Caregiver post-traumatic growth monitoring system and method based on multi-modal situational awareness

Abstract

This invention relates to a caregiver post-traumatic growth monitoring system and method based on multimodal situational awareness, comprising a sequentially functioning, closed-loop system: an independent sensing terminal, a role-situation recognition module, a situation-physiological stress decoupling module, a PTG semantic feature extraction module, an AIGC narrative reconstruction intervention module, and a closed-loop feedback unit. Through multimodal data acquisition and situational recognition, it achieves precise decoupling of physical fatigue and psychological stress; combining semantic analysis and generative AI technology, it generates personalized narrative reconstruction intervention content for different situations; and based on closed-loop feedback, it achieves dynamic strategy optimization. This invention, through the collaboration of an independent sensing terminal and multiple modules, accurately distinguishes between caregiver physical fatigue and psychological stress; combines semantic analysis and AIGC technology to generate personalized narrative interventions; and relies on closed-loop feedback and adaptive adjustment of the treatment cycle to achieve intelligent cognitive reassessment, comprehensively promoting post-traumatic growth.

Description

Technical Field

[0001] This invention relates to the field of digital medical processing technology, and in particular to a caregiver post-traumatic growth monitoring system and method based on multimodal context awareness. Background Technology

[0002] Cancer treatment is a long, complex, and uncertain process. Unlike sudden-onset disabling illnesses such as stroke, cancer treatment is characterized by significant periodicity, anticipated suffering, and anxiety about death. During this process, the patient's family caregivers are often referred to as "second victims."

[0003] Existing research indicates that cancer caregivers face a dual dilemma:

[0004] (a) High-intensity care burden: managing chemotherapy side effects, accompanying patients during radiotherapy, managing medications, etc.

[0005] (ii) Severe psychological trauma: vicarious trauma caused by witnessing the suffering of loved ones, fear of relapse and metastasis, and a sense of loss of self due to long-term sacrifice of personal life.

[0006] However, the psychological theory of "post-traumatic growth (PTG)" points out that if caregivers can undergo proper cognitive restructuring after experiencing trauma, they can gain enhanced personal strength, deepen interpersonal relationships, and re-examine the meaning of life.

[0007] In the prior art, patent CN113257382B discloses a "Cloud Computing-Based Post-Traumatic Growth Management System and Method for Stroke Patients." This technology primarily targets the functional recovery of the patient and relies on cloud computing data chains and smart contracts to match rehabilitation scenarios. However, for cancer caregivers, the existing technology has the following fundamental shortcomings:

[0008] (i) Reliance on patient data and neglect of caregiver independence: Most current systems treat caregivers as "accessories" of patients, alerting them by monitoring patients' vital signs. This design actually exacerbates caregivers' "hypervigilance" and deprives them of their right to be cared for as independent individuals.

[0009] (ii) Lack of context awareness: The system cannot distinguish whether the caregiver's heart rate is elevated because they are doing heavy housework or because they are anxiously waiting outside the chemotherapy room. Interventions that do not distinguish context are often ineffective or even disruptive.

[0010] (iii) The intervention methods lack depth: they mostly use fixed scales or general psychological platitudes, and cannot conduct in-depth narrative reconstruction based on the caregiver's specific inner monologue.

[0011] Therefore, there is a need for an intelligent system that is designed specifically for caregivers, can infer their stress situation by perceiving the environment, and can use generative AI for in-depth psychological narrative intervention. Summary of the Invention

[0012] In view of the shortcomings of the prior art, the purpose of this invention is to provide a caregiver post-traumatic growth monitoring system and method based on multimodal context awareness, so as to solve one or more problems in the prior art.

[0013] To achieve the above objectives, the technical solution of the present invention is as follows:

[0014] A caregiver post-traumatic growth monitoring system based on multimodal context awareness includes sequentially acting, closed-loop mechanisms:

[0015] The caregiver's independent sensing terminal is configured to collect real-time data on the caregiver's own physiological signs, body movement and behavior, as well as the physical environment characteristics of the environment in which the caregiver is located.

[0016] The role context recognition module, in conjunction with the caregiver's independent sensing terminal, is configured to identify the caregiver's current social role context based on physical environment feature data and body movement behavior data.

[0017] The context-physiological stress decoupling analysis module, in conjunction with the caregiver's independent perception terminal and the role context recognition module, is configured to align physiological sign data with the identified social role context in time sequence, construct a "context-stress decoupling model" to distinguish between physiological fatigue and psychological stress, and calculate the pure psychological stress index;

[0018] The PTG semantic feature extraction module is configured to acquire the caregiver's voice or text records through a human-computer interaction interface and extract semantic feature vectors that reflect the dimensions of post-traumatic growth.

[0019] The AIGC narrative reconstruction intervention module, in conjunction with the role situation recognition module, the situation-physiological stress decoupling analysis module, and the PTG semantic feature extraction module, is configured to generate targeted cognitive reconstruction narrative guidance scripts or psychological resilience training programs based on the pure psychological stress index, the current social role situation, and the PTG semantic feature vector.

[0020] The closed-loop feedback adjustment unit is configured to assess the effectiveness of the intervention based on the recovery of physiological signs data after the intervention, and feed back to the caregiver's independent sensing terminal to dynamically adjust subsequent monitoring and intervention strategies.

[0021] Furthermore, the caregiver-independent sensing terminal includes:

[0022] A physiological sensor array for quantifying the level of arousal in the autonomic nervous system includes a photoplethysmography (PPG) sensor and a skin conductance sensor; the PPG sensor is configured to collect heart rate variability indices of the caregiver, and the skin conductance sensor is configured to collect skin conductance response data of the caregiver.

[0023] An environmental sensing sensor array includes an environmental microphone, an ambient light sensor, an inertial measurement unit, and a positioning module. The environmental microphone is configured to collect the acoustic spectrum of ambient background noise. The ambient light sensor is used to monitor the caregiver's circadian rhythm exposure. The inertial measurement unit is used to monitor hand movement frequency and body posture. The positioning module is used to identify whether the caregiver is within the geofence of the cancer treatment center.

[0024] Furthermore, the social role situations include "high-intensity care situations", "medical care waiting situations", and "self-detachment and rest situations".

[0025] Furthermore, in the aforementioned medical care waiting scenario, high physiological arousal is labeled as purely psychological anxiety, and the medical care waiting scenario is defined as follows:

[0026] Location module data shows that the caregiver is within the hospital's geofence;

[0027] Physical behavior data showed that caregivers were in a state of prolonged stillness or minimal movement;

[0028] The ambient microphone detected a decibel level below a preset threshold and the presence of a specific frequency medical device alert tone.

[0029] Furthermore, the dimensional semantic feature vectors extracted by the PTG semantic feature extraction module include "interpersonal relationship depth", "discovery of new possibilities", "enhancement of personal power", "spiritual change" and "love of life".

[0030] Furthermore, the AIGC narrative reconstruction intervention module includes:

[0031] The negative schema capture submodule is configured to analyze caregiver input and identify negative cognitive schemas.

[0032] The contextualized reconstruction submodule is configured to combine the identified negative schemas with the social role context to generate narrative text through a large language model.

[0033] The multimodal output submodule is configured to transform the generated narrative text into narrative reconstruction intervention content and output it.

[0034] A monitoring method, applied to the aforementioned multimodal context-aware caregiver post-traumatic growth monitoring system, includes the following steps:

[0035] S1. During the cancer treatment cycle, data on caregivers are collected through their own independent sensing terminals to establish a baseline for their personal life and caregiving behavior.

[0036] S2. The social role situation of the caregiver is identified in real time through the role situation recognition module, and the situation-physiological stress decoupling analysis module is used in combination with the collected physiological sign data to determine whether there is psychological stress.

[0037] S3. When abnormal psychological stress is detected or the preset recording time is reached, guide the caregiver to make voice or text recordings through the PTG semantic feature extraction module.

[0038] S4 and AIGC narrative reconstruction intervention modules generate narrative reconstruction intervention content from integrated data and push it to the closed-loop feedback adjustment unit.

[0039] S5, the caregiver's independent sensing terminal continuously monitors the recovery status of indicators based on the intervention content, forming a closed loop.

[0040] Furthermore, the situation-physiological stress decoupling analysis module in step S2 determines the presence of psychological stress through the following steps:

[0041] Receive data and align it in time, including acquiring physiological data based on a physiological sensor array and acquiring environmental data based on an environmental perception sensor array;

[0042] Physiological fatigue is determined by whether the intensity of physical activity exceeds the exercise threshold. If it does, it is determined to be physiological fatigue and the system will not conduct psychological intervention. If it does not exceed the threshold, psychological stress is determined.

[0043] For the determination of psychological stress, if the heart rate is abnormally low and the skin conductance is high, it is determined to be psychological stress, and the system will trigger intervention; otherwise, it is determined to be normal rest, and the system will not intervene.

[0044] Furthermore, the narrative reconstruction intervention content generated in step S4 includes example questioning strategies, meaning construction strategies, and self-care strategies.

[0045] Furthermore, in step S1, the cancer treatment cycle employs an adaptive adjustment mechanism, which includes the following steps in sequence:

[0046] Increase the frequency of sleep monitoring during chemotherapy administration;

[0047] During periods of heightened side effects, improve the identification of waiting scenarios for medical care.

[0048] During the intermittent recovery period, the frequency of physiological data collection should be reduced.

[0049] Compared with the prior art, the beneficial technical effects of the present invention are as follows:

[0050] This invention's system collects caregiver's physiological, behavioral, and environmental data through an independent sensing terminal, constructing a dedicated monitoring baseline to fundamentally guarantee their independent caregiving status. By leveraging the synergistic effect of a role-context recognition module and a context-physiological stress decoupling analysis module, it aligns social role contexts with the temporal sequence of physiological data, effectively distinguishing between physiological fatigue and psychological stress, achieving precise stress context perception. Combined with a PTG semantic feature extraction module and an AIGC narrative reconstruction intervention module, it deeply analyzes caregiver semantic features and negative cognitive schemas, generating personalized narrative reconstruction content for different contexts, achieving intelligent intervention for cognitive reassessment. Relying on a closed-loop feedback adjustment unit, it dynamically evaluates and optimizes intervention strategies, while simultaneously matching the cancer treatment cycle to form an adaptive monitoring rhythm, comprehensively promoting the caregiver's post-traumatic growth process. Attached Figure Description

[0051] Figure 1 The diagram shows a system architecture schematic of a caregiver post-traumatic growth monitoring system and method based on multimodal context awareness, according to an embodiment of the present invention.

[0052] Figure 2 The diagram shows the logic flowchart of the context-physiological stress decoupling analysis module of the caregiver post-traumatic growth monitoring system and method based on multimodal context awareness, according to an embodiment of the present invention.

[0053] Figure 3 The flowchart illustrates the internal processing of the AIGC narrative reconstruction intervention module in the caregiver post-traumatic growth monitoring system and method based on multimodal context awareness, according to an embodiment of the present invention.

[0054] Figure 4 The diagram shows a timeline of an adaptive monitoring strategy for the cancer treatment cycle based on a multimodal context-aware caregiver post-traumatic growth monitoring system and method according to an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the following detailed description of the caregiver post-traumatic growth monitoring system and method based on multimodal situational awareness, in conjunction with the accompanying drawings and specific embodiments, will further illustrate these points. The advantages and features of this invention will become clearer from the following description. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, used only to facilitate and clearly illustrate the purpose of the embodiments of this invention. Please refer to the accompanying drawings for a clearer understanding of the objectives, features, and advantages of this invention. It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are only for illustrative purposes and to enable those skilled in the art to understand and read them. They are not intended to limit the implementation conditions of this invention and therefore have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to the size, without affecting the effects and objectives achieved by this invention, should still fall within the scope of the technical content disclosed in this invention.

[0056] Please refer to the following: Figure 1 A caregiver post-traumatic growth monitoring system based on multimodal context awareness, comprising sequentially acting components forming a closed loop:

[0057] a. Caregiver Independent Sensing Terminal: Worn by the caregiver in a cancer family, this terminal is configured to collect real-time data on the caregiver's physiological signs, body movement, and the physical environment of their surroundings. As the system's data entry point, this independent sensing terminal completely disconnects from the patient's device, protecting the caregiver's privacy. Specifically, it includes:

[0058] Physiological sensor suite: This suite quantifies the level of autonomic nervous system arousal, including a photoplethysmography (PPG) sensor and a skin conductance sensor. The PPG sensor is configured to collect heart rate variability indices from the caregiver. The skin conductance sensor is configured to collect skin conductance response data from the caregiver.

[0059] An environmental sensing sensor array includes an environmental microphone, an ambient light sensor, an inertial measurement unit (IMU), and a positioning module. The environmental microphone is configured to collect the acoustic spectrum of ambient background noise. Preferably, the environmental microphone does not have a recording function; it only extracts acoustic features such as decibels and frequency distribution, and the original audio is deleted immediately after extraction to protect privacy. The ambient light sensor monitors the caregiver's circadian rhythm exposure. The IMU monitors hand movement frequency and body posture. The positioning module identifies whether the caregiver is within the geofence of the cancer treatment center. The environmental sensing sensor array is designed to answer questions such as "Where is the caregiver now?", "Is the environment noisy?", and "Is the caregiver busy or sitting quietly?"

[0060] b. A role-context recognition module, which works in conjunction with the caregiver's independent sensing terminal and is communicatively connected to it, is configured to identify and define the caregiver's current social role context based on physical environment feature data and body movement behavior data, using a multimodal scene classification method. The social role context includes "high-intensity nursing context", "medical care waiting context", and "self-detachment and rest context".

[0061] In this embodiment, in the medical care waiting situation, high physiological arousal is labeled as pure psychological anxiety, and the medical care waiting situation is defined as follows:

[0062] Location module data shows that the caregiver is within the hospital's geofence;

[0063] Physical behavior data showed that caregivers were in a state of prolonged stillness or minimal movement;

[0064] The ambient microphone detected a decibel level below a preset threshold and the presence of a specific frequency medical device alert tone.

[0065] For example, in high-intensity care situations, such as when the positioning module is displayed at home or in a hospital ward, the inertial measurement unit captures high-frequency irregular movements, and the microphone detects background noise at home. In self-distraction and rest situations, such as when the positioning module is displayed in a park, shopping mall, or study, the inertial measurement unit captures smooth movements, and the microphone detects white noise or music.

[0066] c. The context-physiological stress decoupling analysis module, in conjunction with the caregiver's independent sensing terminal and the role-context recognition module, is configured to chronologically align physiological sign data with the identified social role context to construct a "context-stress decoupling model" to distinguish between physiological fatigue and psychological stress, and to calculate a purely psychological stress index. This module effectively solves the problem that psychological stress cannot be determined solely by the caregiver's heart rate.

[0067] The calculation of the pure psychological stress index relies on a situation-physiological stress decoupling analysis module, whose algorithm is based on well-known psychological stress quantification techniques. Heart rate variability and skin conductance data are acquired from a group of physiological sensors as raw features characterizing the state of the autonomic nervous system. Then, based on the output of the role-situation recognition module, only physiological data from low-movement situations (such as waiting for medical care, self-distraction and rest) are retained for psychological stress analysis, excluding interference from physical activity. Finally, the selected physiological characteristics are compared with a pre-established individual resting baseline. Using a linear or nonlinear mapping function commonly used in the field, the deviation of the physiological indicators is converted into an index value within a preset range; the higher the value, the more significant the psychological stress.

[0068] d. The PTG semantic feature extraction module is configured to obtain the caregiver's voice or text records through a human-computer interaction interface and use natural language processing technology to extract semantic feature vectors that reflect the post-traumatic growth dimension.

[0069] The PTG semantic feature extraction module extracts dimensional semantic feature vectors including "depth of interpersonal relationships," "discovery of new possibilities," "enhanced personal strength," "spiritual changes," and "love of life." By processing keywords reflecting different dimensional semantic feature vectors, caregivers are guided to record their feelings. Preferably, the PTG semantic feature extraction module can be configured to have voice tone analysis capabilities, enabling it to determine emotional valence and guide caregivers to record their feelings.

[0070] e. The AIGC narrative reconstruction intervention module, in conjunction with the role situation recognition module, the situation-physiological stress decoupling analysis module, and the PTG semantic feature extraction module, is configured to generate targeted cognitive reconstruction narrative guidance scripts or psychological resilience training programs based on the pure psychological stress index, the current social role situation, and the PTG semantic feature vector. The aim is to reconstruct the caregiver's negative expressions into positive narratives through a large language model.

[0071] Furthermore, the AIGC narrative reconstruction intervention module includes:

[0072] The Negative Schema Capture submodule is configured to analyze caregiver input and identify negative cognitive schemas such as “self-sacrifice,” “helplessness,” “guilt,” or “catastrophic visions of the future.”

[0073] The contextualized reconstruction submodule is configured to combine identified negative schemas with social role contexts to generate narrative text through a large language model. For example, if the context is "high-intensity care," the statement "I'm so tired, I can't hold on much longer" will be reconstructed into an achievement narrative of "I've made tremendous efforts to protect my family, and my exhaustion is a badge of love." If the context is "self-detachment and rest," then a permission suggestion of "rest is not betrayal, but rather a preparation for a better start" will be generated.

[0074] The multimodal output submodule is configured to convert the generated narrative text into narrative reconstruction intervention content and output it by transforming it into anthropomorphic voice or visual text and graphics, and finally push it to the caregiver through a smart terminal.

[0075] f. A closed-loop feedback adjustment unit is configured to assess the effectiveness of the intervention based on the recovery of physiological signs after the intervention and feed this assessment back to the caregiver's independent sensing terminal to dynamically adjust subsequent monitoring and intervention strategies. The AIGC narrative reconstruction intervention module generates targeted cognitive reconstruction narrative guidance scripts or psychological resilience training programs, monitors the caregiver's physiological indicators after the intervention (such as whether heart rate variability has increased) to determine the effectiveness of the intervention, and adjusts the generation strategy accordingly for the next intervention.

[0076] A monitoring method, applied to the aforementioned multimodal context-aware caregiver post-traumatic growth monitoring system, includes the following steps:

[0077] S1. During the cancer treatment cycle, data on caregivers are collected through independent sensing terminals to establish a baseline of their personal life and caregiving behavior, providing a benchmark reference value for subsequent assessment of psychological stress.

[0078] The cancer treatment cycle employs an adaptive adjustment mechanism; please refer to [the relevant documentation / reference]. Figure 4 The intervention is structured cyclically, including the chemotherapy administration period, the side effect outbreak period, and the intermittent recovery period. During the chemotherapy administration period, the system increases the frequency of sleep monitoring, continuously recording the dynamic changes in nocturnal heart rate variability using a photoplethysmography (PPG) sensor. Combined with sleep onset and wake-up times monitored by an ambient light sensor, the system analyzes the caregiver's sleep structure and autonomic nervous system regulation, employing physiological stress reduction interventions focused on "allowing oneself to rest" and "physiological stress reduction techniques" to prevent physiological breakdown. During the side effect outbreak period, the system focuses on improving the sensitivity to medical care waiting scenarios. This is indicated when the location module shows the patient is within the hospital's geofence, body movement data shows prolonged periods of stillness or slight movement, and the ambient microphone detects specific frequency medical device prompts. Interventions in this period emphasize "accepting uncertainty" and "emotional catharsis." During the intermittent recovery period, the frequency of physiological data collection is appropriately reduced to minimize interference with the caregiver's daily activities. High-frequency monitoring is only activated when the inertial measurement unit detects abnormal body movement patterns or the ambient light sensor shows circadian rhythm disruption. A "reflective writing" task is added to guide caregivers in reviewing the challenges of the previous cycle, extracting "what I accomplished," strengthening the PTG experience, and promoting the transformation from trauma to growth.

[0079] S2. The social role situation identification module identifies the caregiver's situation in real time, and the situation-physiological stress decoupling analysis module, combined with the collected physiological data, determines whether there is psychological stress.

[0080] For details, please refer to the following: Figure 2 The context-physiological stress decoupling analysis module determines the presence of psychological stress through the following steps:

[0081] The system receives and aligns data in time, including acquiring physiological data based on a physiological sensor array and environmental data based on an environmental sensor array. By aligning the environmental context label sequence with the physiological data sequence along the timeline, it achieves synchronous analysis of contextual features and physiological indicators. For example, it binds the "medical care waiting situation" label to heart rate variability and electrodermal data for the corresponding time period. The system invokes different physiological baselines based on the context label. In a "high-intensity care situation," the caregiver's baseline heart rate is set at a higher level because physical activity naturally leads to an increased heart rate. In this case, unless the heart rate is extremely abnormal, the system does not classify it as psychological stress. However, in a "medical care waiting situation," the baseline heart rate is set at the resting level. In this case, if a heart rate spike, a significant decrease in heart rate variability, and increased electrodermal activity are detected, the system classifies it as "resting psychological stress."

[0082] Physiological fatigue is determined by whether the intensity of physical activity exceeds the exercise threshold. If it does, it is determined to be physiological fatigue, and the system will not conduct psychological intervention. If it does not exceed the threshold, psychological stress is determined.

[0083] For the determination of psychological stress, if the heart rate is abnormally low and the skin conductance is high, it is determined to be psychological stress, and the system will trigger intervention; otherwise, it is determined to be normal rest, and the system will not intervene.

[0084] This system calculates an empathy fatigue index, which is derived by combining the aforementioned assessments of physiological fatigue and psychological stress with situational labels. This algorithm effectively avoids misinterpreting caregivers' fatigue from housework as a sign of psychological breakdown, thus allowing intervention only for genuine psychological trauma.

[0085] S3. When abnormal psychological stress is detected or the preset recording time is reached, the PTG semantic feature extraction module guides the caregiver to make voice or text records. The human-computer interaction interface pushes guiding questions related to the current social role situation to the caregiver, actively triggers the voice log recording request, guides the caregiver to express their current feelings, and provides an accurate psychological state profile for the generation of subsequent intervention content.

[0086] The S4 and AIGC narrative reconstruction intervention modules integrate data to generate narrative reconstruction intervention content and push it to the closed-loop feedback adjustment unit. Using generative artificial intelligence technology, personalized narrative reconstruction intervention content is generated based on the caregiver's current situation, stress level, and semantic characteristics. For details, please refer to [link / reference needed]. Figure 3The negative schema capture submodule first performs deep semantic analysis on the caregiver's voice or text recordings to accurately identify negative cognitive schemas such as "self-sacrifice," "helplessness," "guilt," or "catastrophic imagination of the future." For example, when a caregiver enters "I'm just waiting here, unable to do anything, and I feel useless" in a "medical care waiting situation," the system will identify negative schemas of "helplessness" and "self-denial." Subsequently, the contextual reconstruction submodule closely integrates these negative schemas with the current "medical care waiting situation," using a large language model to generate targeted positive narrative text. For example, in the above example, the system might generate a reconstructed narrative such as "Waiting is also a form of protection; your companionship itself provides the greatest comfort and strength to the patient; this persistence and patience are invaluable," helping the caregiver to understand the value of their actions from a new perspective. Finally, the multimodal output submodule transforms the generated narrative text into anthropomorphic voices preferred by caregivers, such as a gentle and soothing female voice or a steady and powerful male voice, or creates visual graphic cards containing warm colors and positive imagery. These are then pushed to caregivers through the screen of their independent perception terminal or the accompanying app, ensuring that the intervention content can be received clearly and comfortably.

[0087] Furthermore, the strategies for generating narrative reconstruction intervention content include the following:

[0088] Exceptional questioning strategy: When a caregiver is detected to be expressing overwhelming negativity, text is generated to guide them to recall past experiences of successfully coping with a treatment crisis.

[0089] Meaning construction strategy: When it is detected that the caregiver is in the interval between cancer treatment, text is generated to guide them to think about the positive significance of the care experience on personal character development.

[0090] Self-care strategy: When a “self-detachment and rest situation” is detected but physiological indicators are still not relaxed, generate mindfulness meditation guided audio and clear mental permission instructions.

[0091] Furthermore, and exemplary, examples of using AIGC technology to promote post-traumatic growth include the following scenarios:

[0092] Scenario description: The caregiver is on the 3rd day after chemotherapy. The system detects that the caregiver is in a "late-night sitting scenario" and the empathy fatigue index continues to rise.

[0093] System-initiated trigger:

[0094] A gentle notification pops up on the mobile app: "You've worked hard these past few days. Now it's your time. Want to share how you feel today?"

[0095] Voice input and analysis:

[0096] Caregiver's recording: "Seeing him keep vomiting and unable to eat, I feel like the food I cooked was useless. I didn't take good care of him, and he's suffering." Natural Language Analysis: Identified keywords "vomiting / unable to eat" (objective fact), "useless / ineffective care" (internal attribution / self-deprecation / negative schema), and "suffering" (vicarious trauma). Semantic characteristics: High negative emotional valence, low self-efficacy.

[0097] Project notification:

[0098] The system automatically generates prompts and sends them to a large cloud-based model. Role: A professional narrative therapist specializing in post-traumatic growth guidance. Task: To conduct a "cognitive reappraisal" of the user's self-blame. Techniques: Using "externalizing the problem" (side effects are a medication reaction, not the caregiver's fault) and "uncovering the positive aspects" (cooking itself is an expression of love). Input: User corpus.

[0099] Generate reconstructed text:

[0100] AIGC's comment: "Watching a loved one suffer is indeed very painful, which shows that you have a deep emotional connection with him. But please remember that vomiting is a violent reaction when chemotherapy drugs attack tumors, not because the food you cooked is bad. On the contrary, in such a difficult time, you still insist on preparing food for him. This persistence and companionship are the most powerful medicine in themselves. You are fighting alongside him in your own way."

[0101] Intervention push:

[0102] The above text is transformed into a warm female voice and played, and a visual card with the keywords "resilience" and "protection" is generated and stored in the corresponding "growth profile" in the APP to help caregivers accumulate positive psychological capital.

[0103] S5. The caregiver's independent perception terminal continuously monitors the recovery status of indicators based on the intervention content, forming a closed loop and dynamically optimizing the subsequent monitoring frequency and intervention plan to correct the caregiver's psychological resilience profile and enhance their confidence and ability to cope with the pressure of long-term care.

[0104] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0105] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A caregiver post-traumatic growth monitoring system based on multimodal context awareness, characterized in that, Including those that act sequentially and form a closed loop: The caregiver's independent sensing terminal is configured to collect real-time data on the caregiver's own physiological signs, body movement and behavior, as well as the physical environment characteristics of the environment in which the caregiver is located. The role context recognition module, in conjunction with the caregiver's independent sensing terminal, is configured to identify the caregiver's current social role context based on physical environment feature data and body movement behavior data. The context-physiological stress decoupling analysis module, in conjunction with the caregiver's independent perception terminal and the role context recognition module, is configured to align physiological sign data with the identified social role context in time sequence, construct a "context-stress decoupling model" to distinguish between physiological fatigue and psychological stress, and calculate the pure psychological stress index; The PTG semantic feature extraction module is configured to acquire the caregiver's voice or text records through a human-computer interaction interface and extract semantic feature vectors that reflect the dimensions of post-traumatic growth. The AIGC narrative reconstruction intervention module, in conjunction with the role situation recognition module, the situation-physiological stress decoupling analysis module, and the PTG semantic feature extraction module, is configured to generate targeted cognitive reconstruction narrative guidance scripts or psychological resilience training programs based on the pure psychological stress index, the current social role situation, and the PTG semantic feature vector. The closed-loop feedback adjustment unit is configured to assess the effectiveness of the intervention based on the recovery of physiological signs data after the intervention, and feed back to the caregiver's independent sensing terminal to dynamically adjust subsequent monitoring and intervention strategies.

2. The caregiver post-traumatic growth monitoring system based on multimodal context awareness as described in claim 1, characterized in that, The caregiver-independent sensing terminal includes: A physiological sensor array for quantifying the level of arousal in the autonomic nervous system includes a photoplethysmography (PPG) sensor and a skin conductance sensor; the PPG sensor is configured to collect heart rate variability indices of the caregiver, and the skin conductance sensor is configured to collect skin conductance response data of the caregiver. An environmental sensing sensor array includes an environmental microphone, an ambient light sensor, an inertial measurement unit, and a positioning module. The environmental microphone is configured to collect the acoustic spectrum of ambient background noise. The ambient light sensor is used to monitor the caregiver's circadian rhythm exposure. The inertial measurement unit is used to monitor hand movement frequency and body posture. The positioning module is used to identify whether the caregiver is within the geofence of the cancer treatment center.

3. The caregiver post-traumatic growth monitoring system based on multimodal context awareness as described in claim 1, characterized in that: The social role scenarios include "high-intensity nursing scenarios", "medical care waiting scenarios", and "self-detachment and rest scenarios".

4. The caregiver post-traumatic growth monitoring system based on multimodal context awareness as described in claim 3, characterized in that: In the aforementioned medical companionship waiting scenario, high physiological arousal is labeled as purely psychological anxiety. The medical companionship waiting scenario is defined as follows: Location module data shows that the caregiver is within the hospital's geofence; Physical behavior data showed that caregivers were in a state of prolonged stillness or minimal movement; The ambient microphone detected a decibel level below a preset threshold and the presence of a specific frequency medical device alert tone.

5. The caregiver post-traumatic growth monitoring system based on multimodal context awareness as described in claim 1, characterized in that: The dimensional semantic feature vectors extracted by the PTG semantic feature extraction module include "interpersonal relationship depth", "discovery of new possibilities", "enhancement of personal power", "spiritual change" and "love of life".

6. The caregiver post-traumatic growth monitoring system based on multimodal context awareness as described in claim 1, characterized in that, The AIGC narrative reconstruction intervention module includes: The negative schema capture submodule is configured to analyze caregiver input and identify negative cognitive schemas. The contextualized reconstruction submodule is configured to combine the identified negative schemas with the social role context to generate narrative text through a large language model. The multimodal output submodule is configured to transform the generated narrative text into narrative reconstruction intervention content and output it.

7. A monitoring method applied to the caregiver posttraumatic growth monitoring system based on multi-modal context awareness according to any one of claims 1 to 6, characterized in that, Includes the following steps: S1. During the cancer treatment cycle, data on caregivers are collected through their own independent sensing terminals to establish a baseline for their personal life and caregiving behavior. S2. The social role situation of the caregiver is identified in real time through the role situation recognition module, and the situation-physiological stress decoupling analysis module is used in combination with the collected physiological sign data to determine whether there is psychological stress. S3. When abnormal psychological stress is detected or the preset recording time is reached, guide the caregiver to make voice or text recordings through the PTG semantic feature extraction module. S4 and AIGC narrative reconstruction intervention modules generate narrative reconstruction intervention content from integrated data and push it to the closed-loop feedback adjustment unit. S5, the caregiver's independent sensing terminal continuously monitors the recovery status of indicators based on the intervention content, forming a closed loop.

8. The monitoring method as described in claim 7, characterized in that: Step S2, the situation-physiological stress decoupling analysis module, determines the presence of psychological stress through the following steps: Receive data and align it in time, including acquiring physiological data based on a physiological sensor array and acquiring environmental data based on an environmental perception sensor array; Physiological fatigue is determined by whether the intensity of physical activity exceeds the exercise threshold. If it does, it is determined to be physiological fatigue and the system will not conduct psychological intervention. If it does not exceed the threshold, psychological stress is determined. For the determination of psychological stress, if the heart rate is abnormally low and the skin conductance is high, it is determined to be psychological stress, and the system will trigger intervention; otherwise, it is determined to be normal rest, and the system will not intervene.

9. A monitoring method as described in claim 8, characterized in that: Step S4 generates narrative reconstruction intervention content including example questioning strategies, meaning construction strategies, and self-care strategies.

10. The monitoring method of claim 7, wherein: Step S1, which involves an adaptive adjustment mechanism for cancer treatment cycles, includes the following in sequence: Increase the frequency of sleep monitoring during chemotherapy administration; During periods of heightened side effects, improve the identification of waiting scenarios for medical care. During the intermittent recovery period, the frequency of physiological data collection should be reduced.

Images