Multi-dimensional health state assessment method for elderly care robot

Abstract

The invention discloses a multi-dimensional health state assessment method for an old-age care robot, and relates to the technical field of old-age health monitoring, and the method comprises the steps: collecting the physiological parameters of a cared old person through a built-in multi-sensor module of the old-age care robot, and forming a physiological parameter set; on the basis of the physiological parameter set, combining a preset physiological parameter threshold value, and performing weighted fusion on standardized numerical values of various physiological parameters to obtain a physiological health index through calculation; analyzing facial expressions and voice features of the cared old people by adopting an emotion recognition algorithm, and extracting psychological state data of the cared old people; and inputting the physiological health indexes and the psychological state characteristic parameters into a health state evaluation model to generate a health state evaluation report. Potential health risks can be found in time, a scientific decision basis can be provided for medical staff and family members, and the health management level and life quality of old people are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring for the elderly, and particularly to a multi-dimensional health status assessment method for elderly care robots. Background Art

[0002] Traditional health monitoring for the elderly mainly relies on manual care and regular physical examinations, which has certain limitations. In recent years, the deep integration of intelligent robot technology and the medical and health field has given rise to elderly care robots. However, there are still obvious limitations in the health status assessment of existing elderly care robots. Most current assessment systems use single physiological parameter monitoring or simple multi-parameter linear superposition methods, lacking the integrated analysis of the physiological and psychological states of the elderly. In addition, traditional assessment methods usually use fixed threshold judgments, failing to fully consider individual differences and the mutual influence between parameters, resulting in insufficient accuracy and real-time performance of the assessment results.

[0003] Existing elderly health monitoring systems still face some deficiencies. First, most existing systems only rely on the single assessment of physiological parameters and ignore the influencing factors of the mental health of the elderly. Although some systems have started to incorporate emotion recognition and facial expression analysis, the accuracy and sensitivity of their methods still need to be improved. Second, the existing health assessment methods generally have the problems of low standardization and insufficient consideration of individual differences. Most systems use unified health indicators and assessment criteria, failing to fully combine the physiological and psychological states of individuals, resulting in relatively rough assessment results of the health status. In addition, most of the health assessment models of existing technologies rely on single algorithms or traditional medical knowledge systems, lacking multi-dimensional comprehensive assessment capabilities and unable to effectively integrate various data, resulting in limitations in the accuracy and credibility of the assessment results. Therefore, there is still much room for improvement in personalization, precision, and multi-dimensional comprehensive assessment in existing technologies.

[0004] In view of the above problems, the present invention proposes a multi-dimensional health status assessment method for elderly care robots, which comprehensively assesses and real-time monitors the health status of the elderly by collecting physiological parameters through multi-sensor collaboration, analyzing the psychological state in combination with emotion recognition algorithms, and constructing a comprehensive assessment model based on support vector machines. Summary of the Invention

[0005] In view of the fact that existing health assessment methods mainly focus on single physiological data monitoring or rough psychological state analysis and lack accurate assessment of the overall health status of the elderly, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is how to improve the accuracy and personalization of the health status assessment of the elderly through multi-dimensional comprehensive assessment of physiological and psychological data, so as to achieve more comprehensive and real-time health monitoring.

[0007] To solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a multi-dimensional health status assessment method for an elderly care robot, which includes collecting physiological parameters of the elderly being cared for through a multi-sensor module built into the elderly care robot to form a set of physiological parameters; based on the set of physiological parameters, combining preset physiological parameter thresholds, and calculating a physical health index by weighted fusion of the standardized values of each physiological parameter; using an emotion recognition algorithm to analyze the facial expressions and voice characteristics of the elderly being cared for, and extracting the psychological state data of the elderly being cared for; inputting the physical health index and the psychological state characteristic parameters into a health status assessment model to generate a health status assessment report.

[0009] As a preferred solution of the multi-dimensional health status assessment method for the elderly care robot of the present invention, wherein: the method for generating the health status assessment report is to perform data standardization processing on the physical health index, and convert the qualitative characteristics in the psychological state data into quantitative characteristic parameters to form psychological state characteristic parameters; input the processed physical health index and psychological state characteristic parameters as feature vectors into a health status assessment model; calculate the comprehensive physical and mental health score of the elderly being cared for through the health status assessment model, and at the same time perform multi-level classification on the health status of the elderly being cared for through the comprehensive physical and mental health score; combine the historical health data of the elderly being cared for, analyze the time evolution trend of the health status, and identify health risk factors; generate a comprehensive health status assessment report according to the analysis results and the comprehensive physical and mental health score, wherein the comprehensive health status assessment report includes a health status summary, abnormal index prompts, health trend charts, and personalized health suggestions; send the comprehensive health status assessment report to the terminal devices designated by the family members and medical staff through a secure encrypted channel to achieve remote real-time monitoring of the health status.

[0010] As a preferred solution of the multi-dimensional health status assessment method for the elderly care robot of the present invention, wherein: the health status assessment model is constructed using the support vector machine algorithm with a radial basis kernel function, and the model is trained through health status samples annotated by experts to optimize the kernel function parameters and penalty factors; the specific formula of the health status assessment model is as follows:

[0011]

[0012] wherein, H is the comprehensive physical and mental health score, w i is the weight coefficient of the i-th physiological parameter, P i is the actual value of the i-th physiological parameter, P min and P maxare the minimum and maximum values of the corresponding parameters respectively, λ is the emotional state influence factor, and E j is the j-th emotional feature value, n is the total number of emotional features, and T k is the weight of the k-th time series feature, and S k is the change slope of the k-th historical health indicator, and m is the number of historical health indicators.

[0013] As a preferred solution of the multi-dimensional health status assessment method for the elderly care robot described in the present invention, wherein: the method for extracting the mental state data is to detect the facial area of the elderly being cared for through the high-definition camera module built in the elderly care robot, obtain facial image data, and perform feature point positioning on the facial image data through a deep learning model to extract the coordinates of key feature points; based on the coordinates of the key feature points, calculate the relative distances and angular changes between the feature points to obtain facial geometric feature parameters; collect the voice signals of the elderly being cared for and preprocess the voice signals to extract acoustic feature parameters; input the facial geometric feature parameters and the acoustic feature parameters into a pre-trained multi-modal emotion recognition model, wherein the multi-modal emotion recognition model is constructed based on a convolutional neural network and a long short-term memory network, and adaptively fuses facial features and voice features through an attention mechanism; generate a mental state feature vector according to the probability distribution of the emotional state of the elderly being cared for output by the multi-modal emotion recognition model in combination with the emotional change trend, wherein the mental state feature vector includes emotion category, emotion intensity, and emotion stability; input the mental state feature vector into a health status assessment model for comprehensive analysis and output a comprehensive physical and mental health score.

[0014] As a preferred solution of the multi-dimensional health status assessment method for the elderly care robot described in the present invention, wherein: the calculation method of the physical health index is to compare the physiological parameters with the preset physiological parameter thresholds, and calculate the degree of deviation from the normal range of the physiological parameters by using the Z-score method to generate a standardized value; assign weight coefficients according to the importance of the physiological parameters for health status assessment, and multiply and accumulate the standardized value and the corresponding weight coefficients to form a physical health index; map the physical health index to a percentile scale and display the physical health status of the elderly being cared for in a way that combines color coding and numerical values.

[0015] As a preferred solution of the multi-dimensional health status assessment method for the elderly care robot described in the present invention, wherein: the specific formula of the physical health index is as follows:

[0016]

[0017] wherein, I PH is the physical health index, N is the total number of physiological parameters, and w iis the weight coefficient of the ith physiological parameter, x i is the index of the ith physiological parameter, Z(x i ) is the collected value of the ith physiological parameter index, f(x) is the Z-score of the ith physiological parameter, and the specific formula is as follows:

[0018]

[0019] Among them, I min and I max are the minimum and maximum values ​​of the health index respectively.

[0020] As a preferred solution of the multi-dimensional health status assessment method of the elderly care robot described in the present invention, the method for forming the physiological parameter set is to collect physiological parameters through a multi-sensor module, wherein the multi-sensor module includes a thermopile temperature sensor, a photoelectric capacitance pulse wave sensor, a vibration blood pressure sensor, an electrochemical oxygen sensor, a radar respiratory monitor and a microprocessor; based on the collected physiological parameters, combined with the cooperation degree and physical condition of the elderly being cared for, the microprocessor dynamically adjusts the collection frequency and collection intensity; the physiological parameters are preliminarily screened and outliers are identified, erroneous data are eliminated, and valid data are organized into a structured physiological parameter set, wherein the physiological parameters include body temperature, blood pressure, heart rate, blood oxygen and respiratory rate.

[0021] In the second aspect, an embodiment of the present invention provides a multi-dimensional health status assessment system for an elderly care robot, which includes: an acquisition module, which is used to acquire physiological parameters of the elderly being cared for through the multi-sensor module built into the elderly care robot to form a physiological parameter set; a calculation module, which is based on the physiological parameter set, combined with preset physiological parameter thresholds, and calculates the physiological health index by weighted fusion of standardized numerical values ​​of various physiological parameters; an extraction module, which is used to analyze the facial expressions and voice features of the elderly being cared for using an emotion recognition algorithm, and extract the psychological state data of the elderly being cared for; a construction and generation module, which is used to input the physiological health index and the psychological state characteristic parameters into a health status assessment model to generate a comprehensive health status assessment report for the elderly.

[0022] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the multi-dimensional health status assessment method for the elderly care robot as described in the first aspect of the present invention are implemented.

[0023] Fourthly, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein: when the computer program instructions are executed by a processor, the steps of the multi-dimensional health status assessment method for the elderly care robot as described in the first aspect of the present invention are implemented.

[0024] The beneficial effects of the present invention are as follows: By synchronously collecting multiple physiological parameters such as body temperature, blood pressure, heart rate, blood oxygen, and respiratory rate through the multi-sensor module built in the elderly care robot, a set of physiological parameters is formed, reducing the trouble caused to the elderly by multiple measurements and improving the timeliness and accuracy of data collection; By weighted fusion of various standardized physiological parameters, a unified physiological health index is generated, enabling comprehensive evaluation of physiological indicators with different dimensions and ranges, facilitating medical staff to quickly judge the overall physiological health status of the elderly; At the same time, an emotion recognition algorithm is used to analyze the facial expressions and voice features of the elderly to obtain psychological state data, breaking through the limitation of traditional health assessment that only focuses on physiological indicators and making the health assessment more comprehensive and in-depth; Finally, the physiological health index and psychological state characteristic parameters are input into the health status assessment model, and through the fusion analysis of multi-dimensional data, a comprehensive health status assessment report is generated. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0026] Figure 1 It is a flowchart of the multi-dimensional health status assessment method for the elderly care robot in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the drawings in the specification.

[0028] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0029] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0030] Example 1

[0031] Reference Figure 1 , which is the first embodiment of the present invention, and provides a multi-dimensional health status assessment method for an elderly care robot, comprising:

[0032] S1: The multi-sensor module built into the elderly care robot collects the physiological parameters of the elderly being cared for to form a physiological parameter set.

[0033] Specifically, physiological parameters are collected through a multi-sensor module, wherein the multi-sensor module includes a thermopile temperature sensor, a photoelectric capacitance pulse wave sensor, a vibration blood pressure sensor, an electrochemical oxygen sensor, a radar respiratory monitor and a microprocessor.

[0034] It should be noted that the thermopile temperature sensor is arranged in the front sensing area of ​​the elderly care robot, and performs non-contact body temperature measurement within 5-10 cm from the forehead of the elderly being cared for, avoiding the risk of cross-infection caused by traditional contact thermometers; the photoelectric volumetric pulse wave sensor and the vibrating blood pressure sensor are integrated in the multi-functional arm of the elderly care robot, wherein the photoelectric volumetric pulse wave sensor obtains the heart rate and blood oxygen data of the elderly being cared for by emitting light of a specific wavelength and detecting changes in the reflected light signal, and the vibrating blood pressure sensor uses a gentle pressurization method to measure the systolic and diastolic blood pressure of the elderly being cared for; the electrochemical oxygen sensor and the radar respiratory monitor are installed in the chest area of ​​the elderly care robot, wherein the electrochemical oxygen sensor is used to assist in the verification measurement of blood oxygen saturation, and the radar respiratory monitor uses millimeter wave technology to penetrate thin clothing to monitor the tiny chest movements of the elderly being cared for, thereby calculating the respiratory rate.

[0035] Furthermore, based on the collected physiological parameters and in combination with the cooperation and physical condition of the elderly being cared for, the collection frequency and intensity are dynamically adjusted through a microprocessor.

[0036] Furthermore, physiological parameters are preliminarily screened and outliers are identified to eliminate erroneous data, and valid data are organized into a structured set of physiological parameters, including body temperature, blood pressure, heart rate, blood oxygen, and respiratory rate.

[0037] S2: Based on the physiological parameter set and the preset physiological parameter thresholds, the physiological health index is calculated by weighted fusion of the standardized values ​​of various physiological parameters.

[0038] Specifically, the physiological parameters are compared with the preset physiological parameter thresholds, and the Z-score method is used to calculate the degree of deviation from the normal range of the physiological parameters to generate a standardized value.

[0039] It should be noted that the normal range standard of the preset physiological parameter threshold is a multi-level standard subdivided according to the physiological characteristics of the elderly in different age groups; if the body temperature exceeds the normal range interval of 36.1°C to 37.2°C, the standard deviation multiple of the body temperature deviation from the median value is calculated to generate the body temperature Z-score; if the blood pressure exceeds the normal range interval of systolic blood pressure 90-140 mmHg and diastolic blood pressure 60-90 mmHg, the standard deviation multiples of the systolic blood pressure and diastolic blood pressure deviations from their respective reference values are calculated respectively, and the average value is taken as the blood pressure Z-score; if the heart rate exceeds the normal range interval of 60-100 beats per minute, the standard deviation multiple of the heart rate deviation from the median value of the normal reference range is calculated to generate the heart rate Z-score; if the blood oxygen is lower than the normal threshold of 95%, the blood oxygen Z-score is calculated according to the degree of deviation, and for each 1 percentage point decrease in the degree of deviation, the Z-score increases by 0.5 standard deviations; if the respiratory rate exceeds the normal range interval of 12-20 times per minute, the standard deviation multiple of the respiratory rate deviation from the median value of the normal reference range is calculated to generate the respiratory rate Z-score.

[0040] Furthermore, weight coefficients are assigned according to the importance of physiological parameters for health status assessment, and the standardized values are multiplied by the corresponding weight coefficients and accumulated to form a physiological health index. The specific formula is as follows:

[0041]

[0042] Among them, I PH is the physiological health index, N is the total number of physiological parameters, w i is the weight coefficient of the i-th physiological parameter, x i is the index of the i-th physiological parameter, Z(x i ) is the collected value of the i-th physiological parameter index, f(x) is the Z-score of the i-th physiological parameter. The specific formula is as follows:

[0043]

[0044] Among them, I min and I max are the minimum and maximum values of the health index respectively, μ i is the normal mean value of the i-th physiological parameter, σ i is the standard deviation of the i-th physiological parameter.

[0045] It should be noted that the weight coefficients are determined through the expert knowledge base and clinical medical statistical data, and are configured differently for the elderly in different age groups and with different types of underlying diseases.

[0046] Further, map the physiological health index onto a hundred-point scale, and display the physiological health status of the cared-for elderly through a combination of color coding and numerical values.

[0047] S3: Analyze the facial expressions and voice features of the cared-for elderly using an emotion recognition algorithm to extract the psychological state data of the cared-for elderly.

[0048] Specifically, use the high-definition camera module built into the elderly care robot to detect the facial area of the cared-for elderly, obtain facial image data, and perform feature point localization on the facial image data through a deep learning model to extract the coordinates of key feature points.

[0049] It should be noted that the high-definition camera module is equipped with an optical compensation system suitable for low-light environments to ensure clear acquisition of facial images under insufficient indoor light; the deep learning model is constructed based on an artificial neural network, and by simulating the hierarchical connection structure of human brain neurons, it uses large-scale data to train and optimize parameters to automatically learn features and perform feature point localization.

[0050] Further, based on the coordinates of the key feature points, calculate the relative distances and angular changes between the feature points to obtain facial geometric feature parameters; collect the voice signals of the cared-for elderly, and preprocess the voice signals to extract acoustic feature parameters.

[0051] It should be noted that the facial geometric feature parameters include the distance between the eyebrows, the degree of eyelid opening and closing, and the angle of the mouth rising; the acoustic feature parameters include the fundamental frequency, energy, and formants of the voice signal.

[0052] Further, input the facial geometric feature parameters and acoustic feature parameters into a pre-trained multi-modal emotion recognition model, where the multi-modal emotion recognition model is constructed based on a convolutional neural network and a long short-term memory network, and adaptively fuses the facial features and voice features through an attention mechanism.

[0053] Specifically, according to the multi-modal emotion recognition model, output the probability distribution of the emotion state of the cared-for elderly, and combine the emotion change trend to generate a psychological state feature vector, where the psychological state feature vector includes emotion category, emotion intensity, and emotion stability.

[0054] It should be noted that the emotion state includes three basic categories: positive, calm, and negative, and each emotion state corresponds to a probability value.

[0055] Further, input the psychological state feature vector into a health status assessment model for comprehensive analysis and output a comprehensive physical and mental health score.

[0056] S4: Input the physiological health index and psychological state characteristic parameters into the health status assessment model to generate a health status assessment report.

[0057] Specifically, perform data standardization processing on the physiological health index, and convert the qualitative characteristics in the psychological state data into quantitative characteristic parameters to form psychological state characteristic parameters; input the processed physiological health index and psychological state characteristic parameters as feature vectors into the health status assessment model.

[0058] Furthermore, the health status assessment model is constructed using the support vector machine algorithm with a radial basis kernel function, and the model is trained through health status samples annotated by experts to optimize the kernel function parameters and penalty factors. The specific formula of the health status assessment model is as follows:

[0059]

[0060] Among them, H is the comprehensive physical and mental health score, w i is the weight coefficient of the i-th physiological parameter, P i is the actual value of the i-th physiological parameter, P min and P max are the minimum and maximum values of the corresponding parameters respectively, λ is the emotional state influence factor, E j is the j-th emotional characteristic value, n is the total number of emotional characteristics, T k is the weight of the k-th time series feature, S k is the change slope of the k-th historical health indicator, and m is the number of historical health indicators.

[0061] Even further, calculate the comprehensive physical and mental health score of the cared-for elderly through the health status assessment model, and at the same time perform multi-level classification on the health status of the cared-for elderly based on the comprehensive physical and mental health score.

[0062] It should be noted that when the comprehensive physical and mental health score ≥ health threshold TH1, it is determined to be healthy, and the elderly care robot conducts daily monitoring and provides normal escort interaction; when health threshold TH2 ≤ comprehensive physical and mental health score < health threshold TH1, it is determined to be mildly abnormal, and the elderly care robot actively reminds the elderly to pay attention to health management and recommends basic health checks; when health threshold TH3 ≤ comprehensive physical and mental health score < health threshold TH2, it is determined to be moderately abnormal, and the elderly care robot sends a health warning to the caregivers and provides a personalized health regulation plan; when the comprehensive physical and mental health score < health threshold TH3, it is determined to be severely abnormal, and the elderly care robot immediately activates the emergency response mechanism, notifies the medical institution and emergency contacts, and continuously monitors the abnormal data; the health threshold is determined comprehensively based on historical health data statistical analysis, medical expert experience, and individual physiological characteristics.

[0063] Specifically, in combination with the historical health data of the elderly under care, analyze the time evolution trend of the health status to identify health risk factors; generate a comprehensive health status assessment report based on the analysis results and the comprehensive physical and mental health score, where the comprehensive health status assessment report includes a summary of the health status, prompts for abnormal indicators, health trend charts, and personalized health advice.

[0064] It should be noted that the health trend chart takes time as the horizontal axis, and simultaneously shows the change curves of the physical health index and various mental state data, and marks the key change points and their correlations.

[0065] Furthermore, send the comprehensive health status assessment report to the terminal devices designated by the family members and medical staff through a secure encrypted channel to achieve remote real-time monitoring of the health status.

[0066] Furthermore, this embodiment also provides a multi-dimensional health status assessment system for an elderly care robot, including: a collection module for collecting the physiological parameters of the elderly under care through a multi-sensor module built in the elderly care robot to form a set of physiological parameters; a calculation module for calculating the physical health index by weighted fusion of the standardized values of each physiological parameter based on the set of physiological parameters and in combination with a preset physiological parameter threshold; an extraction module for analyzing the facial expressions and voice features of the elderly under care using an emotion recognition algorithm to extract the mental state data of the elderly under care; a construction and generation module for inputting the physical health index and the mental state characteristic parameters into a health status assessment model to generate a comprehensive health status assessment report for the elderly.

[0067] This embodiment also provides a computer device applicable to the multi-dimensional health status assessment method of an elderly care robot, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the multi-dimensional health status assessment method of the elderly care robot as proposed in the above embodiment.

[0068] The computer device may be a terminal, which includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0069] In summary, through the multi-sensor module built in the elderly care robot, this invention synchronously collects multiple physiological parameters such as body temperature, blood pressure, heart rate, blood oxygen, and respiratory rate to form a set of physiological parameters, reducing the trouble caused to the elderly by multiple measurements and improving the timeliness and accuracy of data collection; by weighted fusion of various standardized physiological parameters, a unified physiological health index is generated, enabling comprehensive evaluation of physiological indicators with different dimensions and ranges, facilitating medical staff to quickly judge the overall physiological health status of the elderly; at the same time, an emotion recognition algorithm is used to analyze the facial expressions and voice features of the elderly to obtain psychological state data, breaking through the limitation of traditional health assessment that only focuses on physiological indicators and making the health assessment more comprehensive and in-depth; finally, the physiological health index and psychological state characteristic parameters are input into the health status assessment model, and through the fusion analysis of multi-dimensional data, a comprehensive health status assessment report is generated.

[0070] Embodiment 2

[0071] Referring to Table 1, this is the second embodiment of the present invention. This embodiment provides a multi-dimensional health status assessment method for an elderly care robot. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0072] Specifically, a 30-day health status assessment experiment is conducted on the elderly population aged 60 - 85 in a certain nursing home. An intelligent care robot equipped with a high-definition camera module (resolution 4K / 60fps) and a high-precision sensor array (including a medical-grade thermopile temperature sensor, an FDA-certified oscillometric blood pressure sensor, a German imported photoplethysmograph sensor, etc.) is used as the test platform.

[0073] Furthermore, physical examinations were conducted on all the elderly participants in the experiment to exclude subjects with serious diseases. Finally, 40 subjects were determined, including 22 males and 18 females, with an average age of 72.5 years. The subjects were randomly divided into an experimental group (20 people) and a control group (20 people). The experimental group adopted the multi-dimensional health status assessment method of the present invention, and the control group adopted the traditional single physiological index monitoring method.

[0074] Furthermore, before the experiment started, all the subjects underwent a 3-day adaptation training to familiarize them with the interaction mode with the care robot. The elderly care robot collected the physiological parameters of the subjects at regular times every day (7:00 am, 12:00 noon, 19:00 pm). During the collection process, the robot guided the subjects to maintain the correct posture through voice interaction. At the same time, the elderly care robot also collected the facial expressions and voice data of the subjects during daily interactions (such as morning greetings, meal reminders, etc.) for psychological state assessment. After the collection of physiological parameters was completed, the system automatically performed data preprocessing, including outlier detection, signal smoothing and other operations. For facial expression recognition, an improved VGG-Face deep learning model was used to extract features, with an accuracy rate of 94.7%; for voice emotion recognition, a bidirectional LSTM network structure was used, and the recognition accuracy rate was 92.3%. Finally, the processed multi-dimensional data was input into the health status assessment model to generate a daily assessment report.

[0075] Specifically, as shown in Table 1, the method of the present invention has significant advantages compared with the traditional method. First of all, in terms of the recognition accuracy rate of physiological indexes, the method of the present invention reaches 97.5%, which is 5.6 percentage points higher than the traditional method. This benefits from the multi-sensor fusion technology and the improved data processing algorithm. Especially in terms of emotion state recognition, the accuracy rate of the method of the present invention reaches 93.5%, which is 9.7 percentage points higher than the traditional method, fully reflecting the superiority of the multi-modal emotion recognition model. More importantly, the method of the present invention performs outstandingly in health risk early warning, and can send out early warning signals 18.6 hours in advance, which is 10.2 hours earlier than the traditional method, with an increase of 121.4%. This is of great significance for preventing sudden health problems of the elderly.

[0076] Table 1. Comparison table of the method of the present invention and the traditional method

[0077]

[0078] Furthermore, in terms of the abnormal state recognition rate, the method of the present invention reaches 96.2%, which is 8.5 percentage points higher than the traditional method, indicating that this method can capture abnormal health states more accurately. At the same time, the method of the present invention also has significant advantages in terms of the evaluation report generation efficiency. The average generation time is only 12.3 seconds, which is 52.1% shorter than the traditional method. From the perspective of user experience, the method of the present invention has obtained a relatively high satisfaction score of 8.9 points, which is 23.6% higher than the traditional method, indicating that this method better meets the actual needs of the elderly. The recognition degree of medical staff for the method of the present invention reaches 9.1 points, which is 21.3% higher than the traditional method, indicating that the health assessment results provided by this method have relatively high professional reference value.

[0079] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A multi-dimensional health status assessment method for an elderly care robot, characterized by: include, The multi-sensor module built into the elderly care robot collects the physiological parameters of the elderly being cared for to form a physiological parameter set; Based on the physiological parameter set, combined with the preset physiological parameter threshold, the physiological health index is calculated by weighted fusion of the standardized values ​​of various physiological parameters; Use emotion recognition algorithms to analyze the facial expressions and voice features of the elderly being cared for, and extract the psychological state data of the elderly being cared for; The physiological health index and the psychological state characteristic parameters are input into a health state assessment model to generate a health state assessment report.

2. The multi-dimensional health status assessment method of the elderly care robot according to claim 1, characterized in that: The method for generating the health status assessment report is as follows: Perform data standardization on physiological health indexes and transform qualitative features in psychological state data into quantitative feature parameters to form psychological state feature parameters; The processed physiological health index and psychological state characteristic parameters are input into the health state assessment model as feature vectors; Calculating the comprehensive physical and mental health score of the elderly being cared for by the health status assessment model, and classifying the health status of the elderly being cared for into multiple levels by using the comprehensive physical and mental health score; Combine the historical health data of the elderly being cared for to analyze the time evolution trend of health status and identify health risk factors; Generate a comprehensive health status assessment report based on the analysis results and the comprehensive physical and mental health score, wherein the comprehensive health status assessment report includes a health status summary, abnormal indicator prompts, health trend charts, and personalized health recommendations; The comprehensive health status assessment report is sent to the terminal devices designated by family members and medical staff through a secure encrypted channel to achieve remote real-time monitoring of health status.

3. The multi-dimensional health status assessment method of the elderly care robot as claimed in claim 2, characterized in that: The health status assessment model is constructed using a support vector machine algorithm with a radial basis kernel function, and the model is trained using health status samples annotated by experts to optimize kernel function parameters and penalty factors; the specific formula of the health status assessment model is as follows: Among them, H is the comprehensive score of physical and mental health, w i is the weight coefficient of the ith physiological parameter, P i is the actual value of the ith physiological parameter, P min and P max are the minimum and maximum values ​​of the corresponding parameters, λ is the emotional state influencing factor, E j is the jth emotional feature value, n is the total number of emotional features, T k is the kth time series feature weight, S k is the change slope of the kth historical health indicator, and m is the number of historical health indicators.

4. The multi-dimensional health status assessment method of the elderly care robot according to claim 2 or 3, characterized in that: The method for extracting the psychological state data is: The facial area of ​​the elderly person being cared for is detected by the built-in high-definition camera module of the elderly care robot to obtain facial image data, and the feature points of the facial image data are located by a deep learning model to extract the coordinates of key feature points; Based on the coordinates of the key feature points, the relative distance and angle change between the feature points are calculated to obtain the facial geometric feature parameters; Collecting the voice signal of the elderly being cared for, preprocessing the voice signal, and extracting acoustic feature parameters; Inputting the facial geometric feature parameters and the acoustic feature parameters into a pre-trained multimodal emotion recognition model, wherein the multimodal emotion recognition model is constructed based on a convolutional neural network and a long short-term memory network, and adaptively fuses facial features and speech features through an attention mechanism; Outputting the probability distribution of the emotional state of the elderly being cared for according to the multimodal emotion recognition model and combining the emotion change trend to generate a psychological state feature vector, wherein the psychological state feature vector includes emotion category, emotion intensity and emotion stability; The psychological state feature vector is input into the health state assessment model for comprehensive analysis, and a comprehensive score of physical and mental health is output.

5. The multi-dimensional health status assessment method of the elderly care robot according to claim 2 or 3, characterized in that: The calculation method of the physiological health index is: Comparing the physiological parameter with a preset physiological parameter threshold, and calculating the degree of deviation of the physiological parameter from the normal range using the Z-score method to generate a standardized value; Allocate a weight coefficient according to the importance of the physiological parameter to the health status assessment, and multiply and accumulate the standardized value and the corresponding weight coefficient to form a physiological health index; The physiological health index is mapped onto a percentage scale, and the physiological health status of the elderly being cared for is displayed by combining color coding and numerical values.

6. The multi-dimensional health status assessment method of the elderly care robot according to claim 5, characterized in that: The specific formula of the physiological health index is as follows: Among them, I PH is the physiological health index, N is the total number of physiological parameters, w i is the weight coefficient of the ith physiological parameter, x i is the index of the ith physiological parameter, Z(x i ) is the collected value of the ith physiological parameter index, f(x) is the Z-score of the ith physiological parameter, and the specific formula is as follows: Among them, I min and I max are the minimum and maximum values ​​of the health index respectively.

7. The multi-dimensional health status assessment method of the elderly care robot according to claim 6, characterized in that: The method for forming the physiological parameter set is: Acquiring physiological parameters through a multi-sensor module, wherein the multi-sensor module includes a thermopile temperature sensor, a photoplethysmography sensor, a vibrating blood pressure sensor, an electrochemical oxygen sensor, a radar respiration monitor, and a microprocessor; Based on the collected physiological parameters, combined with the cooperation degree and physical condition of the elderly being cared for, the microprocessor dynamically adjusts the collection frequency and collection intensity; The physiological parameters are preliminarily screened and outliers are identified, erroneous data are eliminated, and valid data are organized into a structured set of physiological parameters, wherein the physiological parameters include body temperature, blood pressure, heart rate, blood oxygen and respiratory rate.

8. A multi-dimensional health status assessment system for an elderly care robot, based on the multi-dimensional health status assessment method for an elderly care robot according to any one of claims 1 to 7, characterized in that: include, The acquisition module is used to collect the physiological parameters of the elderly being cared for through the multi-sensor module built into the elderly care robot to form a physiological parameter set; A calculation module, based on the physiological parameter set and in combination with preset physiological parameter thresholds, calculates a physiological health index by weighted fusion of standardized values ​​of various physiological parameters; An extraction module is used to analyze the facial expressions and voice features of the elderly being cared for using an emotion recognition algorithm to extract the psychological state data of the elderly being cared for; The construction and generation module is used to input the physiological health index and the psychological state characteristic parameters into the health status assessment model to generate a comprehensive health status assessment report for the elderly.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-dimensional health status assessment method of the elderly care robot described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-dimensional health status assessment method of the elderly care robot according to any one of claims 1 to 7 are implemented.

Images