A method and device for detecting abnormality of human physiological and motion state

Abstract

The application is suitable for the technical field of machine learning, and provides a human physiological and motion state anomaly detection method and device.The human physiological and motion state anomaly detection method comprises the following steps: acquiring real-time physical sign data collected by a physical sign collector and standard physical sign data corresponding to a plurality of preset states respectively and obtained through pre-training; performing semantic classification on the real-time physical sign data to obtain real-time semantic information; each real-time semantic information corresponds to one or more preset states; and determining whether the real-time physical sign data appears an abnormal state according to the standard physical sign data corresponding to the real-time semantic information.The technical scheme introduces an anomaly detection method based on semantic classification, significantly improves the accuracy of physical sign data analysis in a wearable device, and enhances the real-time performance and intelligent level of the device.

Description

Technical Field

[0001] This invention belongs to the technical field of machine learning, and in particular relates to a method and device for detecting abnormalities in human physiological and motor states. Background Technology

[0002] Wearable devices, as mechanical devices capable of sensing and enhancing human movement capabilities, have been widely used in recent years in fields such as medical rehabilitation, assisted walking, and physical strength enhancement. Particularly in the field of medical rehabilitation, wearable devices can monitor the body's condition, helping patients with limited mobility to perform daily activities, thereby improving their quality of life.

[0003] However, effectively detecting and handling abnormal states that may occur during human movement when using wearable devices has become a crucial problem that urgently needs to be solved. Traditional methods of analyzing vital signs often struggle to accurately identify a small number of abnormal states, such as (backward) falls, (sudden) turns, and (accidental) falls. These abnormal states not only involve changes in movement patterns but also involve comprehensive changes in various individual physiological and motor parameters, leading to low accuracy of traditional methods for analyzing abnormal vital signs. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method and apparatus for detecting abnormalities in human physiological and motor states, in order to solve the technical problem of low accuracy in traditional abnormal physical sign data analysis methods.

[0005] A first aspect of this invention provides a method for detecting abnormalities in human physiological and motor states. This method is applied to a processor in a wearable device, which includes a vital sign acquisition device and a processor. The method for detecting abnormalities in human physiological and motor states includes:

[0006] Acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training;

[0007] The real-time vital sign data are semantically classified to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states.

[0008] Based on the standard vital sign data corresponding to the real-time semantic information, determine whether the real-time vital sign data shows an abnormal state.

[0009] Furthermore, the step of determining whether the real-time vital sign data exhibits an abnormal state based on the standard vital sign data corresponding to the real-time semantic information includes:

[0010] Obtain the standard vital sign data corresponding to the real-time semantic information;

[0011] Calculate the distance between the real-time vital signs data and the standard vital signs data;

[0012] If the distance is greater than the threshold, then the real-time vital signs data are determined to be in an abnormal state.

[0013] Furthermore, before the step of acquiring the real-time vital sign data collected by the vital sign acquisition device and the standard vital sign data corresponding to various preset states obtained through pre-training, the method further includes:

[0014] Acquire the first training vital sign data collected by the vital sign collector; the first training vital sign data is the vital sign data of the target user in a normal state collected by the vital sign collector;

[0015] Acquire first historical vital sign data corresponding to each of multiple preset states; the first historical vital sign data is collected by the vital sign collector from the vital sign data of other users in the preset states;

[0016] The first training vital sign data and the first historical vital sign data are input into the target generation model to obtain the standard vital sign data output by the generation model.

[0017] Further, the step of inputting the first training vital sign data and the first historical vital sign data into the target generation model to obtain the standard vital sign data output by the generation model includes:

[0018] The first historical vital sign data is noise-added using the noise-adding network in the target generation model to obtain the first feature data.

[0019] Match the first training vital sign data corresponding to the real-time semantic information;

[0020] The first feature data is fused with the first training vital sign data corresponding to the real-time semantic information to obtain the second feature data;

[0021] The second feature data is denoised using the denoising network in the target generation model to obtain standard vital sign data.

[0022] Further, before the step of inputting the first training vital sign data and the first historical vital sign data into the target generation model to obtain the standard vital sign data output by the target generation model, the method further includes:

[0023] Acquire the second training vital sign data collected by the vital sign acquisition device;

[0024] Acquire the second historical vital sign data corresponding to each of the multiple preset states;

[0025] Obtain the initial generative model, initial semantic information, and label data;

[0026] Based on the second historical vital sign data, the initial semantic information, and the label data, the initial generation model is trained to obtain the target generation model.

[0027] Further, the step of training the initial generation model based on the second historical vital sign data, the initial semantic information, and the label data to obtain the target generation model includes:

[0028] The second historical vital sign data is noise-added using the noise-adding network in the initial generation model to obtain the third feature data;

[0029] Obtain the second training trait data corresponding to the initial semantic information;

[0030] The third feature data is fused with the second training symptom data corresponding to the initial semantic information to obtain the fourth feature data;

[0031] The fourth feature data is denoised using the denoising network in the initial generation model to obtain the output data.

[0032] Calculate the error between the output data and the label data, and adjust the initial generation model based on the error;

[0033] Based on different first historical vital sign data, the step of adding noise to the second historical vital sign data through the noise-adding network in the initial generation model to obtain the first feature data, and subsequent steps, are performed iteratively to obtain the target generation model.

[0034] Furthermore, the step of semantically classifying the real-time vital sign data to obtain semantic information includes:

[0035] The real-time vital signs data are input into the supervised model to obtain the semantic information output by the supervised model.

[0036] A second aspect of the present invention provides a device for detecting abnormalities in human physiological and motor states, comprising:

[0037] The acquisition unit is used to acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training.

[0038] A classification unit is used to perform semantic classification on the real-time vital sign data to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states.

[0039] The judgment unit is used to determine whether the real-time vital signs data is in an abnormal state based on the standard vital signs data corresponding to the real-time semantic information.

[0040] A third aspect of the present invention provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method for detecting abnormal human physiological and motor states described in the first aspect.

[0041] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for detecting abnormal human physiological and motor states described in the first aspect.

[0042] The beneficial effects of this invention compared to existing technologies are as follows: Utilizing pre-trained standard vital sign data corresponding to various preset states, this method can accurately identify and classify multiple abnormal movement states, such as falls, turning, and tumbling. This semantic classification-based detection method is more targeted than traditional methods, reducing the possibility of misjudgments and missed judgments. By performing semantic classification on real-time vital sign data collected by the vital sign collector, real-time semantic information of preset states can be quickly generated, thereby instantly determining whether abnormalities have occurred in the vital sign data. This real-time detection capability is of great significance for user safety and monitoring of movement status, especially in some high-risk sports or rehabilitation training. This method utilizes a pre-trained model to achieve accurate identification of complex movement states through semantic classification, demonstrating a high level of intelligent processing capabilities. The processor can autonomously analyze and judge movement states. This method is applicable to various types of wearable devices and can meet the needs of different users in various sports and daily activity scenarios. Whether used for rehabilitation training, assisted walking, or enhancing motor skills, it can provide reliable anomaly detection support. Through high-precision and real-time anomaly detection, it can promptly provide feedback on the user's movement status, prevent potential dangers, and thus improve the user's sense of security and user experience. Meanwhile, it reduces false alarms and increases user trust and reliance on the device. In summary, this technical solution significantly improves the accuracy of vital sign data analysis in wearable devices and enhances the device's real-time performance and intelligence level by introducing an anomaly detection method based on semantic classification. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A schematic flowchart of a method for detecting abnormalities in human physiological and motor states provided by the present invention is shown.

[0045] Figure 2 A schematic flowchart of another method for detecting abnormal human physiological and motor states provided by the present invention is shown.

[0046] Figure 3 A schematic flowchart of another method for detecting abnormal human physiological and motor states provided by the present invention is shown.

[0047] Figure 4 This diagram illustrates a human physiological and motor state abnormality detection device according to an embodiment of the present invention.

[0048] Figure 5 A schematic diagram of a terminal device provided in an embodiment of the present invention is shown. Detailed Implementation

[0049] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0050] This invention provides a method and apparatus for detecting abnormalities in human physiological and motor states, thereby addressing the technical problem of low accuracy in traditional methods for analyzing abnormal physical signs.

[0051] First, this invention provides a method for detecting abnormalities in human physiological and motor states. Please refer to [link / reference]. Figure 1 , Figure 1 A schematic flowchart of a method for detecting abnormalities in human physiological and motor states provided by the present invention is shown. Figure 1 As shown, the method for detecting abnormal human physiological and motor states may include the following steps: Step 101: Acquire the real-time vital sign data collected by the vital sign collector and the standard vital sign data corresponding to various preset states obtained through pre-training.

[0052] The method for detecting abnormal human physiological and motor states is applied to the processor in a wearable device, which includes a vital sign acquisition device and a processor. The vital sign acquisition device is used to collect human physiological and vital sign data. Application scenarios for the vital sign acquisition device include, but are not limited to, motion data acquisition in exoskeleton devices and physiological (snoring and sleep apnea, etc.) and motion detection (leg movements and sleepwalking, etc.) in sleep monitoring. That is, the application scenarios of this application are not limited to, but include, the fields of exoskeleton abnormal state detection and sleep abnormality monitoring, etc. No limitations are made on the application scenarios here; any application scenario suitable for the technical solution of this application is acceptable.

[0053] Vital signs acquisition devices include, but are not limited to, motion sensors and physiological monitoring devices. Vital signs acquisition devices include, but are not limited to, inertial navigation sensors, force sensors, and optical sensors. Physiological monitoring devices include, but are not limited to, electroencephalogram (EEG) monitors, electrooculogram (EOG) monitors, electromyogram (EMG) monitors, pulse oximeters, blood pressure monitors, electrocardiogram (ECG) monitors, respiration monitors, and nasal airflow monitors.

[0054] Vital signs data include real-time and standard vital signs data. Vital signs data includes, but is not limited to, physiological data, kinetic data, and kinematic data. Kinematic data includes, but is not limited to, one or more combinations of data such as position, velocity, acceleration, and angular velocity. Kinetic data includes, but is not limited to, one or more combinations of data such as force, torque, center of mass position, and torque. Physiological data includes, but is not limited to, one or more combinations of data such as physiological electrical signals, respiration, heart rate, blood oxygen, blood pressure, air pressure, and air temperature. Body parts monitored by kinematics include, but are not limited to, toes, heels, calves, knees, thighs, hands, forearms, elbows, upper arms, shoulders, and hips.

[0055] For example, in the application scenario of exoskeleton abnormality detection, the vital signs data are kinematic and / or kinetic data. In the application scenario of sleep disorder monitoring, the vital signs data are kinematic and physiological data. The vital signs data corresponding to different application scenarios can be selected according to actual needs, and no restrictions are imposed here.

[0056] Wearable devices include, but are not limited to, exoskeleton devices, physiological monitoring devices, and biomechanical monitoring devices. The combination of one or more devices can be determined based on the actual application scenario and is not limited here.

[0057] Different preset states correspond to different standard vital sign data. Standard vital sign data includes, but is not limited to, standard exercise data, standard physiological data, and standard biomechanical data. Standard exercise data includes, but is not limited to, normal walking data, turning data, jumping data, squatting data, and climbing stairs data. Standard physiological data includes normal breathing data and normal heart rate data. Standard biomechanical data includes normal torque data and normal moment data.

[0058] Taking standard movement data as an example, normal walking data is relatively easy to obtain for different users (by guiding the user to walk normally during the first use of the device, and then obtaining the corresponding walking data), while turning data, jumping data, squatting data, and climbing stairs data are more difficult to obtain. Therefore, this application uses data synthesis technology to generate personalized data (standard vital sign data) for different users. For example, by combining the current user's normal walking data with the jumping data of other users, the current user's jumping data can be obtained. Personalized data treats each sample as an individual with independent characteristics, aiming to synthesize data across semantics. For example, a large amount of normal walking data of a user can be combined with standard vital sign data of other people with target semantics (target semantics include but are not limited to turning, jumping, and squatting) to synthesize the user's standard vital sign data. The advantage of doing so is that it can combine an individual's normal vital sign data with a large amount of standard vital sign data of other individuals with target semantics, thus eliminating the need to collect all of an individual's standard vital sign data. This is different from traditional predictive models, which often focus on generating data with the same semantics within a short time window. For example, for a patient with hip compensatory force, his walking posture may be considered abnormal under conventional standards, but it may be normal for him personally. Therefore, it is necessary to use walking data containing his unique walking posture to predict standard vital signs data, which can effectively reduce the problems of poor quality and insufficient continuity of the generated data.

[0059] The specific method for synthesizing standard vital sign data is as follows:

[0060] Optionally, before step 101, steps A1 to A3 are also included:

[0061] Step A1: Obtain the first training vital sign data collected by the vital sign collector; the first training vital sign data is the vital sign data of the target user in a normal state collected by the vital sign collector;

[0062] It is worth noting that step 101 is executed during the daily use of the wearable device, where real-time vital sign data is collected for anomaly detection. Step A1, on the other hand, is used to initialize the wearable device to train the deep learning model within it, thereby adapting to the physiological characteristics and walking habits of different users.

[0063] The first training vital signs data includes, but is not limited to, motion characteristic data, physiological characteristic data, and mechanical data.

[0064] Taking wearable devices as exoskeletons as an example, during the initialization phase of the wearable device, the user needs to wear the wearable device and walk normally according to their own walking habits. At this time, the vital signs collector collects the first training vital signs data.

[0065] Step A2: Obtain the first historical vital sign data corresponding to each of the multiple preset states; the first historical vital sign data is collected by the vital sign collector from the vital sign data of other users in the preset states;

[0066] The first historical vital sign data is the historical standard vital sign data of other users, or it can be the standard vital sign data set by humans, which is used to merge and generate the standard vital sign data of the current user.

[0067] Step A3: Input the first training vital sign data and the first historical vital sign data into the target generation model to obtain the standard vital sign data output by the target generation model.

[0068] The target generation model fuses the first training vital sign data and the first historical vital sign data. The specific logic is as follows:

[0069] Specifically, step A3 includes steps A31 to A34:

[0070] Step A31: Noise is added to the first historical vital sign data through the noise-adding network in the target generation model to obtain the first feature data;

[0071] By gradually adding noise to the first historical vital sign data through a noisy network, the first feature data is obtained, making it gradually become random.

[0072] Step A32: Match the first training vital sign data corresponding to the real-time semantic information;

[0073] Step A33: Fuse the first feature data with the first training symptom data corresponding to the real-time semantic information to obtain the second feature data;

[0074] Step A34: Denoise the second feature data using the denoising network in the target generation model to obtain standard vital sign data.

[0075] The denoising network gradually removes noise from the second feature data to obtain standard vital sign data.

[0076] The system collects first training vital sign data of the target user while walking normally using a vital sign collector, and obtains first historical vital sign data corresponding to multiple preset states. This historical data is composed of vital sign data from other users in the preset states. By collecting vital sign data from different users in preset states, a diverse standard vital sign dataset can be established. The first historical vital sign data is then denoised using a noisy network in the target generation model to generate first feature data. Noisy processing effectively simulates the uncertainty and noise characteristics in the data, enabling the model to better handle various complex situations in real-world applications. The first training vital sign data corresponding to real-time semantic information is matched and fused with the denoised first feature data to obtain second feature data. This process combines the features of normal vital sign data and standard vital sign data, integrating the user's normal vital sign data with the standard vital sign data of others, thereby generating personalized standard vital sign data for the user. The second feature data is then denoised using a denoising network in the target generation model to obtain the final standard vital sign data. The denoising process filters out noise and interference in the data, making the extracted standard vital sign data more accurate and reliable. In summary, this technical solution, by fusing training data from normal walking with historical data from preset states and utilizing noise-adding and denoising networks, can effectively improve the accuracy and robustness of abnormal motion state detection. This method can also adapt to individual differences among different users.

[0077] Objective generative models, through a progressive denoising process, can generate highly diverse samples. Due to the randomness of their generation process, objective generative models can generate different standard characteristic data given the same latent variable as input. Objective generative models support guided diversity, meaning that the generation process can be guided by adding additional conditional information (such as semantic information), improving the quantity, quality, and consistency of generated samples. Compared to Generative Adversarial Networks (GANs), the training process of objective generative models is more stable. Training GANs often requires complex balancing techniques to counteract the generator and discriminator, while objective generative models, through explicit loss functions and progressive denoising, are simpler and more stable to train, reducing training difficulty and instability. The training logic of objective generative models is as follows:

[0078] Optionally, steps B1 to B4 are included before step A34:

[0079] Step B1: Obtain the second training vital sign data collected by the vital sign collector;

[0080] It is worth noting that the execution time of step 101 is to collect real-time vital sign data for anomaly detection during daily use of the wearable device. Step B1 is used to train the target generation model. The second training vital sign data can be the training vital sign data of any other user, collected by the vital sign collector while the user wears the wearable device and walks normally according to their own walking habits.

[0081] Step B2: Obtain the second historical vital sign data corresponding to each of the multiple preset states;

[0082] The second historical vital signs data can be the historical vital signs data of any other user.

[0083] Step B3: Obtain the initial generative model, initial semantic information, and label data;

[0084] The initial generative model refers to the model whose parameters are initialized. The initial semantic information consists of multiple preset states. The label data consists of the labels corresponding to each of the multiple initial semantic information states.

[0085] Step B4: Based on the second historical vital sign data, the initial semantic information, and the label data, train the initial generation model to obtain the target generation model.

[0086] Specifically, step B4 includes steps B41 to B46:

[0087] Step B41: Noise is added to the second historical vital sign data through the noise-adding network in the initial generation model to obtain the third feature data;

[0088] Step B42: Obtain the second training trait data corresponding to the initial semantic information;

[0089] Step B43: Fuse the third feature data with the second training symptom data corresponding to the initial semantic information to obtain the fourth feature data;

[0090] Step B44: Denoise the fourth feature data using the denoising network in the initial generation model to obtain output data;

[0091] Step B45: Calculate the error between the output data and the label data, and adjust the initial generation model based on the error;

[0092] Step B46: Based on different first historical vital sign data, iteratively execute the step of adding noise to the second historical vital sign data through the noise-adding network in the initial generation model to obtain the first feature data, and subsequent steps, to obtain the target generation model.

[0093] The target generation model can also be VAE, GAN, diffusion model and its variants, and the feature extraction network includes but is not limited to CNN, LSTM, U-net, Transformer and its variants.

[0094] As an optional embodiment of this application, for the target generation model, daily vital sign data can be collected during subsequent user use, and the daily vital sign data can be used as the second training vital sign data in step B43 to update and train the model, so that the target generation model can better meet the user's needs.

[0095] In this embodiment, by acquiring second historical vital sign data corresponding to multiple preset states and adding noise to them, the model can learn and adapt to different preset states during training, thereby improving the model's robustness and generalization ability when facing abnormal situations in real application scenarios. By introducing a noise-adding network and a denoising network into the initial generation model, the model can effectively remove noise and extract useful features when processing noisy data. This process improves the model's adaptability and accuracy in complex environments. By fusing third feature data with second training vital sign data corresponding to the initial semantic information, the model can more accurately capture and reflect the motion features collected by the vital sign collector, improving the model's understanding and generation ability of vital sign data. During training, the error between the output data and the label data is calculated, and the initial generation model is adjusted based on the error, enabling the model to continuously optimize and improve its performance, gradually approaching the ideal state of the target generation model. By iteratively training based on different first historical vital sign data, the model can better utilize the information in the historical data, improving the model's understanding and processing ability of the diversity and complexity of vital sign data. In summary, the proposed technical solution significantly improves the performance and effectiveness of the target generation model in processing vital sign data through a series of innovative processing and optimization steps.

[0096] Step 102: Perform semantic classification on the real-time vital sign data to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states;

[0097] The real-time vital sign data is input into a supervised model to obtain semantic information output by the supervised model. Supervised models include, but are not limited to, CNN, LSTM, GNN, Transformer, and their variants.

[0098] Step 103: Determine whether the real-time vital signs data is in an abnormal state based on the standard vital signs data corresponding to the real-time semantic information.

[0099] This application performs semantic classification on real-time vital sign data to obtain real-time semantic information for preset states. Then, it matches multiple corresponding standard vital sign data (data subsets) based on the real-time semantic information. Different preset states correspond to different subsets. This data subset classification significantly reduces computation (eliminating the need for matching calculations with all data). The data subsets are generated based on training data using a target generation model; that is, the target generation model pre-classifies different preset states into multiple data subsets, and the data distribution of these subsets exhibits high consistency and stability. The semantic classification logic for real-time vital sign data is as follows:

[0100] Specifically, step 103 includes steps 1031 to 1033:

[0101] Step 1031: Obtain the standard vital sign data corresponding to the real-time semantic information;

[0102] The standard vital sign data corresponding to different preset states are obtained through pre-training of the target generation model. Step 1032: Calculate the distance between the real-time vital sign data and the standard vital sign data;

[0103] Methods for calculating the distance between real-time vital signs data and standard vital signs data include, but are not limited to, Euclidean distance, cosine similarity, Manhattan distance, Minkowski distance, Mahalanobis distance, Chebyshev distance combined with threshold determination, as well as using Weibull distribution, parametric Sigmoid distribution, generalized Pareto distribution, Frechet distribution, Gumbel distribution combined with confidence determination.

[0104] Step 1033: If the distance is greater than the threshold, then it is determined that the real-time vital signs data is in an abnormal state.

[0105] By calculating the distance between real-time vital sign data and predefined standard vital sign data, this solution can accurately determine the presence of abnormal states. Using distance as the judgment criterion ensures the objectivity of the judgment, reduces the probability of false positives and false negatives, and improves the reliability of the detection results. Real-time acquisition of vital sign data and distance calculation can quickly determine the presence of abnormalities at the moment of movement. This real-time detection capability plays a crucial role in monitoring user safety and movement status, especially in high-risk sports or rehabilitation training, enabling timely warnings and appropriate measures. By setting appropriate thresholds, the detection sensitivity can be flexibly adjusted according to different application scenarios and user needs. For example, a lower threshold can be set in high-risk environments to quickly respond to potential dangers; a higher threshold can be set in normal sports to avoid excessive false alarms, thereby optimizing the user experience. This solution utilizes real-time semantic information and standard vital sign data for calculation, achieving a high level of intelligent processing. The processor can autonomously analyze and judge the movement status, not only improving the automation of detection but also reducing reliance on manual monitoring, demonstrating a high degree of intelligence. This technical solution is applicable to various types of wearable devices and other sports monitoring devices, meeting the needs of different users in various sports and daily activity scenarios. Whether used for rehabilitation training, assisted walking, or enhancing motor skills, this device provides reliable anomaly detection support. Through high-precision and real-time anomaly detection, it can promptly report the user's movement status, preventing potential dangers and thus improving user safety and experience. Simultaneously, it reduces false alarms, increasing user trust and reliance on the device. In summary, this technical solution significantly improves the accuracy and real-time performance of anomaly detection by calculating the distance between real-time and standard vital sign data, enhancing the device's intelligence and applicability.

[0106] Abnormal conditions include, but are not limited to, motion abnormalities, physiological abnormalities, and mechanical abnormalities. Motion abnormalities include, but are not limited to, falls, sudden stops, and drops. Physiological abnormalities include, but are not limited to, respiratory abnormalities and heart rate abnormalities. Mechanical abnormalities include, but are not limited to, torque abnormalities and torque-force abnormalities.

[0107] Abnormal data corresponding to abnormal states are usually too small to be used as a class for independent training. They are characterized by rarity and lack of cases, and can have a negative impact on the operation of many systems, such as unnatural and sudden movement data in the gait signals of the lower limb exoskeleton due to the influence of human body state and external environment.

[0108] It's important to note that in real-world applications, for preset falls, sudden stops, drops, and pathological conditions, if data collection and wearable device calibration are possible, these processed data are also considered "normal," i.e., used as standard vital sign data. For example, with falls, preset falls are often forward falls; backward falls have an extremely low probability of occurrence, making them uncollectible. Therefore, these anomalies are usually too small to be used as a separate training data category, possessing characteristics such as rarity and lack of precedent.

[0109] In this embodiment, by utilizing pre-trained standard vital sign data corresponding to various preset states, this method can accurately identify and classify multiple abnormal movement states, such as falls, turning, and tumbling. This semantic classification-based detection method is more targeted than traditional methods, reducing the possibility of false positives and false negatives. By performing semantic classification on the real-time vital sign data collected by the vital sign collector, real-time semantic information of preset states can be quickly generated, thereby instantly determining whether the vital sign data is abnormal. This real-time detection capability is of great significance for user safety and monitoring of movement status, especially in some high-risk sports or rehabilitation training. This method utilizes a pre-trained model to achieve accurate identification of complex movement states through semantic classification, demonstrating a high level of intelligent processing capabilities. The processor can autonomously analyze and judge movement states. This method is applicable to various types of wearable devices and can meet the needs of different users in various sports and daily activity scenarios. Whether used for rehabilitation training, assisted walking, or enhancing motor skills, it can provide reliable anomaly detection support. Through high-precision and real-time anomaly detection, it can promptly provide feedback on the user's movement status, prevent potential dangers, and thus improve the user's sense of security and user experience. Meanwhile, it reduces false alarms and increases user trust and reliance on the device. In summary, this technical solution significantly improves the accuracy of vital sign data analysis in wearable devices and enhances the device's real-time performance and intelligence level by introducing an anomaly detection method based on semantic classification.

[0110] like Figure 4 This invention provides a device for detecting abnormalities in human physiological and motor states. Please refer to [link / reference]. Figure 4 , Figure 4 This diagram illustrates a human physiological and motor state abnormality detection device provided by the present invention, as shown below. Figure 4 The device shown includes a human physiological and motor state abnormality detection device.

[0111] The acquisition unit 21 is used to acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training.

[0112] Classification unit 22 is used to perform semantic classification on the real-time vital sign data to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states;

[0113] The judgment unit 23 is used to determine whether the real-time vital signs data is in an abnormal state based on the standard vital signs data corresponding to the real-time semantic information.

[0114] This invention provides a device for detecting abnormal human physiological and motor states. Utilizing pre-trained standard vital sign data corresponding to various preset states, this method can accurately identify and classify multiple abnormal motor states, such as falls, turning, and tumbling. This semantic classification-based detection method is more targeted than traditional methods, reducing the possibility of false positives and false negatives. By performing semantic classification on real-time vital sign data collected by the vital sign collector, real-time semantic information for preset states can be quickly generated, thereby instantly determining whether abnormalities have occurred in the vital sign data. This real-time detection capability is of great significance for user safety and monitoring of motor state, especially in high-risk sports or rehabilitation training. This method utilizes a pre-trained model to achieve accurate identification of complex motor states through semantic classification, demonstrating a high level of intelligent processing capabilities. The processor can autonomously analyze and judge the motor state. This method is applicable to various types of wearable devices, meeting the needs of different users in various sports and daily activities. Whether used for rehabilitation training, assisted walking, or enhancing motor ability, it provides reliable abnormality detection support. Through high-precision and real-time abnormality detection, it can promptly provide feedback on the user's motor state, preventing potential dangers and thus improving the user's sense of security and user experience. Meanwhile, it reduces false alarms and increases user trust and reliance on the device. In summary, this technical solution significantly improves the accuracy of vital sign data analysis in wearable devices and enhances the device's real-time performance and intelligence level by introducing an anomaly detection method based on semantic classification.

[0115] Figure 5 This is a schematic diagram of a terminal device provided in an embodiment of the present invention. Figure 5 As shown, a terminal device 3 in this embodiment includes: a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and executable on the processor 30, such as a human physiological and motor state abnormality detection program. When the processor 30 executes the computer program 32, it implements the steps in each of the above embodiments of the human physiological and motor state abnormality detection method, for example... Figure 1 Steps 101 to 105 are shown. Alternatively, when the processor 30 executes the computer program 32, it implements the functions of each unit in the above-described device embodiments, for example... Figure 2 The function of the unit shown.

[0116] For example, the computer program 32 can be divided into one or more units, which are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 32 in the terminal device 3. For example, the specific functions of each unit of the computer program 32 can be divided as follows:

[0117] The acquisition unit is used to acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training.

[0118] A classification unit is used to perform semantic classification on the real-time vital sign data to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states.

[0119] The judgment unit is used to determine whether the real-time vital signs data is in an abnormal state based on the standard vital signs data corresponding to the real-time semantic information.

[0120] The terminal device includes, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of a terminal device 3 and does not constitute a limitation on a terminal device 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0121] The processor 30 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0122] The memory 31 can be an internal storage unit of the terminal device 3, such as a hard disk or memory of the terminal device 3. The memory 31 can also be an external storage device of the terminal device 3, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the terminal device 3. Furthermore, the memory 31 can include both internal and external storage units of the terminal device 3. The memory 31 is used to store the computer program and other programs and data required by the roaming control device. The memory 31 can also be used to temporarily store data that has been output or will be output.

[0123] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0124] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0125] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments 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 as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0126] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0127] This invention provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps described in the above-described method embodiments.

[0128] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0129] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0130] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0131] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0132] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units.

[0133] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0134] It should also be understood that the term “and / or” as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0135] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if [the described condition or event] is detected" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once [the described condition or event] is detected," or "in response to detection of [the described condition or event]."

[0136] Furthermore, in the description of this invention and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0137] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0138] The above-described 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for detecting abnormalities in human physiological and motor states, characterized in that, The method for detecting abnormal human physiological and motor states is applied to a processor in a wearable device, the wearable device including a vital sign collector and a processor, and the method for detecting abnormal human physiological and motor states includes: Acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training; The real-time vital sign data are semantically classified to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states. Based on the standard vital sign data corresponding to the real-time semantic information, determine whether the real-time vital sign data is in an abnormal state; The step of semantically classifying the real-time vital sign data to obtain real-time semantic information includes: The real-time vital signs data are input into the supervisory model to obtain the real-time semantic information output by the supervisory model; Prior to the step of acquiring the real-time vital sign data collected by the vital sign collector and the standard vital sign data corresponding to various preset states obtained through pre-training, the method further includes: Acquire the first training vital sign data collected by the vital sign collector; the first training vital sign data is the vital sign data of the target user in a normal state collected by the vital sign collector; Acquire first historical vital sign data corresponding to each of multiple preset states; the first historical vital sign data is collected by the vital sign collector from the vital sign data of other users in the preset states; The first training vital sign data and the first historical vital sign data are input into the target generation model to obtain the standard vital sign data output by the target generation model. The step of inputting the first training vital sign data and the first historical vital sign data into the target generation model to obtain the standard vital sign data output by the target generation model includes: The first historical vital sign data is noise-added using the noise-adding network in the target generation model to obtain the first feature data. Match the first training vital sign data corresponding to the real-time semantic information; The first feature data is fused with the first training vital sign data corresponding to the real-time semantic information to obtain the second feature data; The second feature data is denoised using the denoising network in the target generation model to obtain standard vital sign data.

2. The method for detecting abnormalities in human physiological and motor states as described in claim 1, characterized in that, The step of determining whether the real-time vital signs data exhibits an abnormal state based on the standard vital signs data corresponding to the real-time semantic information includes: Obtain the standard vital sign data corresponding to the real-time semantic information; Calculate the distance between the real-time vital signs data and the standard vital signs data; If the distance is greater than the threshold, then the real-time vital signs data are determined to be in an abnormal state.

3. The method for detecting abnormalities in human physiological and motor states as described in claim 1, characterized in that, Before the step of inputting the first training vital sign data and the first historical vital sign data into the target generation model to obtain the standard vital sign data output by the target generation model, the method further includes: Acquire the second training vital sign data collected by the vital sign acquisition device; Acquire the second historical vital sign data corresponding to each of the multiple preset states; Obtain the initial generative model, initial semantic information, and label data; Based on the second historical vital sign data, the initial semantic information, and the label data, the initial generation model is trained to obtain the target generation model.

4. The method for detecting abnormalities in human physiological and motor states as described in claim 3, characterized in that, The step of training the initial generation model based on the second historical vital sign data, the initial semantic information, and the label data to obtain the target generation model includes: The second historical vital sign data is noise-added using the noise-adding network in the initial generation model to obtain the third feature data; Obtain the second training trait data corresponding to the initial semantic information; The third feature data is fused with the second training symptom data corresponding to the initial semantic information to obtain the fourth feature data; The fourth feature data is denoised using the denoising network in the initial generation model to obtain the output data. Calculate the error between the output data and the label data, and adjust the initial generation model based on the error; Based on different first historical vital sign data, the step of adding noise to the second historical vital sign data through the noise-adding network in the initial generation model to obtain the third feature data, and subsequent steps, are performed iteratively to obtain the target generation model.

5. A device for detecting abnormalities in human physiological and motor states, wherein the device is used to implement the steps in the method for detecting abnormalities in human physiological and motor states as described in claim 1, characterized in that, The human physiological and motor state abnormality detection device includes: The acquisition unit is used to acquire real-time vital sign data collected by the vital sign collector and standard vital sign data corresponding to various preset states obtained through pre-training. A classification unit is used to perform semantic classification on the real-time vital sign data to obtain real-time semantic information; each piece of real-time semantic information corresponds to one or more of the preset states. The judgment unit is used to determine whether the real-time vital signs data is in an abnormal state based on the standard vital signs data corresponding to the real-time semantic information.

6. A terminal device, characterized in that, The terminal device includes: a memory, a processor, and a human physiological and motor state abnormality detection program stored in the memory and executable on the processor, wherein the human physiological and motor state abnormality detection program is configured to implement the steps in the human physiological and motor state abnormality detection method as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps in the method for detecting abnormal human physiological and motor states as described in any one of claims 1 to 4.

Images