Personalized rehabilitation nursing method based on ecological momentary assessments, and terminal device

Abstract

A personalized rehabilitation nursing method based on ecological momentary assessments, the method comprising: acquiring historical health information of a patient, wherein the historical health information comprises instant situational information of the patient at a plurality of moments within a past period of time and feedback information from ecological momentary assessments, the instant situational information comprising time information and / or location information of the patient; on the basis of the historical health information and current health information of the patient, determining health prompt information matching the current health information, wherein the current health information comprises instant situational information of the patient at the current moment and / or feedback information from ecological momentary assessments, and the health prompt information comprises notification information about a health problem that may be occurring to the patient at the current moment and / or autonomous intervention information targeting the health problem; and outputting the health prompt information. By means of the method, the nursing requirements of a patient can be accurately responded to in real time, thereby improving the timeliness and accuracy of rehabilitation nursing.

Description

Personalized rehabilitation nursing methods and terminal equipment based on instantaneous ecological assessment

[0001] This application claims priority to Chinese Patent Application No. 202510042614.4, filed on January 10, 2025, entitled "Personalized Rehabilitation Nursing Method and Terminal Equipment Based on Ecological Instantaneous Assessment", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of rehabilitation nursing technology, and in particular to a personalized rehabilitation nursing method and terminal device based on ecological instantaneous assessment. Background Technology

[0003] Patients with certain diseases often require rehabilitation nursing services after treatment or during their illness to promote functional recovery and improve their quality of life. Because patients' nursing needs during the rehabilitation phase may change dynamically over time, rehabilitation nursing services need to be able to respond to these needs in real time. Technical issues

[0004] The purpose of this application is to provide a personalized rehabilitation nursing method, rehabilitation nursing system, terminal equipment, computer-readable storage medium, and computer program product based on instantaneous ecological assessment. Technical solutions

[0005] The technical solution adopted in the embodiments of this application is:

[0006] In a first aspect, a personalized rehabilitation nursing method based on instantaneous ecological assessment is provided, comprising: acquiring the patient's historical health information, wherein the historical health information includes the patient's real-time contextual information and feedback information from instantaneous ecological assessment at multiple moments over a past period, the real-time contextual information including time information and / or the patient's location information; determining health prompt information matching the current health information based on the historical health information and the patient's current health information, wherein the current health information includes: the patient's real-time contextual information and / or feedback information from instantaneous ecological assessment at the current moment, the health prompt information including notification information of potential health problems the patient may encounter at the current moment and / or self-intervention information for the health problems; and outputting the health prompt information.

[0007] Secondly, a personalized rehabilitation nursing system based on instantaneous ecological assessment is provided, comprising: a first acquisition module for acquiring the patient's historical health information, wherein the historical health information includes the patient's real-time contextual information and feedback information from instantaneous ecological assessment at multiple moments over a past period, the real-time contextual information including time information and / or the patient's location information; a first determination module for determining health prompt information matching the current health information based on the historical health information and the patient's current health information, wherein the current health information includes: the patient's real-time contextual information and / or feedback information from instantaneous ecological assessment at the current moment, the health prompt information including notification information of potential health problems the patient may encounter at the current moment and / or self-intervention information for the health problems; and an output module for outputting the health prompt information.

[0008] Thirdly, a terminal device is provided, 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 above-mentioned personalized rehabilitation care method based on ecological instantaneous assessment.

[0009] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the aforementioned personalized rehabilitation care method based on instantaneous ecological assessment.

[0010] Fifthly, a computer program product is provided that, when run on a terminal device, enables the terminal device to implement the steps in the aforementioned personalized rehabilitation care method based on instantaneous ecological assessment. Beneficial effects

[0011] The beneficial effects of the personalized rehabilitation nursing method based on instantaneous ecological assessment provided in this application are as follows: It can promptly and accurately determine health alerts matching the patient's current health information based on the patient's historical and current health information, and can output these alerts in a timely manner, enabling patients to pay attention to potential health problems and effectively manage themselves. On the one hand, it reduces the cost of manual intervention by professional nursing staff, resulting in lower nursing costs and better applicability of the plan. On the other hand, because it can automatically and promptly determine health alerts matching the patient's current health information, it can respond to the patient's nursing needs in real time and accurately. Therefore, it can improve the timeliness and accuracy of rehabilitation nursing, and further improve the patient's rehabilitation effect and accelerate the rehabilitation process. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or exemplary technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 shows a flowchart of a personalized rehabilitation care method based on instantaneous ecological assessment provided in one embodiment of this application;

[0014] Figure 2 shows a schematic diagram of a list of health alert information in different scenarios provided in an embodiment of this application;

[0015] Figure 3 shows a flowchart of a personalized rehabilitation care method based on instantaneous ecological assessment provided in another embodiment of this application;

[0016] Figure 4 shows a schematic diagram of the structure of a personalized rehabilitation care system based on instantaneous ecological assessment provided in one embodiment of this application;

[0017] Figure 5 shows a schematic diagram of the structure of a terminal device provided in one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] 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 this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0020] It should be understood that, when used in this application 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 a collection thereof.

[0021] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0022] Furthermore, in the description of this application 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.

[0023] References to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application 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 all mean "including but not limited to," unless otherwise specifically emphasized. To illustrate the technical solutions provided by this application, a detailed description is provided below in conjunction with specific drawings and embodiments.

[0024] Current rehabilitation nursing methods typically involve professional medical caregivers conducting periodic (e.g., monthly) static assessments of patients, followed by adjustments to the nursing plan based on the assessment results, and the provision of periodic nursing recommendations. This approach relies on professional medical caregivers, incurs high labor costs, and struggles to provide real-time responses to patients' nursing needs.

[0025] Taking neurological diseases (such as stroke, cerebral infarction, cerebral hemorrhage, traumatic brain injury, Parkinson's disease, cerebral palsy, etc.) as an example, these diseases may cause health problems such as limb movement disorders, speech disorders, and cognitive impairments, and may also be accompanied by emotional health problems such as depression. This necessitates targeted nursing care for each patient's specific health issues. Furthermore, for the same patient, the health problems requiring attention may differ at different times, or the severity of the same health problem may vary at different times. For example, for stroke patients with depressive symptoms, these symptoms are often intermittent. During periods of emotional stability (when depressive symptoms are not present), emotional intervention is usually unnecessary. However, intervention is required when depressive symptoms are about to occur or are occurring. The severity of these depressive symptoms can also vary. Because professional medical care personnel can only conduct periodic static assessments of patients, the assessment results are directly related to the patient's state at the time of assessment. Therefore, these static assessment results cannot accurately reflect the patient's state at other times, and the nursing plans or recommendations developed accordingly are often inaccurate. If patients are cared for according to predetermined nursing plans or recommendations during disease flare-ups, it is highly likely to result in ineffective care, over-care (e.g., taking medication when it is not needed), or even incorrect care. Therefore, existing rehabilitation nursing plans struggle to respond to patients' needs in real time, potentially leading to prolonged rehabilitation periods, poor rehabilitation outcomes, and even adverse effects on patients' health (e.g., worsening of their condition). To at least partially address these technical problems, embodiments of this application provide a personalized rehabilitation nursing method based on instantaneous ecological assessment, a personalized rehabilitation nursing system based on instantaneous ecological assessment, a computer-readable storage medium, and a computer program product. This system can promptly and accurately determine health alerts matching the patient's current health information based on their historical and current health information, and can output these alerts in a timely manner to enable patients to pay attention to their current health problems and engage in effective self-care.

[0026] First, this application provides a personalized rehabilitation nursing method based on instantaneous ecological assessment. This method is applicable to the rehabilitation nursing of patients with various diseases, including but not limited to neurological diseases (such as stroke, cerebral infarction, cerebral hemorrhage, traumatic brain injury, Parkinson's disease, cerebral palsy, etc.), visceral diseases (such as coronary heart disease, essential hypertension, chronic obstructive pulmonary disease, diabetes, etc.), psychological diseases (such as depression, anxiety, phobias, obsessive-compulsive disorder, adjustment disorder, etc.), and mental illnesses (Alzheimer's disease, traumatic brain injury, epilepsy, schizophrenia, bipolar disorder, sleep disorders, etc.). For simplicity, the following description uses the rehabilitation nursing of a stroke patient as an example. Exemplarily, the personalized rehabilitation nursing method based on instantaneous ecological assessment provided in this application can be applied to various forms of terminal devices carried by the patient. In particular, it can be applied to the patient's portable terminal devices, such as smartphones, tablets, and wearable terminal devices such as smartwatches, smart glasses, and smart bracelets. Specifically, the terminal device may have an application installed (e.g., an APP named "Rehabilitation Nursing Partner"). When the application runs on the terminal device, it can implement the various steps of the personalized rehabilitation nursing method based on ecological instantaneous assessment provided in the embodiments of this application.

[0027] As shown in Figure 1, the personalized rehabilitation nursing method based on instantaneous ecological assessment provided in this application includes the following steps:

[0028] Step S110: Obtain the patient's historical health information, wherein the historical health information includes real-time contextual information and feedback information of ecological instantaneous assessment at multiple moments in the past period of time, and the real-time contextual information includes time information and / or the patient's location information.

[0029] In this embodiment, the patient's historical health information includes at least real-time contextual information and feedback information from ecological instantaneous assessments at multiple points within a past period. The past period can be set according to actual needs and can be longer than the onset cycle of each of the patient's symptoms (e.g., depression). For example, the past period could be 5 days, 7 days, 10 days, etc., before the formal commencement of rehabilitation care.

[0030] In this embodiment, the real-time contextual information may include time and location information for each ecological instantaneous assessment. For example, the network time can be obtained through the patient's mobile phone network, and the patient's location information can be obtained through the phone's GPS.

[0031] Ecological Momentary Assessment (EMA) is an assessment method that evaluates subjects’ behavior, attitudes, emotions, cognition, and other information in real time within the natural context of an event.

[0032] For example, the "Rehabilitation Care Companion" app on a patient's phone may include an EMA module. This module provides an instant ecological assessment interface, which may include an ecological assessment questionnaire. Patients can complete the questionnaire and input feedback information through interaction with the app interface. After the patient completes the questionnaire, the mobile app can receive the patient's ecological assessment feedback information in real time. Taking the ecological assessment of stroke patients as an example, the subjective questionnaire displayed on the interface can be used to quickly and accurately assess the patient's emotions, cognition, and physical activity in real time, collecting the patient's emotional feedback, cognitive feedback, physical activity feedback, social environment feedback, etc.

[0033] For example, for seven days prior to stroke patient rehabilitation care, the "Rehabilitation Care Partner" app installed on the stroke patient's mobile phone can collect feedback information on the patient's instantaneous ecological assessment (EMA) at different times each day, along with the assessment time and the patient's location. Optionally, for each of these seven days (e.g., from 9:00 AM to 9:00 PM), the app can pop up prompts such as "Please complete the EMA assessment" at a preset frequency to remind the patient to complete the EMA questionnaire multiple times a day. When the patient provides feedback, the app can automatically record the feedback time and the patient's location. Alternatively, for each of these seven days, the app can also randomly pop up prompts such as "Please complete the EMA assessment" to randomly obtain the patient's EMA feedback information at different times of the day. For example, prompts can be made at six different times each day for these seven days. That is, for seven consecutive days, the app collects the patient's real-time contextual information and EMA feedback information at six different times each day.

[0034] For example, a patient's historical health information may also include their medical records and nursing preferences. Medical records may include the patient's name, age, gender, occupation, address, medical history (e.g., stroke history), daily activity patterns, symptom categories, and an overview of various abilities (cognitive ability, physical mobility). Nursing preferences may include the patient's exercise preferences, dietary preferences, sleep preferences, EMA (Emergency Medical Advice) prompting time and frequency, and health alert prompting time and method (e.g., voice prompts, text prompts, visual prompts). In a specific example, the "Rehabilitation Nursing Companion" app installed on the patient's mobile phone may include a personal profile entry module. This module provides a profile entry window, allowing the patient to input their medical records and nursing preferences through interaction with the controls within the profile entry window.

[0035] For example, after collecting a patient's historical health information over a period of time, this information can be encrypted and stored in a structured manner on a cloud server, which can fully protect the security of the patient's personal information and facilitate subsequent analysis and personalized care for the patient.

[0036] Step S120: Based on the historical health information and the patient's current health information, determine health alert information that matches the current health information. The current health information includes: the patient's immediate contextual information and / or feedback information from an instantaneous ecological assessment at the current moment; the health alert information includes notification information about potential health problems the patient may encounter at the current moment and / or information on self-intervention for those health problems.

[0037] In this embodiment, during the rehabilitation care of a patient, the patient's current health information can be obtained in a timely manner. For example, the patient's immediate contextual information can be obtained in real time. For instance, the patient's location and time information can be obtained in real time via a mobile phone. For example, the timing and frequency of obtaining the patient's Ecological Instantaneous Assessment (EMA) feedback information can be determined based on the patient's care preferences. For example, after rehabilitation care, the EMA module can prompt the patient at least once a day to complete EMA feedback. The timing and frequency of the prompts can match the patient's preferred EMA prompt time and frequency. For example, at any time during the rehabilitation care process, the patient can also actively provide feedback on their health status through the EMA module. That is, during the rehabilitation care process, the EMA feedback information of the patient at the current moment can be obtained in response to the patient's operation on the Ecological Instantaneous Assessment interface. In this case, the patient can provide feedback on their current health status in one or more aspects according to their actual needs. For example, users can proactively use a mobile app to provide feedback on their current emotional state (e.g., feeling depressed, lonely, or anxious), cognitive state (e.g., completing a cognitive ability test questionnaire), physical activity state (e.g., what they are doing), or social environment state (e.g., who they are with).

[0038] In this embodiment, health alert information matching the patient's current health information can be determined by comprehensively considering the patient's historical health information and current health information. For example, at least based on the real-time acquired current health information, it can be determined whether the patient's current risk of developing a health problem meets the risk alert conditions. If so, health alert information matching the current health information can be determined by comprehensively considering the patient's historical health information and current health information. That is, health alert information matching the patient's current health information can be determined in real time when the patient's current risk of developing a health problem is high. In one example, if the real-time acquired feedback information of the patient's EMA indicates that the patient's current risk of developing a health problem is high, health alert information matching the current health information can be determined based on the patient's historical health information and current health information. In another example, if the patient's current risk of developing a health problem is determined by comprehensively considering the patient's historical health information (e.g., the patient's symptom occurrence pattern obtained by analyzing historical health information) and the real-time acquired immediate contextual information, health alert information matching the immediate contextual information can be determined based on the patient's historical health information (e.g., care preference information) and immediate contextual information. In another example, if the patient's historical health information and real-time contextual information determine that the patient is at high risk of developing a health problem, the patient can be further reminded to complete the EMA feedback. Then, based on the real-time EMA feedback information, the contextual information, and the patient's care preference information, health prompts that match the current situation and are consistent with the patient's preferences can be determined.

[0039] In this embodiment, the health alert information may include only notifications of potential health problems the patient may encounter at the current moment, only self-intervention information for potential health problems, or both. For example, the notification information in the health alert may include at least the content of the potential health problem, and may also include the cause and context of the health problem. For instance, a notification of "depression + loneliness" indicates that the patient experiences depression while feeling lonely. For example, the self-intervention information in the health alert may include self-care suggestions customized based on the patient's current situation. For example, if the patient reports feeling depressed at work, the app can send a work-related self-care reminder.

[0040] In this embodiment of the application, various suitable methods can be used to determine health prompt information that matches the current health information based on the historical health information and the patient's current health information.

[0041] In one example, statistical analysis and logical operations can be used to determine health alert information that matches the patient's current health information, based on the patient's historical health information and current health information. For instance, statistical analysis of the patient's historical health information can be performed to fit a mathematical model for determining health alert information. Then, the real-time acquired current health information can be matched with the mathematical model, and appropriate health alert information can be determined based on the matching result.

[0042] In another example, a pre-defined large language model can be used to determine health alert information that matches the current health information.

[0043] In one implementation, step S120 determines health prompt information matching the current health information based on the historical health information and the patient's current health information, including the following steps: step S121, using the patient's historical health information to train a preset large language model to obtain a trained large language model; step S122, inputting the patient's current health information into the trained large language model to obtain health prompt information matching the current health information.

[0044] For example, a pre-defined large language model can be trained using historical health information collected from stroke patients before the start of rehabilitation care (patient's medical records, nursing preferences, real-time contextual information collected over 7 consecutive days, and feedback from ecological instantaneous assessments) stored on a cloud server. Exemplarily, the pre-defined large language model can be an open-source large language model, such as the Baidu Large Language Model. Exemplarily, the Baidu Large Language Model can also be trained by combining historical health information from other stroke patients (e.g., all stroke patients participating in rehabilitation care) to construct a comprehensive information database related to stroke patient care. Exemplarily, the comprehensive information database can include diverse sets of reference prompts that can be associated with different stroke patients. Various suitable training methods can be used to train the large language model, enabling it to accurately infer health prompts that match the current health information of the current patient. In this way, during rehabilitation care, the real-time situational information of the patient and / or feedback information from the EMA can be input into the trained large language model. This allows for the generation of accurate health prompts inferred by the large language model in real time, which simultaneously match the patient's current situation, health status, and personalized needs. This improves the real-time nature, accuracy, and effectiveness of rehabilitation care.

[0045] Step S130: Output the health alert information.

[0046] In one example, after identifying a health alert that matches the current health information, the alert can be output immediately. For instance, if a patient is predicted to have a high risk of developing depression, a real-time alert can be sent to the patient's phone: "Depression, please take timely care." Another example is if a patient's symptoms are predicted to be severe, requiring timely intervention, a real-time self-intervention alert (such as an app alert: "Please take XX medication promptly") can be sent to the patient's phone. In another example, the timing of outputting health alerts can be determined based on the patient's care preferences. For instance, if a patient's symptoms are predicted to be mild, relevant self-care information can be sent during the patient's preferred self-care time. Similarly, if the self-intervention information is determined to be exercise intervention, the patient can be prompted to perform exercise during their preferred exercise time.

[0047] In this embodiment, the output of health alert information can take various forms, including but not limited to audio alerts, video alerts, text alerts, and visual alerts. The form of outputting health alert information can also be determined according to the patient's preferences to reduce interference with the patient and improve patient satisfaction.

[0048] The personalized rehabilitation nursing method based on instantaneous ecological assessment in this application can determine health tips that match the patient's current health information in a timely and accurate manner based on the patient's historical and current health information. This allows the method to output these tips promptly, enabling patients to pay attention to potential health problems and take effective self-care measures. On the one hand, it reduces the cost of manual intervention by professional nursing staff, resulting in lower nursing costs and better applicability of the approach. On the other hand, because it can automatically and promptly determine health tips that match the patient's current health information, it can respond to the patient's nursing needs in real time and accurately. Therefore, it can improve the timeliness and accuracy of rehabilitation nursing, further enhancing the patient's rehabilitation effect and accelerating the rehabilitation process.

[0049] In one implementation, the historical health information further includes the category information of the patient's condition. Step S120, which involves determining health alert information matching the current health information based on the historical health information and the patient's current health information, includes the following steps:

[0050] Step S1201: Determine the health group to which the patient belongs based on the group characteristic information in the historical health information, wherein the group characteristic information includes the category information of the patient's disease;

[0051] Step S1202: Based on the individual characteristic information in the historical health information and the current health information, select reference prompt information that matches the individual characteristic information and the current health information from the reference prompt information set corresponding to the health group to which the patient belongs, and determine the health prompt information that matches the current health information based on the selected reference prompt information. The individual characteristic information includes at least one of the patient's ecological instantaneous assessment of emotional feedback information, physical activity feedback information, cognitive feedback information, and social environment feedback information.

[0052] In this embodiment, the group characteristic information in the historical health information may include at least the patient's disease category information, or it may also include basic information such as the patient's age and gender. The patient's disease category information may include at least the category of the patient's symptoms, and may also include the category of the patient's health status (e.g., the category of stroke severity). It is understood that different stroke patients may have different symptoms; some patients have depressive symptoms, some have cognitive impairment symptoms, and some have physical dysfunction symptoms. Some patients have one of the above symptoms, while others may have multiple symptoms. In this embodiment, the health group to which the patient belongs can be determined at least based on the patient's disease category information. For patients with different disease category information, their respective health groups are different. For example, a patient with only cognitive impairment and a patient with only physical dysfunction belong to different health groups.

[0053] For example, in training a large language model using a patient's historical health information, the model can be jointly trained using the historical health information of other stroke patients, constructing a comprehensive information database related to stroke patient care. This comprehensive information database can include a set of reference prompts associated with each patient. For instance, patients can be divided into multiple healthy groups based on their symptom categories. Then, for each healthy group, a set of reference prompts corresponding to that group can be determined based on the symptom categories of each patient within that group, EMA feedback information in different contexts, and each patient's care preference information. Each set of reference prompts can include diverse reference prompts matching different contexts, different feedback information, and different care preferences.

[0054] For example, before or during rehabilitation care, the patient's health group can be determined based on the symptom categories in the patient's historical health information. During rehabilitation care, based on the patient's real-time situational information and / or received EMA feedback information, combined with individual characteristic information from the patient's historical health information (e.g., the patient's care preferences, previous EMA feedback information, and real-time situational information), optimal reference prompts can be selected from the set of reference prompts corresponding to the patient's health group. This ensures that the selected reference prompts match both the patient's current real-time situational information and / or received EMA feedback information, as well as the patient's care preferences. Subsequently, the selected reference prompts can be directly used as health prompts matching the current health information, or the selected reference prompts can be fine-tuned, and the fine-tuned reference prompts can be determined as health prompts matching the current health information.

[0055] The above solution can ensure that the health tips pushed to patients not only match the patient's situation but also fully meet the patient's personalized needs, and can also improve the real-time nature of health tips push.

[0056] In one embodiment, before determining the health group to which the patient belongs based on the group characteristic information in the historical health information in step S1201, the personalized rehabilitation nursing method based on ecological instantaneous assessment provided in this application embodiment further includes the following steps S101 to S103.

[0057] Step S101: Obtain historical health information for multiple patients.

[0058] Understandably, if the number of patients in step S110 is one, the number of patients in this step can include the patients in step S110. That is, step S101 can include the aforementioned step S110. The number of patients in this step can be the number of patients currently participating in rehabilitation care, for example, 60. Exemplarily, as the number of patients participating in rehabilitation care increases, the historical health information of newly added patients can also be continuously supplemented.

[0059] Step S102: Based on the disease category information of the multiple patients, the multiple patients are divided into multiple healthy groups, wherein the patients in each healthy group have the same disease category.

[0060] In a specific example, 60 stroke patients participating in rehabilitation care can be divided into seven healthy groups based on the symptom categories they exhibit: "Depression Group," "Cognitive Impairment Group," "Functional Impairment Group," "Depression + Cognitive Impairment Group," "Depressive Disorder + Functional Impairment Group," "Cognitive Impairment + Functional Impairment Group," and "Depressive + Cognitive + Functional Impairment Group." For instance, patients in the "Depression Group" only exhibit depressive symptoms; patients in the "Cognitive Impairment + Functional Impairment Group" exhibit both cognitive impairment symptoms and physical functional impairment symptoms.

[0061] Step S103: For each health group, determine the reference prompt information set corresponding to the health group based on the individual characteristic information in the historical health information of each patient in the health group. The reference prompt information set includes early warning notification information indicating that the patient has developed the first symptom, as well as various self-care suggestions for the patient when the first symptom occurs in different situations.

[0062] For example, a "depression group" includes 10 stroke patients. The individual characteristics of these 10 stroke patients' historical health information can be analyzed, and combined with Baidu's big data language model, a set of reference prompts can be generated for this health group. Specifically, the big data language model can be used to generate diverse reference prompts, including various health prompts that match different care preferences, different EMA feedback information, and different immediate contextual information for these 10 stroke patients. For example, the reference prompt set for each health group can include at least 100,000 different health prompts (e.g., options that match at least 8 types of EMA feedback information and 10 types of immediate contextual information). These reference prompt sets can be stored in a cloud server database for convenient subsequent personalized push notifications.

[0063] In this embodiment, the reference prompt information set includes early warning notification information indicating the onset of a primary symptom and various self-care suggestions for patients experiencing the primary symptom in different contexts. The early warning notification information can be used to indicate potential health problems (such as symptom names) and can also provide explanations such as the cause and context of the health problem. For example, "depression + loneliness." The self-care suggestions can be personalized self-care recommendations to alleviate a patient's symptoms in different contexts. For example, music therapy suggestions to alleviate depressive symptoms over a certain period, or aerobic exercise suggestions suitable for the patient.

[0064] In one embodiment, the individual characteristic information further includes the patient's nursing preference information, which includes at least one of the following: the patient's exercise preference information, diet preference information, and sleep preference information.

[0065] Step S1202, which involves selecting reference prompt information that matches the individual characteristic information and the current health information from the reference prompt information set corresponding to the health group to which the patient belongs, includes: selecting self-care suggestion information that matches the nursing preference information and the current health information from the reference prompt information set corresponding to the health group to which the patient belongs, as health prompt information that matches the current health information; the personalized rehabilitation nursing method based on ecological instantaneous assessment provided in this application embodiment further includes: determining the time for outputting the self-care suggestion information according to the nursing preference information.

[0066] The above solution fully considers patients' nursing preferences. On one hand, it ensures that the content of real-time self-care suggestions not only matches the patient's current health status and immediate situation but also accurately matches their nursing preferences, making the suggestions more personalized and increasing patient compliance and satisfaction. On the other hand, the solution ensures that the timing of self-care suggestions is more aligned with individual user needs, minimizing disruption to the patient's daily life and further enhancing compliance and satisfaction. Overall, the solution improves nursing outcomes and provides a better user experience.

[0067] In one implementation, step S120, which involves determining a health alert information matching the current health information based on the historical health information and the patient's current health information, includes steps S123 to S125.

[0068] Step S123: Based on the patient's historical health information, determine the pattern information of the patient's health problems.

[0069] For example, the real-time contextual information and ecological instantaneous assessment feedback information of patients collected over seven consecutive days can be analyzed to determine the patterns of symptom occurrence. For cases where historical health information from multiple patients has been collected, the group occurrence patterns of the same symptom across multiple patients can be analyzed, as well as the individual occurrence patterns of each symptom for each patient. For instance, the patient's EMA feedback information and real-time contextual information at the same time can be concatenated to form a health data set, thus obtaining a time series of the patient's health data over seven consecutive days. Machine learning models (such as ARIMA or exponential smoothing) can be used to analyze and infer the time series of this health data to obtain trends, fluctuations, or abnormalities in the patient's emotional and physical states throughout the day. Predictive models (such as random forests, support vector machines, and long short-term memory networks) can be used to predict the patterns of health problems in patients. For example, patterns in the time, place, social environment (e.g., who the patient is with), physical activity (e.g., what the patient is doing), and emotional state of a patient's depression can be obtained. For example, stroke patients have a higher risk of developing depressive symptoms at 2 PM every day. Similarly, anxiety patients have a higher risk of developing anxiety on Monday mornings. For example, stroke patients are at higher risk of developing depressive symptoms when they are alone. For example, stroke patients are at higher risk of developing depressive symptoms when they are at home.

[0070] Step S124: Based on the pattern information and the current health information, determine whether the current risk warning conditions are met.

[0071] In this embodiment, the pattern information of the patient's health problem and the patient's current health information are used as the basis for determining whether the current risk warning conditions are met. In some cases, the pattern information and the current health information can be combined to determine whether the current risk warning conditions are met. For example, if the real-time acquired contextual information of the patient meets the contextual conditions for the patient's health problem indicated in the pattern information, it is determined that the current risk warning conditions are met. In other cases, the current health information may be sufficient to determine whether the current risk warning conditions are met. For example, if the real-time acquired contextual information of the patient does not meet the contextual conditions for the patient's health problem indicated in the pattern information, but the currently received feedback information from the patient's instantaneous ecological assessment indicates that the patient is currently experiencing the health problem, it is determined that the current risk warning conditions are met.

[0072] Step S125: If the risk warning conditions are met, determine the health warning information that matches the current health information.

[0073] In the above-described solution, by analyzing the patient's historical health information, patterns of the health problems can be accurately predicted. Based on these predicted patterns and the patient's current health information, it is possible to quickly and accurately determine whether the risk of the patient's current symptoms meets the risk warning criteria. Furthermore, if the risk meets these criteria, timely and accurate health warnings matching the current health information can be identified. This allows for real-time prediction of the patient's symptom risk during recovery, automatic and timely risk warnings, and the timely delivery of personalized rehabilitation care suggestions tailored to the current situation. This can further improve patient satisfaction, promote recovery outcomes, and reduce medical costs.

[0074] In one implementation, the patient's ecological transient assessment feedback information at the current moment includes emotional feedback information and / or situational feedback information. Emotional feedback information may include mood feedback information (e.g., feeling depressed), and situational feedback information may include social environment feedback information (who the patient is with) and physical activity feedback information (what the patient is doing).

[0075] Step S124, which determines whether the risk warning conditions are met based on the pattern information and the current health information, includes: Step S1241, if the current health information includes emotional feedback information and the emotional feedback information indicates that the patient is currently experiencing the health problem, then it is determined that the risk warning conditions are met.

[0076] For example, a stroke patient can proactively report feeling depressed through the EMA module of the app. In this case, even if the current time and location do not meet the situational conditions for the patient's health problem as indicated in the regular information, it can still be determined that the current risk warning conditions are met, and appropriate early warning notifications and depression care suggestions can be pushed to the patient in a timely manner.

[0077] Step S125, which involves determining health alert information that matches the current health information when the risk alert conditions are met, includes steps S1251 and / or S1252.

[0078] Step S1251: If the current health information does not include the contextual feedback information, determine the cause information of the patient's health problem based on the pattern information, and determine that the notification information includes the health problem and the cause information.

[0079] For example, if a patient only reports their subjective feeling about a certain health problem but does not provide other background information, and if the current time and place do not meet the situational conditions for the patient to experience the health problem, but the information about the pattern of the patient's health problem indicates that the patient is prone to experiencing the health problem in a specific social environment or when doing a specific thing, then it can be determined that the specific social environment or the specific thing the patient does is the cause of the patient's health problem.

[0080] The following is an application scenario for the "non-contextualized prompts" type, as shown in Figure 3. For example, if a patient reports feeling depressed but doesn't specify the context (e.g., no feedback on who they were with or what they were doing), and if the patient's history of depression indicates they are prone to depression when alone, the app can pop up a warning notification such as "Depression + Loneliness," informing the patient that they may be feeling depressed due to loneliness. Alternatively, it can send nursing tips to alleviate depression that match the current immediate context. For example, if it's daytime and the patient is at home, it could suggest they talk to a friend.

[0081] Step S1252: If the current health information includes the contextual feedback information, determine the notification information and / or the autonomous intervention information based on the combination of the feedback information and the contextual feedback information.

[0082] The following is an application scenario for the "Contextualized Prompt" type, as shown in Figure 3. For example, if a patient reports feeling depressed at home and alone, the system can send a warning notification and / or self-care advice message tailored to the specific environment. For instance, the patient's mobile app could pop up a notification such as "Depression + Home Alone," and a self-care advice message such as "Go out and chat with a friend!"

[0083] Refer to the application scenarios of the "Proactively Triggered Prompt" type in Figure 3. For example, if a patient proactively reports feeling depressed at work through the EMA module, the system can send a work-related self-care prompt. For instance, the patient's mobile app can display a notification such as "Depression + Work," and also self-care suggestions such as "Stand up and move around!"

[0084] In the aforementioned approach, during rehabilitation nursing, health alerts can be promptly identified based on the patient's proactive emotional feedback indicating a health problem, allowing for timely responses to the patient's nursing needs. This immediate feedback mechanism helps reduce nursing delays, ensuring patients receive appropriate guidance and support when needed. Furthermore, different health alerts can be determined based on whether the patient proactively provides situational feedback. This ensures that the identified and outputted health alerts are more aligned with the actual nursing scenario. This personalized approach not only meets the patient's specific needs but also improves the relevance and satisfaction of nursing alerts, contributing to increased motivation and participation in the rehabilitation process. A comprehensive analysis combining emotional and situational feedback allows for a more accurate assessment of the patient's health status and needs, leading to more precise and effective nursing recommendations and further enhancing the effectiveness of rehabilitation nursing.

[0085] In one implementation, the current health information includes the real-time context information. Step S124, determining whether the risk warning condition is met based on the pattern information and the current health information, includes: Step S1242, determining that the risk warning condition is met if the real-time context information meets the contextual condition for the patient's health problem as indicated in the pattern information. For example, if the current time meets the time condition for the patient's health problem as indicated in the pattern information, the risk warning condition is met. Another example is if GPS detects that the patient's current location meets the location condition for the patient's health problem as indicated in the pattern information, the risk warning condition is met.

[0086] Step S125, which involves determining health alert information matching the current health information when the risk alert conditions are met, includes: Step S1253, determining the notification information and / or the self-intervention information based on the health problem and the immediate context information.

[0087] For example, consider the application scenario of the "passive trigger prompt" type in Figure 3. If analysis reveals that a stroke patient is prone to depression every afternoon, and the current time is afternoon or near afternoon, it can be directly determined that the current risk prompt condition is met. Alternatively, it can be determined that the current risk prompt condition is met if the patient's current EMA feedback information (which can be actively or passively provided by the patient) meets preset conditions. When the risk prompt condition is determined to be met, the notification information and / or the self-intervention information can be determined at least based on the patient's potential health problems (such as depression) and the immediate contextual information (such as the current time and the patient's location). In one example, if the patient's EMA feedback is not obtained, the notification information and / or the self-intervention information can be determined solely based on the patient's potential health problems and the immediate contextual information. In another example, if the patient's EMA feedback is obtained, the notification information and / or the self-intervention information can be determined by combining the patient's potential health problems, the immediate contextual information, and the patient's EMA feedback. For example, the patient's mobile app can display a warning prompt such as "Depression + Afternoon," and can also display corresponding self-care prompts for dealing with depression in the afternoon. Specifically, the app can display self-care suggestions based on the patient's current location. For example, if the app detects that the patient is at home, it can display a message such as "Listen to some light music to relax!"

[0088] In the aforementioned approach, even when patients do not actively report health problems, the system can predict the risk of health problems in real time and accurately based on immediate contextual information and analyzed patterns of such problems. When a high risk is predicted, it provides timely warnings and effective self-care suggestions tailored to the current situation. This proactive identification of potential health risks, even before the patient is aware of the problem, helps them take timely preventative measures and significantly reduces health risks. Furthermore, the immediate triggering of the warning mechanism upon high-risk prediction provides timely self-care advice relevant to the current context. This immediate response mechanism significantly improves the efficiency of rehabilitation care, ensuring patients receive timely and effective guidance at critical moments. Therefore, this approach provides patients with more comprehensive, timely, and effective rehabilitation care services. It not only reduces health risks and improves care efficiency but also enhances patient autonomy and the development of healthy behaviors, further improving the effectiveness of rehabilitation care.

[0089] In one embodiment, step S1242 may further include the following steps:

[0090] Step S1242.1: When the real-time contextual information satisfies the contextual conditions for the patient to have the health problem as indicated in the regularity information, a first window is displayed on the user interface operated by the patient, wherein the first window includes questionnaire information for ecological instantaneous assessment of the patient.

[0091] Step S1242.2: Based on the patient's operation information on the first window, determine the feedback information of the patient's instantaneous ecological assessment at the current moment;

[0092] Step S1242.3: If the feedback information meets the preset requirements, determine that the current risk warning conditions are met.

[0093] For example, referring again to the "passive trigger prompt" type application scenario in Figure 3, if analysis reveals that a stroke patient is prone to depression every afternoon, and the current time is afternoon or near afternoon, a prompt message such as "High risk of depression, please provide feedback for confirmation!" can be sent via the user's mobile phone. If the patient opens the mobile app, an EMA questionnaire assessment window can be displayed on the user interface. This assessment window can display simple questions (e.g., "Who are you with now?" and "What are you doing?"). After the patient answers the questions, it can be further determined whether the patient's feedback meets preset conditions. If so, it can be determined that the current risk prompt conditions are met. The preset conditions can be set according to actual needs. For example, if the patient reports being alone, it can be determined that the feedback meets the preset conditions.

[0094] In this way, by monitoring real-time contextual information and passive feedback from patients, the risk of a patient developing a particular health problem can be determined more accurately. This reduces the probability of misjudgment and incorrect prompts, thereby making rehabilitation care more effective and precise.

[0095] The following description, in conjunction with Figure 3, illustrates another embodiment of the personalized rehabilitation care method based on instantaneous ecological assessment of this application. As shown in Figure 3, the process of this method can be divided into two stages.

[0096] In the first phase, historical health information can be collected from multiple stroke patients receiving rehabilitation care (e.g., 50 stroke patients), and machine learning analysis can be performed on the collected historical health information to obtain meaningful analytical results.

[0097] For example, a patient's medical records, nursing preferences, feedback from their email assessment (EMA) over seven consecutive days, and their immediate situational information can be collected via a mobile app. Specifically, nursing preferences can include: the patient's exercise preferences, dietary preferences, sleep preferences, EMA prompting time and frequency, and health prompting time and method (e.g., voice prompts, text prompts, visual prompts). For instance, the EMA module in the patient's mobile app can collect feedback from their EMA over seven consecutive days, gathering real-time feedback on their depression levels, cognitive function, daily activities, and daily situational factors at different times of the day. For example, the EMA module might randomly issue six feedback prompts between 9:00 AM and 9:00 PM each day, prompting the patient to complete a brief assessment of their mood, cognitive tasks, physical activity, and situation. Specifically, Likert scales can be used to assess the patient's mood, such as feelings of sadness or despair. Cognitive task assessment involves simple memory recall exercises. Physical activity can be assessed by tracking the patient's activity level and sedentary behavior. Information about the patient's current social environment can also be obtained and assessed. Furthermore, the patient's location at different times can be obtained via mobile networks and the phone's GPS, especially the time and location when EMA feedback is conducted. For example, after collecting historical health information from 50 stroke patients, the collected information can be encrypted and stored on a cloud server for easy analysis and retrieval later.

[0098] For example, after collecting historical health information from 50 stroke patients, machine learning analysis can be used to analyze and process the patients' historical health information (mainly their EMA feedback information) and derive meaningful patterns or rules, laying the foundation for providing personalized care recommendations later. Specifically, firstly, the patients' historical health information can be preprocessed. The collected data usually contains missing or outlier values, which can be filled using the kNN algorithm, and outliers can be detected using the Z-score algorithm. Next, feature extraction can be performed on the preprocessed historical health information. For example, features can be extracted from the EMA feedback information of each patient at different times based on time, aggregated indicators (the overall measure obtained by summarizing and integrating multiple measures or indicators), and context. For example, extracting time-related features, such as the time of day and the day of the week, helps to detect patterns of emotional fluctuations. As another example, extracting context-related features, such as social environment features, can be used to analyze the impact of the social environment on the occurrence of health problems. Specifically, ARIMA or exponential smoothing algorithms can be used to analyze the time series of patient feedback information collected through the EMA module, thereby identifying trends, fluctuations, or abnormalities in the patient's emotional and physical state throughout the day. Subsequently, predictive models (such as random forests, support vector machines, and long short-term memory networks) can be used to predict the occurrence patterns of the patient's health problems and determine the optimal timing and personalized intervention measures. Furthermore, preference matching and behavior prediction can be achieved by analyzing the patient's historical health information, ensuring that the patient's feedback preferences align with their actual behavior, thus allowing for more personalized interventions. For example, if a patient expresses a desire to increase physical activity but consistently fails to achieve this goal, the system can flag this mismatch and suggest intervention. In addition, the system can learn the patient's behavioral patterns and determine the optimal time to prompt specific activities (such as physical exercise), subsequently pushing health tips to the patient at that optimal time. Furthermore, reinforcement learning algorithms can be integrated to dynamically optimize the timing and content of health tips. The data obtained through machine learning analysis in the first stage can also be applied to the training of the large language model in the second stage.

[0099] For example, in the second stage, a comprehensive information database can be developed by training a pre-set large language model to provide personalized care recommendations for post-stroke patients. This database can be powered by Baidu's large language model, which is trained using a large amount of patients' historical health information. This enhances the model's ability to accurately output push notifications that meet contextual and personalized needs, thus more accurately guiding the rehabilitation process of post-stroke patients. For example, additional training samples can be generated using Baidu's large language model to process and enhance the training samples for EMA feedback information, thereby enhancing the model's robustness and effectiveness. For example, Baidu's large language model can achieve diversity in model output, expanding to include more and richer health tips, helping to provide patients with health tips that better meet their unique needs, preferences, and contexts.

[0100] For example, Baidu's Big Language Model can be used to analyze the group characteristics (common symptom categories and health status) and individual characteristics (e.g., personalized data from EMA feedback) in each patient's historical health information to classify patients and build a comprehensive information database. For instance, group characteristics in historical health information include common symptoms faced by stroke patients, such as depression, cognitive impairment, and functional limitations. Based on the patient's symptom combinations, stroke patients can be divided into seven health groups: "Depression Group," "Cognitive Impairment Group," "Functional Impairment Group," "Depression + Cognitive Impairment Group," "Depression + Functional Impairment Group," "Cognitive Impairment + Functional Impairment Group," and "Depressive + Cognitive + Functional Impairment Group." Personalized characteristics can include each patient's care preferences and EMA feedback. Training the Baidu Big Language Model with the personalized characteristics of each patient in each health group enables it to generate diverse health alerts. Specifically, for each health group, the Baidu Big Language Model can be trained to generate 700,000 different push messages (7 groups × 8 categories of feedback × 10 options). These messages will be stored in the database and can be used for subsequent personalized notifications.

[0101] For example, the Baidu Big Language Model can also be trained to output personalized health tips that match the actual needs of patients in various application scenarios. Various suitable methods can be used to train the Baidu Big Language Model. Specifically, the historical health information of the aforementioned 50 stroke patients can be encoded and input into the Baidu Big Language Model for training. For example, the input data for training the model can include: data on depression, cognitive function, and daily activities reported by stroke patients through the EMA module, contextual data (e.g., environment, social interaction), and the time and location of the patients' EMA data feedback. This data can be encoded and concatenated according to certain encoding rules and then input into the Baidu Big Language Model for training. For example, contextual data can be processed into unordered categories, encoded using one-hot encoding, and then fed into the Baidu Big Language Model. For example, based on different users' nursing preference data and combined with the machine analysis results obtained in the first stage (such as the occurrence patterns of patient symptoms), labels can be set for each input data point to conform to a preset push logic (e.g., the logic of pushing different types of health tips in different application scenarios in Figure 3). For example, an appropriate loss function can be set based on the machine analysis results obtained in the first stage to train the model to output health tips that match the patient's actual personalized needs in various application scenarios.

[0102] For example, evaluation metrics for model training can include various metrics such as BLEU-4, ROUGE-1, ROUGE-2, and ROUGE-L. These metrics can be used to assess the accuracy and diversity of the model's output, enabling the trained large language model to output high-quality, context-sensitive health tips. For example, to reduce overfitting, additional training samples can be generated using the Baidu Large Language Model, thereby expanding the dataset and improving the model's generalization ability.

[0103] For example, after training the Baidu Big Language Model, in actual rehabilitation care, real-time collected patient contextual information and EMA feedback information can be input into the Baidu Big Language Model, and the health tips output by the model can be promptly pushed to the patient. See Figure 3. For instance, in a "non-contextual prompt" application scenario, if a patient reports feeling depressed but does not specify the context (e.g., no feedback on who they are with or what they are doing), but if the patient's pattern of depression indicates they are prone to depression when alone, the Baidu Big Language Model can output a warning notification such as "depression + loneliness," informing the patient that they may be feeling depressed due to loneliness. As another example, in a "contextual prompt" application scenario, if a patient reports feeling depressed at home and alone, the Baidu Big Language Model can output a warning notification and / or self-care suggestions tailored to the specific environment. As yet another example, in a "proactively triggered prompt" application scenario, if a patient proactively reports feeling depressed at work through the EMA module, the Baidu Big Language Model can output a work-related self-care tip. For example, in the application scenario of "passive trigger prompts", if analysis reveals that a stroke patient is prone to depression every afternoon, and the current time is detected to be afternoon or close to afternoon, Baidu's big language model can output warning prompts such as "depression + afternoon", and can also output corresponding self-care prompts for dealing with depression in the afternoon.

[0104] Furthermore, during the rehabilitation care of patients using the trained large language model, the model can be retrained using continuously collected health information from patients, thereby continuously enhancing its reasoning ability. Thus, Baidu's large language model can be used to achieve real-time, effective, and accurate rehabilitation care for patients, improving the rehabilitation outcomes for each individual.

[0105] The implementation of the personalized rehabilitation nursing method based on instantaneous ecological assessment according to the embodiments of this application has been described in detail above with reference to Figures 1 to 3. The system embodiments of this application will be described in detail below with reference to Figure 4. It should be understood that the personalized rehabilitation nursing system based on instantaneous ecological assessment in the embodiments of this application can execute the steps of the various personalized rehabilitation nursing methods based on instantaneous ecological assessment described in the foregoing embodiments of this application. That is, the specific working processes of the various products described below can be referred to the corresponding processes in the foregoing method embodiments.

[0106] Figure 4 shows a schematic diagram of the structure of a personalized rehabilitation care system based on instantaneous ecological assessment in some embodiments of this application. As shown in Figure 4, the personalized rehabilitation care system 400 based on instantaneous ecological assessment in this application includes:

[0107] The first acquisition module 410 is used to acquire the patient's historical health information, wherein the historical health information includes the patient's real-time contextual information and feedback information of ecological instantaneous assessment at multiple moments in the past period, and the real-time contextual information includes time information and / or the patient's location information.

[0108] The first determining module 420 is used to determine health prompt information matching the current health information based on the historical health information and the patient's current health information. The current health information includes: the patient's immediate context information and / or feedback information of the instantaneous ecological assessment at the current moment. The health prompt information includes notification information of possible health problems that the patient may have at the current moment and / or self-intervention information for the health problems.

[0109] Output module 430 is used to output the health alert information.

[0110] Each unit module of the above-mentioned personalized rehabilitation and nursing system 400 based on ecological instantaneous assessment can execute the corresponding steps in the above method embodiment. Therefore, each unit module will not be described in detail here. Please refer to the description of the corresponding steps above for details.

[0111] It should be noted that the aforementioned personalized rehabilitation and nursing system 400 based on instantaneous ecological assessment is embodied in the form of functional units. The term "unit" here can be implemented in software and / or hardware, without specific limitations.

[0112] Figure 4 is merely an example of a personalized rehabilitation care system based on instantaneous ecological assessment and does not constitute a limitation on such a system. A personalized rehabilitation care system based on instantaneous ecological assessment may include more or fewer components than shown in the figure, or may combine certain components or different components.

[0113] This application embodiment also provides a terminal device. As shown in FIG5, the terminal device 500 provided in this application embodiment includes: at least one processor 510 (only one processor is shown in FIG5), a memory 520, and a computer program 530 stored in the memory 520 and executable on at least one processor 510. When the processor 510 executes the computer program 530, it implements the steps of the above-mentioned personalized rehabilitation nursing method based on ecological instantaneous assessment.

[0114] The processor 510 may be a Central Processing Unit (CPU), or it may be 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. A general-purpose processor may be a microprocessor or any conventional processor.

[0115] In some embodiments, memory 520 may be an internal storage unit of terminal device 500, such as a hard disk or memory of terminal device 500. In other embodiments, memory 520 may be an external storage device of terminal device 500, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on terminal device 500. Furthermore, memory 520 may include both internal and external storage units of terminal device 500. Memory 520 can be used to store thermal performance parameters and loss data tables of power devices for processor access; memory 520 is also used to store operating systems, applications, boot loaders, data, and other programs, such as program code for computer programs. Memory 520 can also be used to temporarily store data that has been output or will be output.

[0116] This application also provides a computer program product that, when executed by a processor, implements the personalized rehabilitation care method based on ecological instantaneous assessment in any of the method embodiments of this application.

[0117] The computer program product can be stored in memory. For example, it is a program that undergoes preprocessing, compilation, assembly, and linking processes to eventually be converted into an executable object file that can be executed by a processor.

[0118] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the personalized rehabilitation care method based on instantaneous ecological assessment of any method embodiment of this application. The computer program may be a high-level language program or an executable object program.

[0119] The computer-readable storage medium is, for example, memory. Memory can be volatile or non-volatile, or it can include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0120] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0121] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes 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 this application.

[0122] 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 implementation should not be considered beyond the scope of this application.

[0123] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0124] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and there may be other division methods in actual implementation; 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, and the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0125] 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; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0126] In addition, the functional units in the various embodiments of this application 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.

[0127] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A personalized rehabilitation nursing method based on instantaneous ecological assessment, characterized in that, include: Obtain the patient's historical health information, wherein the historical health information includes real-time contextual information and feedback information of ecological instantaneous assessment at multiple moments in the past period, and the real-time contextual information includes time information and / or the patient's location information; Based on the historical health information and the patient's current health information, health prompt information matching the current health information is determined, wherein the current health information includes: the patient's immediate contextual information and / or feedback information of the instantaneous ecological assessment at the current moment, and the health prompt information includes notification information of possible health problems of the patient at the current moment and / or self-intervention information for the health problems; Output the health alert information.

2. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 1, characterized in that, The step of determining health alert information matching the current health information based on the historical health information and the patient's current health information includes: Based on the patient's historical health information, determine the pattern information of the occurrence of the patient's health problem; Based on the pattern information and the current health information, determine whether the current conditions for risk warning are met; If the risk warning conditions are met, determine the health warning information that matches the current health information.

3. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 2, characterized in that, The patient's current ecological instantaneous assessment feedback information includes emotional feedback information and / or situational feedback information. The step of determining whether the risk warning conditions are met based on the pattern information and the current health information includes: If the current health information includes the emotional feedback information, and the emotional feedback information indicates that the patient is currently experiencing the health problem, then it is determined that the current risk warning conditions are met. The step of determining health alert information matching the current health information when it is determined that the risk alert conditions are currently met includes: If the current health information does not include the contextual feedback information, based on the pattern information, determine the cause information of the patient's health problem, and determine that the notification information includes the health problem and the cause information; and / or If the current health information includes the contextual feedback information, the notification information and / or the autonomous intervention information are determined based on the combination of the emotional feedback information and the contextual feedback information.

4. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 2, characterized in that, The current health information includes the real-time context information, and determining whether the risk warning conditions are met based on the pattern information and the current health information includes: If the real-time contextual information satisfies the contextual conditions for the patient to experience the health problem as indicated in the regularity information, it is determined that the current risk warning conditions are met. The step of determining health alert information matching the current health information when it is determined that the risk alert conditions are currently met includes: The notification information and / or the autonomous intervention information are determined based on the health problem and the immediate contextual information.

5. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 4, characterized in that, When the immediate contextual information satisfies the contextual conditions for the patient's health problem as indicated in the pattern information, determining that the current risk warning conditions are met includes: When the real-time contextual information satisfies the contextual conditions for the patient to experience the health problem as indicated in the regularity information, a first window is displayed on the user interface operated by the patient, wherein the first window includes questionnaire information for an ecological instantaneous assessment of the patient. Based on the patient's operation information on the first window, determine the feedback information of the patient's instantaneous ecological assessment at the current moment; If the feedback information meets the preset requirements, it is determined that the current risk warning conditions are met.

6. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 1, characterized in that, The historical health information also includes the category information of the patient's condition. The step of determining health alert information matching the current health information based on the historical health information and the patient's current health information includes: Based on the group characteristic information in the historical health information, the health group to which the patient belongs is determined, wherein the group characteristic information includes the category information of the patient's disease; Based on the individual characteristic information in the historical health information and the current health information, reference prompts that match the individual characteristic information and the current health information are selected from the reference prompt information set corresponding to the health group to which the patient belongs. Based on the selected reference prompts, health prompts that match the current health information are determined. The individual characteristic information includes at least one of the patient's ecological instantaneous assessment of emotional feedback information, physical activity feedback information, cognitive feedback information, and social environment feedback information.

7. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 6, characterized in that, Before determining the health group to which the patient belongs based on the group characteristic information in the historical health information, the method further includes: Obtain historical health information from multiple patients; Based on the symptom category information of the multiple patients, the multiple patients are divided into multiple healthy groups, wherein the patients in each healthy group have the same symptom category; For each health group, a set of reference prompts is determined based on the individual characteristics of each patient in the historical health information of that health group. The set of reference prompts includes early warning notifications indicating the onset of the first symptom and various self-care suggestions for patients when the first symptom occurs in different situations.

8. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to claim 6, characterized in that, The individual characteristic information also includes the patient's nursing preference information, which includes at least one of the following: the patient's exercise preference information, dietary preference information, and sleep preference information. The step of selecting reference prompts that match the individual characteristic information and the current health information from the reference prompt information set corresponding to the patient's health group includes: Self-care suggestions that match the nursing preference information and the current health information are selected from the reference prompt information set corresponding to the health group to which the patient belongs, and used as health prompt information that matches the current health information; The method further includes: Based on the nursing preference information, determine the time to output the self-care suggestion information.

9. The personalized rehabilitation nursing method based on instantaneous ecological assessment according to any one of claims 1-8, characterized in that, Based on the historical health information and the patient's current health information, determine health alert information that matches the current health information, including: The patient's historical health information is used to train a pre-set large language model to obtain a trained large language model. The patient's current health information is input into the trained large language model to obtain health prompts that match the current health information.

10. A terminal device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the personalized rehabilitation care method based on ecological instantaneous assessment as described in any one of claims 1-9.

Images