign confidence distribution map for signal quality assessment, while the posture sequence of the target object in the process of turning over is obtained by the radar;S4, after nursing turning over operation is completed, the point position of the highest confidence of the local vital sign confidence distribution map of each point position is regarded as the chest point position change track of target object, the chest point position change track is fused posture sequence, forms current turning over track;S5, according to the current turning over track and the best turning over track are compared, and output turning over nursing normative evaluation result.

Description

Technical Field

[0001] This invention relates to the field of elderly care for disabled elderly, specifically to a method and system for evaluating the turning and care behaviors of disabled elderly based on multi-data fusion. Background Technology

[0002] Disabled elderly people refer to the elderly population who have lost their basic ability to take care of themselves due to illness or accidents. Among them, hemiplegic elderly people are bedridden for a long time, and continuous monitoring of their vital signs (especially heart rate) is of great significance for preventing sudden cardiovascular events.

[0003] For elderly people with hemiplegia who are bedridden for extended periods, regular turning is a core nursing measure to prevent bedridden complications such as pressure sores, lung infections, and deep vein thrombosis. However, due to abnormal muscle tone, muscle atrophy, and limited joint movement in the affected limb, improper turning techniques (such as dragging the affected limb, turning at inappropriate angles, or turning too quickly) not only fail to prevent complications but may also cause secondary injuries such as joint dislocations, muscle strains, and skin abrasions. Therefore, how to objectively, quantitatively, and in real-time evaluate the quality of turning procedures performed by caregivers is a technical problem that needs to be solved in the field of elderly care quality management.

[0004] The purpose of this invention is to design a method and system for evaluating the turning and care behaviors of disabled elderly people based on multi-data fusion, in order to address the problems existing in the prior art. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a method and system for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion, which can effectively solve at least one of the problems existing in the prior art.

[0006] The technical solution of this invention is: A method for evaluating the turning and care behaviors of disabled elderly people based on multi-data fusion includes the following steps: S1, obtain the pre-stored optimal turning trajectory for the target object; S2, before the nursing turning operation, the radar collects the chest cavity micro-motion signal of the target object in real time and extracts the initial optimal vital measurement point; S3, during the nursing turning operation, the control radar takes the initial optimal vital sign measurement point as the turning starting point, performs multi-point scanning in the preset space area of ​​the optimal turning trajectory, collects the echo signal of each preset space area, performs signal quality assessment, and generates the corresponding local vital sign confidence distribution map. At the same time, the radar obtains the posture sequence of the target object during the turning process. S4. After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is taken as the chest cavity point change trajectory of the target object. The chest cavity point change trajectory is fused with the posture sequence to form the current turning trajectory. S5. Based on the comparison between the current turning trajectory and the optimal turning trajectory, output the evaluation result of the standardization of turning care.

[0007] Furthermore, in step S1, the optimal rolling trajectory is obtained through the following steps: A trackable sensor is placed on the chest cavity surface of the target object at the area corresponding to the heartbeat. During a standard turning operation, the spatial position change of the trackable sensor is continuously tracked by a tracker to form the optimal turning trajectory.

[0008] Furthermore, step S2 includes: performing the following before the nursing repositioning procedure: S21 uses radar to scan around a point in a pre-defined, coarse candidate area for the location of the chest cavity and calculates multiple vital signs. S22, if multiple vital signs are similar, the point is determined to be the initial optimal vital sign measurement point; if multiple vital signs are not similar, the point is shifted and returned to S21.

[0009] Furthermore, by calculating the standard deviation of multiple vital signs, if the standard deviation is less than a preset standard deviation threshold, the multiple vital signs are determined to be similar.

[0010] Further, step S3 includes: S3.1 During the nursing turning operation, the initial point of the optimal turning trajectory is set at the turning starting point, and the control radar performs multi-point scanning within a preset space area with a preset radius, centered on each point of the optimal turning trajectory. S3.2, Evaluate the signal quality of the echo signal in each preset spatial area to obtain the signal quality index of each point, and generate a local vital sign confidence distribution map covering the preset spatial area based on the signal quality index using an interpolation algorithm. S3.3 During the execution of S3.1 and S3.2, the radar simultaneously performs a real-time scan of the target object's entire torso and arranges the posture sequence in chronological order to cover the entire turning process.

[0011] Further, step S4 includes: S4.1 After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is fitted using a smooth interpolation algorithm to generate the chest cavity point change trajectory of the target object. S4.2 Align the trajectory of chest cavity position change with the posture sequence on the time axis, extract the trunk centroid of each frame of the posture sequence and the chest cavity position of each frame of the trajectory of chest cavity position change, add the trunk centroid and the chest cavity position according to the preset weight, and use the coordinates obtained from each frame after smooth calculation as the current turning trajectory.

[0012] Furthermore, the weight of the trunk centroid is set to 0.2~0.3, the weight of the chest cavity point is set to 0.7~0.8, and the sum of the weight of the trunk centroid and the weight of the chest cavity point is 1.

[0013] Further, step S5 includes: Calculate the Euclidean distance between each node of the current turning trajectory and the optimal turning trajectory, count the number of nodes whose Euclidean distance is less than a preset tolerance threshold, and output the standardization evaluation result of the turning care as the higher the number of nodes.

[0014] Furthermore, the preset tolerance threshold is any value between 2 and 4 cm.

[0015] Furthermore, a monitoring system for the turning and care behavior of hemiplegic and disabled elderly based on multi-data fusion is provided, which implements the aforementioned evaluation method for the turning and care behavior of disabled elderly based on multi-data fusion during operation.

[0016] Therefore, the present invention provides the following effects and / or advantages: This invention performs multi-point scanning and signal quality assessment of the target's chest cavity region using radar before the turning operation, generating an initial vital sign confidence distribution map. The highest confidence point is extracted as the initial optimal vital sign measurement point, resolving the tracking starting point deviation problem caused by the inconsistency between the pre-stored trajectory and the elderly person's current actual position. During the turning process, this invention controls the radar to perform multi-point scanning within a preset spatial neighborhood, centered on the theoretical point of the optimal turning trajectory. By generating a local vital sign confidence distribution map, it achieves adaptive positioning of the optimal signal source, solving the problem of deviation between the actual optimal signal source and the theoretical trajectory due to individual differences during the turning process. The chest cavity point change trajectory is then fused with the posture sequence to form a current turning trajectory, which is compared with the optimal turning trajectory, thus realizing a quantitative evaluation method for the turning process of disabled hemiplegic elderly people.

[0017] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0018] It should be understood that the above summary and the following detailed description of the invention are exemplary and explanatory, and are intended to provide further explanation of the invention as claimed. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating one embodiment of the present invention. Detailed Implementation

[0020] To facilitate understanding by those skilled in the art, the present invention will now be described in further detail with reference to the embodiments: refer to Figure 1 A method for evaluating the turning and care behaviors of disabled elderly people based on multi-data fusion includes the following steps: S1, obtain the pre-stored optimal turning trajectory for the target object; S2, before the nursing turning operation, the radar collects the chest cavity micro-motion signal of the target object in real time and extracts the initial optimal vital measurement point; S3, during the nursing turning operation, the control radar takes the initial optimal vital sign measurement point as the turning starting point, performs multi-point scanning in the preset space area of ​​the optimal turning trajectory, collects the echo signal of each preset space area, performs signal quality assessment, and generates the corresponding local vital sign confidence distribution map. At the same time, the radar obtains the posture sequence of the target object during the turning process. S4. After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is taken as the chest cavity point change trajectory of the target object. The chest cavity point change trajectory is fused with the posture sequence to form the current turning trajectory. S5. Based on the comparison between the current turning trajectory and the optimal turning trajectory, output the evaluation result of the standardization of turning care.

[0021] This embodiment utilizes a pre-stored optimal turning trajectory for the target subject. This optimal trajectory can be a customized turning movement by professional caregivers based on the target subject's disability or hemiplegia, combined with the subject's personal sensations after turning. Then, starting from the optimal vital sign measurement points of the target subject in a supine position, the radar continuously collects micro-motion signals along the optimal turning trajectory, expanding from each point on the optimal trajectory to a preset spatial area for scanning. These scan signals, through signal quality assessment, can roughly determine the location of the optimal chest vital sign collection points within each preset space, thereby identifying the approximate orientational changes of the chest cavity during turning. Next, combining the posture changes scanned by the radar, a more precise prediction of the chest cavity's orientational changes can be made, resulting in the current turning trajectory. Finally, by comparing the current turning trajectory with the optimal turning trajectory, the difference between the two can be used as a result to evaluate the standardization of turning care, achieving objective quantification of care quality.

[0022] Furthermore, in step S1, the optimal rolling trajectory is obtained through the following steps: A trackable sensor is placed on the chest cavity surface of the target object at the area corresponding to the heartbeat. During a standard turning operation, the spatial position change of the trackable sensor is continuously tracked by a tracker to form the optimal turning trajectory.

[0023] In this embodiment, the sensor can be a passive reflector with a high radar cross-section, such as a corner reflector or a metal patch, or a positioning transmitter, such as a Bluetooth transmitter. The corresponding tracker is a radar or Bluetooth receiver. To optimize vital sign measurement, this embodiment places a trackable sensor on the chest surface of the target subject, corresponding to the area of ​​heart pulsation. This ensures that the sensor remains aligned with the heart pulsation area during the turning process, as this area contains the most prominent vital signs. By continuously tracking the sensor's spatial position changes using radar or positioning equipment during a turning operation performed by a professional nurse according to standard nursing protocols, the sensor's three-dimensional coordinates as it moves during the turning process are recorded, forming and storing the optimal turning trajectory.

[0024] Furthermore, step S2 includes: performing the following before the nursing repositioning procedure: S21 uses radar to scan around a point in a pre-defined, coarse candidate area for the location of the chest cavity and calculates multiple vital signs. S22, if multiple vital signs are similar, the point is determined to be the initial optimal vital sign measurement point; if multiple vital signs are not similar, the point is shifted and returned to S21.

[0025] Furthermore, by calculating the standard deviation of multiple vital signs, if the standard deviation is less than a preset standard deviation threshold, the multiple vital signs are determined to be similar.

[0026] In this embodiment, the initial optimal vital signs measurement point can be the location of the chest cavity corresponding to the heart. The mechanical waves emitted when the heart beats cause the chest cavity to fluctuate around the point corresponding to the heart. If multiple vital signs around this point are similar, then this point can be considered the location corresponding to the heart and can be considered the initial optimal vital signs measurement point.

[0027] In another embodiment, step S2 includes: performing the following before the nursing turning operation: S21, In a pre-defined candidate area for the coarse location of the thoracic cavity, a multi-point scan is performed by radar to extract the first micro-motion signal; S22, the signal quality of the first micro-motion signal at each point is evaluated to obtain the signal quality index of each point, and an interpolation algorithm is used to generate an initial vital sign confidence distribution map covering the coarse chest cavity location candidate area based on the signal quality index. S23, the highest confidence point of the initial vital signs confidence distribution map is taken as the initial optimal vital signs measurement point.

[0028] In this step, since the elderly person with hemiplegia is in a supine position for a long time, the steps before the nursing turning operation can be considered as being performed while the person is lying down. A confidence distribution map of vital signs at each location within the coarse chest cavity location candidate area can be obtained through signal quality assessment. Furthermore, an interpolation algorithm can be used to spatially interpolate the signal quality indicators of each scanning point to generate a high-density confidence distribution matrix covering the coarse chest cavity location candidate area. The grid density of the high-density confidence distribution matrix is ​​greater than the density of the scanning point array, and this high-density confidence distribution matrix is ​​used as the confidence distribution map of vital signs.

[0029] Further, step S22 includes: S22.1 Calculate the signal-to-noise ratio, phase coherence, amplitude stability, and template matching degree for the first micro-motion signal at each point, and then weight and sum the signal-to-noise ratio, phase coherence, amplitude stability, and template matching degree to obtain the signal quality index. S22.2, an initial vital sign confidence distribution map covering the rough candidate region of the thoracic cavity is generated using the inverse distance weighted interpolation method.

[0030] Furthermore, the signal-to-noise ratio is obtained by performing a fast Fourier transform on the first micro-motion signal to obtain the spectrum, and then calculating the ratio of peak power to noise power within the vital signs frequency range. In this step, the respiratory frequency range is determined to be 0.1Hz~0.5Hz and the heart rate range is determined to be 0.8Hz~2.0Hz. The peak frequencies within these ranges are extracted and the corresponding power is taken as the peak power. The average power of all frequency points in the spectrum other than the signal frequency range and its harmonics is taken as the noise power, and the ratio is calculated to obtain the corresponding signal-to-noise ratio.

[0031] The phase coherence is calculated by the following method: the phase sequence of the first micro-motion signal is used to calculate the fluctuation of the phase difference between adjacent frames, and the number of times the fluctuation matches the oscillation cycle of human vital signs within a preset period is counted as the phase coherence. The amplitude stability is calculated by the following method: extracting the amplitude envelope by performing a Hilbert transform on the first micro-motion signal, and counting the number of times the amplitude envelope conforms to the oscillation amplitude of human vital signs within a preset period as the amplitude stability. In this step, the preset period can be set to 5 minutes. The radar scans the point every 15 seconds and calculates the amplitude envelope and volatility through the first micro-motion signal. Then, it calculates the number of times the amplitude envelope and volatility meet the range corresponding to human vital signs within 5 minutes, which is used as the corresponding score.

[0032] The template matching degree is obtained by normalizing the cross-correlation between the first micro-motion signal and the pre-constructed respiratory template and / or heartbeat template, respectively. In this embodiment, the respiratory template adopts a sine wave model, and the heartbeat template adopts a mathematical model simulating the waveform of a cardiac impact curve. Normalized cross-correlation calculations are performed on the first micro-motion signal within a preset period and the corresponding template, and the maximum value of the respiratory template matching degree or the heartbeat template matching degree is taken as the final template matching degree.

[0033] The weights for the signal-to-noise ratio, the phase coherence, the amplitude stability, and the template matching degree are 0.25~0.35, 0.20~0.30, 0.20~0.30, and 0.15~0.20, respectively.

[0034] In this embodiment, since the signal-to-noise ratio (SNR) is the fundamental criterion, it is given the highest weight. Phase coherence and amplitude stability reflect whether the data conforms to the range of human vital characteristics, reflecting the regularity of the micro-motion signal at this location in the time dimension. They are key indicators for distinguishing regular breathing / heartbeat from random noise or occasional limb micro-movements, so they are given a medium weight. Template matching degree verifies the signal from the perspective of waveform morphology. Template matching degree is relatively less affected by changes in body position and has good robustness, so it is given the lowest weight. Specifically, the weights of the SNR, phase coherence, amplitude stability, and template matching degree are 0.3, 0.25, 0.25, and 0.20, respectively.

[0035] Further, step S3 includes: S3.1 During the nursing turning operation, the initial point of the optimal turning trajectory is set at the turning starting point, and the control radar performs multi-point scanning within a preset space area with a preset radius, centered on each point of the optimal turning trajectory. In this embodiment, setting the initial point of the optimal turning trajectory at the starting point of the turning process allows for multi-point scanning within a pre-defined spatial area along the optimal turning trajectory, using the point where vital signs are best obtained in the chest cavity before the target object is turned as the starting point. The pre-defined radius can be set to 30cm. Furthermore, the pre-defined spatial area can be gridded N×N, and scanning can be performed at the center of each grid to obtain multi-point scanning results.

[0036] S3.2, Evaluate the signal quality of the echo signal in each preset spatial area to obtain the signal quality index of each point, and generate a local vital sign confidence distribution map covering the preset spatial area based on the signal quality index using an interpolation algorithm. In this step, the method for signal quality assessment can be the same as that in step S22 above, thereby obtaining a local vital sign confidence distribution map.

[0037] S3.3 During the execution of S3.1 and S3.2, the radar simultaneously performs a real-time scan of the target object's entire torso and arranges the posture sequence in chronological order to cover the entire turning process.

[0038] In this step, the radar echo is processed by range-dimensional FFT and angle estimation to generate three-dimensional point cloud data containing a large number of reflection points on the human body surface. Then, the algorithm is used to extract the key points of the human skeleton from the sparse point cloud to obtain the corresponding human pose.

[0039] Further, step S4 includes: S4.1 After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is fitted using a smooth interpolation algorithm to generate the chest cavity point change trajectory of the target object. In this step, it's important to note that since the target may deviate from the optimal turning trajectory during the turning process, the confidence levels of the vital signs obtained from the local confidence distribution map may all be at a medium or low level. In this case, the gradient information of the local confidence map can be used to predict whether there are higher confidence points outside the map. Specifically, the gradient direction and magnitude of the confidence value of the local vital sign confidence distribution map on the neighborhood boundary are further calculated. If there is a continuous upward trend in a certain direction on the neighborhood boundary, i.e., the boundary treatment confidence value increases along a certain direction, then quadratic polynomial extrapolation or cubic spline extrapolation is performed along that direction to predict whether there are higher confidence points outside the map and their spatial coordinates. If the prediction result shows that there is a higher point outside the map, then this predicted point replaces the highest point in the neighborhood; if the extrapolation result is not significant, i.e., the highest point in the map is a stable extreme value, then the highest point in the neighborhood is still used.

[0040] By interpolating and smoothing these points with the highest confidence levels, a continuous and smooth trajectory of changes in chest cavity points is obtained.

[0041] S4.2 Align the trajectory of chest cavity position change with the posture sequence on the time axis, extract the trunk centroid of each frame of the posture sequence and the chest cavity position of each frame of the trajectory of chest cavity position change, add the trunk centroid and the chest cavity position according to the preset weight, and use the coordinates obtained from each frame after smooth calculation as the current turning trajectory.

[0042] Furthermore, the weight of the trunk centroid is set to 0.2~0.3, the weight of the chest cavity point is set to 0.7~0.8, and the sum of the weight of the trunk centroid and the weight of the chest cavity point is 1.

[0043] In this step, the spatial position change of the chest cavity during the turning process includes two components: the displacement component with the overall movement of the trunk and the local displacement component of the chest cavity relative to the trunk. The former reflects the progress of the turning itself, while the latter reflects whether the chest cavity is subjected to improper support or compression. If only the absolute coordinates of the chest cavity points are used, it is impossible to distinguish between these two components, which may lead to similar changes in the trajectory due to the caregiver's standardized or non-standard operations, resulting in misjudgment. By introducing the trunk centroid as a reference and performing weighted fusion, the core information of the chest cavity points is retained with high weight, while the overall movement constraint of the trunk is introduced with low weight. This ensures that the generated current turning trajectory is dominated by the quality of the chest cavity signal and conforms to the overall physical movement law of the turning process. Preferably, the weight of the trunk centroid is set to 0.25, and the weight of the chest cavity points is set to 0.75. This weight setting can impose reasonable physical constraints on the chest cavity position without significantly increasing the computational burden, and the weight values ​​match the confidence levels of the two types of data.

[0044] Further, step S5 includes: Calculate the Euclidean distance between each node of the current turning trajectory and the optimal turning trajectory, count the number of nodes whose Euclidean distance is less than a preset tolerance threshold, and output the standardization evaluation result of the turning care as the higher the number of nodes.

[0045] In this embodiment, the Euclidean distance reflects the deviation of each node between the current rolling trajectory and the optimal rolling trajectory. Since the target object cannot perfectly fit the optimal rolling trajectory during the rolling process, it only needs to be within a suitable tolerance range. Preferably, the preset tolerance threshold in this application can be set to any value between 2 and 4 cm. The evaluation result can be obtained by multiplying the number of nodes less than the preset tolerance threshold by the total number of nodes by 100%.

[0046] Furthermore, a monitoring system for the turning and care behavior of hemiplegic and disabled elderly based on multi-data fusion is provided, which implements the aforementioned evaluation method for the turning and care behavior of disabled elderly based on multi-data fusion during operation.

[0047] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0049] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0050] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0051] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms should not be construed as necessarily referring to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

Claims

1. A method for evaluating the turning and care behaviors of disabled elderly people based on multi-data fusion, characterized in that: Includes the following steps: S1, obtain the pre-stored optimal turning trajectory for the target object; S2, before the nursing turning operation, the radar collects the chest cavity micro-motion signal of the target object in real time and extracts the initial optimal vital measurement point; S3, during the nursing turning operation, the control radar takes the initial optimal vital sign measurement point as the turning starting point, performs multi-point scanning in the preset space area of ​​the optimal turning trajectory, collects the echo signal of each preset space area, performs signal quality assessment, and generates the corresponding local vital sign confidence distribution map. At the same time, the radar obtains the posture sequence of the target object during the turning process. S4. After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is taken as the chest cavity point change trajectory of the target object. The chest cavity point change trajectory is fused with the posture sequence to form the current turning trajectory. S5. Based on the comparison between the current turning trajectory and the optimal turning trajectory, output the evaluation result of the standardization of turning care.

2. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 1, characterized in that: In step S1, the optimal rolling trajectory is obtained through the following steps: A trackable sensor is placed on the chest cavity surface of the target object at the area corresponding to the heartbeat. During a standard turning operation, the spatial position change of the trackable sensor is continuously tracked by a tracker to form the optimal turning trajectory.

3. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 1, characterized in that: Step S2 includes: Before the nursing repositioning procedure, perform the following: S21 uses radar to scan around a point in a pre-defined, coarse candidate area for the location of the chest cavity and calculates multiple vital signs. S22, if multiple vital signs are similar, the point is determined to be the initial optimal vital sign measurement point; if multiple vital signs are not similar, the point is shifted and returned to S21.

4. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 3, characterized in that: By calculating the standard deviation of multiple vital signs, if the standard deviation is less than a preset standard deviation threshold, the multiple vital signs are determined to be similar.

5. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 1, characterized in that: Step S3 includes: S3.1 During the nursing turning operation, the initial point of the optimal turning trajectory is set at the turning starting point, and the control radar performs multi-point scanning within a preset space area with a preset radius, centered on each point of the optimal turning trajectory. S3.2, Evaluate the signal quality of the echo signal in each preset spatial area to obtain the signal quality index of each point, and generate a local vital sign confidence distribution map covering the preset spatial area based on the signal quality index using an interpolation algorithm. S3.3 During the execution of S3.1 and S3.2, the radar simultaneously performs a real-time scan of the target object's entire torso and arranges the posture sequence in chronological order to cover the entire turning process.

6. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 1, characterized in that: Step S4 includes: S4.1 After the nursing turning operation is completed, the point with the highest confidence in the local vital signs confidence distribution map of each point is fitted using a smooth interpolation algorithm to generate the chest cavity point change trajectory of the target object. S4.2 Align the trajectory of chest cavity position change with the posture sequence on the time axis, extract the trunk centroid of each frame of the posture sequence and the chest cavity position of each frame of the trajectory of chest cavity position change, add the trunk centroid and the chest cavity position according to the preset weight, and use the coordinates obtained from each frame after smooth calculation as the current turning trajectory.

7. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 6, characterized in that: The weight of the trunk centroid is set to 0.2~0.3, the weight of the chest cavity point is set to 0.7~0.8, and the sum of the weight of the trunk centroid and the weight of the chest cavity point is 1.

8. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 1, characterized in that: Step S5 includes: Calculate the Euclidean distance between each node of the current turning trajectory and the optimal turning trajectory, count the number of nodes whose Euclidean distance is less than a preset tolerance threshold, and output the standardization evaluation result of the turning care as the higher the number of nodes.

9. The method for evaluating the turning and care behavior of disabled elderly people based on multi-data fusion according to claim 8, characterized in that: The preset tolerance threshold is any value between 2 and 4 cm.

10. A monitoring system for the turning and care behavior of hemiplegic and disabled elderly people based on multi-data fusion, characterized in that, The method described in any one of claims 1-9 is used to evaluate the turning and care behavior of disabled elderly people based on multi-data fusion.

Images