Healthcare landscape intervention method and device based on user feedback

Abstract

The invention discloses a healing landscape intervention method and device based on user feedback, and the method comprises the steps: carrying out the intervention of a to-be-intervened user according to a preset intervention scheme, and obtaining the feedback data of the to-be-intervened user after the intervention; wherein the preset intervention scheme comprises a preset healing landscape video; collecting user multi-dimensional data of the to-be-intervened user, and preprocessing the feedback data and the user multi-dimensional data to obtain a plurality of personal emotion indexes corresponding to the to-be-intervened user; and according to the plurality of personal emotion indexes, constructing a healing isomerism map, and according to the healing isomerism map, adjusting the preset intervention scheme. According to the invention, the intervention effect of psychological intervention based on the healing landscape can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of mental health management, in particular to a healing landscape intervention method and device based on user feedback. BACKGROUND

[0002] Focusing on the therapeutic effect of non-drug health intervention on emotional problems is conducive to reducing the medical burden of the mental health service system.

[0003] Landscape environment has healing and recovery properties and has a positive effect on people's mental health, such as relieving stress, improving attention, reducing feelings of loneliness, reducing mental fatigue, delaying cognitive decline, increasing positive emotions, reducing negative emotions, and improving well-being. Healing landscapes have shown positive effects in assisting mental health interventions, and exposure to healing landscapes can reduce depression and anxiety levels in patients with depression or anxiety.

[0004] In daily life, it takes time to travel to access real healing landscapes or is limited by factors such as weather, and people are increasingly using electronic products and screens, reducing the opportunity to directly access healing landscapes. With the development of digital economy and Internet of Things technology, digital interventions have become a form of alternative medicine with advantages such as low cost, flexibility, anonymity, and scalability. Smart devices, applications, and websites can be used as a medium for psychological treatment interventions, providing self-intervention and daily health data collection services.

[0005] Existing digital psychological intervention research is mainly based on cognitive therapy and behavioral therapy interventions, with only a few studies considering increasing user exposure to healing landscapes such as natural environments through digital means such as applications or implementing related interventions to reduce mental health problems and medical service burdens. However, these studies only focus on the experience of exposure to real landscape environments. Virtual healing landscapes such as pictures and videos can provide a more convenient way for busy office workers, the elderly with mobility problems, and hospitalized patients to access nature and relax. However, most studies on virtual landscape healing are limited to controlled environments such as laboratories and fail to combine them with digital platforms to implement emotional healing interventions in daily situations. Existing psychological healing systems mainly use music, text, static images, or human-computer dialogue for healing content, and have not yet built a digital intervention medium based on healing landscape materials and an emotional tracking system to verify the emotional health intervention effects of virtual healing landscapes. SUMMARY

[0006] The present application overcomes the shortcomings of the prior art and provides a healing landscape intervention method and device based on user feedback, which can improve the intervention effect of psychological intervention based on healing landscapes.

[0007] An embodiment of the present application provides a healing landscape intervention method based on user feedback, comprising the following steps: intervene in the user to be intervened according to a preset intervention scheme, and obtain feedback data of the user to be intervened after the intervention; wherein the preset intervention scheme comprises a preset healing landscape video; collect user multi-dimensional data of the user to be intervened, and pre-process the feedback data and the user multi-dimensional data to obtain a plurality of personal emotional indicators corresponding to the user to be intervened; construct a healing hetero-graph based on the plurality of personal emotional indicators, and adjust the preset intervention scheme according to the healing hetero-graph.

[0008] Further, the intervention in the user to be intervened according to the preset intervention scheme specifically comprises: preparing the preset intervention scheme for the user to be intervened, and intervening in the user to be intervened through the preset healing landscape video according to the preset intervention scheme; wherein the preset intervention scheme comprises intervention weeks, intervention frequency per week, intervention time length per day, intervention times per day, and a plurality of different types of preset healing landscape videos.

[0009] Further, the obtaining of the feedback data of the user to be intervened specifically comprises: after the user to be intervened watches the preset healing landscape video, making the user to be intervened score the instant emotion and the preset healing landscape video respectively, and obtaining instant emotion feedback data and healing landscape video feedback data correspondingly.

[0010] Further, the collection of the user multi-dimensional data of the user to be intervened specifically comprises: collecting personal basic information data of the user to be intervened; wherein the personal basic information data comprises gender, age, education, electronic screen use time length, and self-evaluation of economic and social status; collecting clinical evaluation data of the user to be intervened; wherein the clinical evaluation data comprises depression score and anxiety score, and the clinical evaluation data is collected at a plurality of preset time points respectively; collecting use behavior data of the user to be intervened; wherein the use behavior data comprises the number of clicks, the watching time length and the watching time of each preset healing landscape video by the user to be intervened, and the time and the latitude and longitude coordinates of the preset healing landscape video watched by the user to be intervened per day.

[0011] Further, the pre-processing of the feedback data and the user multi-dimensional data to obtain a plurality of personal emotional indicators corresponding to the user to be intervened specifically comprises: According to the depression score and the anxiety score, quantitative evaluation is performed by Hamilton Depression Scale and Hamilton Anxiety Scale respectively, and depression indicators and anxiety indicators are obtained correspondingly; According to the instant emotional feedback data, positive emotional indicators and negative emotional indicators of the user to be intervened are calculated; According to the healing landscape video feedback data and the use behavior data, a single video preference indicator and a healing landscape type preference indicator of the user to be intervened corresponding to each preset healing landscape video are calculated. According to the depression indicators, the anxiety indicators, and the positive emotional indicators and the negative emotional indicators of the user to be intervened, an intervention effect indicator is calculated.

[0012] Further, the construction of the healing heterogeneous graph according to the plurality of personal emotional indicators specifically includes: The user portrait corresponding to the user to be intervened is established according to the personal basic information data; A personal therapeutic effect evaluation report corresponding to the user to be intervened is prepared according to the user portrait, the intervention effect indicator, and the single video preference indicator and the healing landscape type preference indicator of the user to be intervened corresponding to each preset healing landscape video, wherein the personal therapeutic effect evaluation report includes user portrait analysis, daily emotional indicator dynamic change analysis, healing landscape type preference and video preference analysis, and intervention effect analysis. According to the personal therapeutic effect evaluation report, a healing heterogeneous graph is constructed through multi-modal data analysis.

[0013] Further, the adjustment of the preset intervention scheme according to the healing heterogeneous graph specifically includes: According to the user portrait analysis and the healing landscape type preference and video preference analysis, the preset healing landscape video type and the number of daily interventions in the preset intervention scheme are adjusted. According to the daily emotional indicator dynamic change analysis, the emotional trend of the user to be intervened is predicted through a preset emotional trend prediction model, and the number of intervention weeks, the weekly intervention frequency, and the daily intervention duration of the preset intervention scheme are adjusted according to the prediction result.

[0014] Preferably, the personal therapeutic effect evaluation report further includes an intervention course scheme completion rate, and a specific calculation formula of the intervention course scheme completion rate is:

[0015] wherein, and are the total duration of the video actually watched by the user and the preset intervention target duration, respectively.

[0016] Another embodiment of the present invention provides a therapeutic landscape intervention device based on user feedback, comprising: an intervention module, an analysis module, and an adjustment module; The intervention module is used to intervene in the user to be intervened according to a preset intervention plan, and to obtain feedback data from the user to be intervened after the intervention; wherein, the preset intervention plan includes a preset therapeutic landscape video; The analysis module is used to collect multidimensional user data of the user to be intervened, and to preprocess the feedback data and multidimensional user data to obtain several personal emotional indicators corresponding to the user to be intervened. The adjustment module is used to construct a healing heterogeneous map based on the several personal emotional indicators, and to adjust the preset intervention plan based on the healing heterogeneous map.

[0017] Furthermore, the intervention module is used to intervene in the user to be intervened according to a preset intervention plan, specifically including: A preset intervention plan is pre-formulated for the user to be intervened, and the user is intervened on using the preset healing landscape video according to the preset intervention plan; wherein, the preset intervention plan includes the number of intervention weeks, the weekly intervention frequency, the daily intervention duration, the daily intervention frequency, and the preset healing landscape video.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention, through emotion data analysis and an automatic adjustment mechanism for landscape intervention, can respond in real time to changes in users' emotional states, ensuring a high degree of alignment between the landscape intervention plan and the user's emotions. This dynamic intervention model significantly improves the accuracy of interventions, thereby enhancing their effectiveness. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a therapeutic landscape intervention method based on user feedback, provided as an embodiment of the present invention.

[0020] Figure 2 This is a schematic diagram of a therapeutic landscape intervention device based on user feedback, provided as another embodiment of the present invention. Detailed Implementation

[0021] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent. It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] Reference Figure 1The following is a flowchart illustrating a user-feedback-based therapeutic landscape intervention method according to an embodiment of the present invention, comprising the following steps: S1: Perform intervention on the user to be intervened according to the preset intervention plan, and obtain feedback data from the user to be intervened after the intervention; wherein, the preset intervention plan includes a preset therapeutic landscape video; S2: Collect the user multidimensional data of the user to be intervened, and preprocess the feedback data and user multidimensional data to obtain several personal emotional indicators corresponding to the user to be intervened; S3: Construct a healing heterogeneous map based on the aforementioned individual emotional indicators, and adjust the preset intervention plan based on the healing heterogeneous map.

[0024] For step S1, specifically, the step of intervening in the user to be intervened according to the preset intervention plan includes: A preset intervention plan is pre-formulated for the user to be intervened, and the user is intervened on using the preset healing landscape video according to the preset intervention plan; wherein, the preset intervention plan includes the number of intervention weeks, the weekly intervention frequency, the daily intervention duration, the daily intervention frequency, and several different types of the preset healing landscape video.

[0025] In a preferred embodiment, a healing landscape intervention program is implemented by providing users with selectable video materials from a healing landscape database in the cloud. Different types of healing landscape videos are provided as intervention content to ensure that users watch videos that suit their emotional needs within a diverse range of landscape interventions. Simultaneously, users can also independently select healing landscape videos to watch from the prescribed program within the application.

[0026] The pre-set intervention plan can be set by researchers in the backend, including the number of intervention weeks, weekly intervention frequency, daily intervention duration, daily intervention frequency, intervention content, and other intervention guidelines. For example, the pre-set intervention plan may include: 4 weeks of intervention, 5 days of intervention per week, 15 minutes of intervention per day, 1 intervention per day, intervention content of a pre-set healing landscape video, and other intervention guidelines such as "It is recommended to record the experiences, locations, and feelings of real contact with healing landscapes in daily life, and take and upload photos of the healing landscapes weekly." Daily contact experiences with healing landscapes can reflect users' landscape preferences and can be used to analyze the landscape elements that users pay attention to daily (trees, shrubs, lawns, flowers, water bodies, mountains, animals, sky, clouds, sounds, smells, climate, etc.) and their associated specific emotions (pleasure, relaxation, excitement, etc.). Based on this, the potential emotional healing effects of different categories of virtual landscapes can be further evaluated. In addition, it can also be used to analyze whether virtual healing landscapes will promote more outdoor activities, thereby improving users' emotional health.

[0027] Specifically, the landscape types and materials in the preset healing landscape videos are pre-evaluated by attending psychiatrists and human geography researchers, and then collected, screened, edited, and uploaded to the cloud by researchers. The healing landscape videos include five categories of landscapes: green landscapes, blue landscapes, geomorphic landscapes, temporal landscapes, and animal landscapes. Green landscapes include forest landscapes, grassland landscapes, park landscapes, and garden / courtyard landscapes; blue landscapes include river / stream landscapes, waterfall landscapes, lake landscapes, and ocean / beach landscapes; geomorphic landscapes include mountain / hill landscapes, glacier / snow mountain landscapes, cave / desert landscapes, and volcanic landscapes; temporal landscapes include sunrise / sunset landscapes, autumn landscapes, snow and ice landscapes, and aurora / midnight landscapes; and animal landscapes include bird landscapes, panda landscapes, forest animal landscapes, and marine animal landscapes.

[0028] Because each type of landscape is intertwined (e.g., a forest landscape in a green landscape video may include a stream landscape), and to avoid making the video landscapes too monotonous and reducing user viewing interest, each type of landscape focuses on videos that primarily showcase its main features, without removing other types of landscapes that may appear in the videos. Each video is approximately 4-10 minutes long. The video audio sound types cover ambient sounds, flowing water sounds, insect chirps, birdsong, and instrumental music, etc. Audio sound combinations include ambient sounds only, flowing water sounds only, insect chirps only, birdsong only, instrumental music only, ambient sounds + instrumental music, flowing water sounds + instrumental music, flowing water sounds + birdsong, flowing water sounds + insect chirps + birdsong + instrumental music, etc.

[0029] Users can execute intervention plans through the healing landscape video intervention module and daily recording module in the user interface. The type of healing landscape and the audio sound combination method serve as tags for the videos. Keyframes, scene transitions, and visual style of the healing landscape videos are extracted using a convolutional neural network to analyze their visual characteristics, including the proportion of landscape elements, the number of key landscape elements, and color features. Audio features such as sound type, rhythm, and melody in the videos are extracted and analyzed using the VGGish deep learning model.

[0030] In this preferred embodiment, the standardized framework and specific content of the healing landscape intervention program are shown in Table 1: Table 1. Default settings for the standardized framework and specific content of the healing landscape intervention program.

[0031] For step S2, specifically, obtaining the feedback data of the user to be intervened in includes: After the user to be intervened watches the preset healing landscape video, the user is asked to rate their own immediate emotions and the preset healing landscape video, and corresponding immediate emotion feedback data and healing landscape video feedback data are obtained.

[0032] In a preferred embodiment, the real-time emotional data, including calm, pleasure, happiness, tension, fatigue, and sadness, is filled out by the user after watching a video during the intervention. The user is asked to answer based on their state of mind that day: "Overall, do you feel calm / pleasant / happy / tension / fatigue / sad?", with a rating in the range of [0,10], where 0 = "not at all" and 10 = "very strong".

[0033] The feedback data from the healing landscape videos includes user ratings and text comments, which are filled out by users after watching the videos during the intervention. Users can rate the videos they watched that day (of the same type of healing landscape) based on their feelings, including three indicators: amazement, restorative potential, and willingness to visit: "The scenes in these videos are amazing and amazed me," "Watching these videos made me feel relaxed, gave me spiritual strength and energy," and "If I had the opportunity, I would like to actually go to the environment in these videos." The rating is in the range of [0,10], where 0 = "strongly disagree" and 10 = "strongly agree." Users can also rate the preference and healing effect of individual videos: "Do you like the environment in this video?" The rating is in the range of [1,5], where 1 = "very dislike," 2 = "dislike," 3 = "neutral," 4 = "like," and 5 = "very much like." "How do you think the environment in this video has a healing effect?" The rating is in the range of [1,4], where 1 = "no effect," 2 = "uncertain," 3 = "neutral effect," and 4 = "can improve anxiety and depression."

[0034] Furthermore, the collection of multidimensional user data of the user to be intervened specifically includes: Collect the basic personal information data of the user to be intervened; wherein, the basic personal information data includes gender, age, education level, screen time, and self-assessed economic and social status; Clinical assessment data of the users to be intervened are collected; wherein, the clinical assessment data includes depression scores and anxiety scores, and the clinical assessment data are collected at several preset time points; Collect the usage behavior data of the user to be intervened; wherein, the usage behavior data includes the number of clicks, viewing duration and viewing time of each preset healing landscape video by the user to be intervened, as well as the time and latitude and longitude coordinates of the user to be intervened watching the preset healing landscape video every day.

[0035] In a preferred embodiment, the basic personal information data is collected before intervention and used for subsequent user profiling.

[0036] The clinical assessment data included depression and anxiety scores, disease stage, and patient disease type. Depression and anxiety scores were assessed by a psychiatrist using the Hamilton Depression Rating Scale (HAM-D17) and the Hamilton Anxiety Rating Scale (HAM-A) to evaluate the user's feelings over the previous week. The scoring criteria were: 0 = "Asymptomatic"; 1 = "Mild"; 2 = "Moderate"; 3 = "Severe"; 4 = "Very Severe". Disease stage was divided into non-remission and remission phases. Patient disease type was determined based on diagnostic data and categorized as anxiety patients, depression patients, and patients with comorbid anxiety and depression. This data was collected before intervention (T0), during the second week of intervention (T1), at the end of intervention (T2), one month after intervention (T3), and three months after intervention (T4).

[0037] In addition, the user's multidimensional data may also include daily healing landscape contact log data. This daily healing landscape contact log data includes photos and text diaries uploaded by the user. The application guides users to observe and engage with healing landscapes in outdoor locations such as outside windows, on roads, or in parks, encouraging them to pay attention to and experience landscape elements including trees, shrubs, lawns, flowers, water, mountains, animals, sky, clouds, sounds, smells, and climate, as well as sensory experiences such as sight, hearing, and touch. The user interface provides a sample record for user reference: "Today I took a 30-minute walk in the park. The weather was sunny and I felt very relaxed. Passing a lake, the sunlight scattered on its surface, and time seemed to stand still. In this environment, I didn't have to think about anything; I just quietly felt it."

[0038] Furthermore, the preprocessing of the feedback data and user multidimensional data yields several personal emotional indicators corresponding to the user to be intervened, specifically including: Based on the depression and anxiety scores, quantitative assessments were performed using the Hamilton Depression Rating Scale and the Hamilton Anxiety Rating Scale, respectively, to obtain corresponding depression and anxiety indicators. Based on the real-time emotional feedback data, the positive and negative emotional indices of the user to be intervened are calculated. Based on the feedback data and usage behavior data of the therapeutic landscape videos, the individual video preference index and therapeutic landscape type preference index of the user to be intervened for each of the preset therapeutic landscape videos are calculated. The intervention effect index is calculated based on the depression index, anxiety index, and the positive and negative emotion indices of the users to be intervened.

[0039] In a preferred embodiment, the user multidimensional data mainly includes quantitative data, which, after preprocessing, can yield depression indicators, anxiety indicators, positive / negative emotion indicators, individual video preference indicators, healing landscape type preference indicators, and intervention effect indicators.

[0040] The depression index is assessed using 17 items of the HAM-D17 scale, with higher total scores indicating more severe depression. The anxiety index is assessed using 14 items of the HAM-A scale, with higher total scores indicating more severe anxiety. The depression and anxiety scores are expressed as follows: =(HAM-D17T0, HAM-D17T1, HAM-D17T2, HAM-D17T3, HAM-D17T4) and =(HAM-AT0, HAM-AT1, HAM-AT2, HAM-AT3, HAM-AT4), where i represents the user being evaluated, and T0, T1, T2, T3 and T4 represent the data collection time points, namely before intervention, the second week during intervention, the end of intervention, 1 month after intervention and 3 months after intervention, respectively. ≥24 is marked as "severe depression" A score of ≥17 indicates "may indicate mild or moderate depression". <7 is marked as "no depressive symptoms". ≥29 is marked as "severe anxiety". A score of ≥21 is marked as "significant anxiety". ≥14 is marked as "definitely has anxiety". ≥7 is marked as "may indicate anxiety". <6 is marked as "no anxiety".

[0041] Daily mood rating indicators can be divided into positive and negative mood indicators. User i's positive mood rating at data collection time k is the average of the calm, pleasant, and happy ratings collected at that time, denoted as: The negative score is the average of the stress, fatigue, and sadness scores collected at that time, expressed as... .

[0042] The individual video preference metric consists of the video's standardized click count, standardized viewing duration, user preference for the video, and therapeutic effect rating, and is expressed as follows: , , , In the formula, For video v, a preference indicator; and These represent the standardized values ​​of the number of clicks and the viewing duration of video v, respectively, ranging from [0,1]. and These are user ratings for their preference for video v and its therapeutic effect; The number of clicks for video v. and These represent the maximum and minimum number of video clicks, respectively. The viewing duration of video v and These represent the maximum and minimum video viewing durations, respectively.

[0043] The healing landscape type preference index refers to user i's preference for each type of healing landscape, expressed as follows: ,in, , , These represent user i's amazement, restorative potential, and willingness to visit each type of healing landscape, with higher scores indicating a stronger preference for that type.

[0044] The intervention effectiveness indicators are calculated from multiple indicators, including those related to depression, anxiety, positive emotions, and negative emotions. Specifically, the intervention effectiveness indicator at data collection time point k can be expressed as: ,in It can refer to the baseline values ​​of depression, anxiety, positive mood, and negative mood at time T0. This can refer to the indicators of depression, anxiety, positive emotions, and negative emotions at the data collection time point k. A decrease in depression, anxiety, and negative emotions compared to the baseline value, and an increase in the negative emotion indicator compared to the baseline value, are marked as "effective intervention," and vice versa, "ineffective intervention."

[0045] Meanwhile, the user's multidimensional data may also include unstructured text data and image data. The unstructured text data includes text comments on videos and healing landscape text diaries. This type of data is analyzed in depth using natural language processing and machine learning algorithms to automatically identify the user's emotional type (such as anxiety, depression, pleasure, etc.) and its intensity, and to extract the healing landscape elements that the user focuses on and their corresponding emotional types. The image data, on the other hand, uses deep learning semantic segmentation in computer image recognition technology to extract visual landscape elements.

[0046] For step S3, specifically, constructing a healing heterogeneous map based on the aforementioned individual emotional indicators includes: A user profile is created for the user to be intervened based on the aforementioned basic personal information data. Based on the user profile, the intervention effect indicators, and the individual video preference indicators and healing landscape type preference indicators of the user to be intervened for each of the preset healing landscape videos, a personal efficacy evaluation report is generated for the user to be intervened; wherein, the personal efficacy evaluation report includes user profile analysis, dynamic change analysis of daily emotional indicators, healing landscape type preference and video preference analysis, and intervention effect analysis; Based on the individual efficacy assessment report, the healing heterogeneity map was constructed through multimodal data analysis.

[0047] Preferably, the individual efficacy evaluation report also includes: the intervention treatment plan completion rate, and the specific formula for calculating the intervention treatment plan completion rate is as follows:

[0048] in, and These are the total duration of the video actually watched by the user and the preset intervention target duration, respectively.

[0049] In a preferred embodiment, the personal efficacy assessment report provides the following: user profile analysis, visualization analysis of the dynamic trends of daily mood score indicators, healing landscape type preference analysis and video preference analysis, application usage preference analysis (including the time periods and locations where users frequently use applications), healing landscape intervention program completion rate, and intervention efficacy. Through the information in the report, users can understand the specific impact of the intervention and receive data support for subsequent intervention decisions.

[0050] The completion rate of the intervention treatment plan is expressed as follows:

[0051] In the formula, and These are the total duration of the video actually watched by the user and the pre-set intervention target duration, respectively.

[0052] After obtaining the individual therapeutic effect assessment report, by integrating multi-dimensional feedback data from users during the intervention process and fusing multimodal video data analysis, a heterogeneous map of healing landscape video-user-emotion-preference can be constructed, namely the healing heterogeneous map. Finally, based on the healing heterogeneous map, a closed-loop system of healing landscape-emotion feedback can be formed by mining higher-order associations through graph neural networks.

[0053] Furthermore, adjusting the preset intervention plan according to the healing heterogeneity map specifically includes: Based on the user profile analysis, healing landscape type preference, and video preference analysis, the preset healing landscape video type and the number of daily interventions in the preset intervention plan are adjusted. Based on the dynamic changes of the daily emotion indicators, the emotion trends of the users to be intervened are predicted by a preset emotion trend prediction model, and the number of weeks, weekly intervention frequency, and daily intervention duration of the preset intervention plan are adjusted according to the prediction results.

[0054] In a preferred embodiment, based on the healing heterogeneity map, user profiles, emotional fluctuations, and preference analysis results, as well as the visual and audio features of the videos, a recommendation index is generated for each healing landscape video. This automatically adjusts the type, order, frequency, and number of interventions in the user's intervention plan, making the intervention plan more personalized and aligned with the user's evolving emotional needs. For example, when a user feels anxious, healing landscape videos with tranquil natural environments, soft colors, and soothing natural sounds are recommended based on the emotional intervention model, helping the user gradually relax and restore a positive emotional state. When a user's negative emotions increase, the frequency and number of interventions are suggested to increase based on application usage data, and healing landscape videos associated with positive emotions are recommended.

[0055] Simultaneously, based on users' historical emotional changes, the system proactively optimizes intervention strategies through a real-time feedback mechanism, ensuring the accuracy of its predictions of users' emotional states. Through big data analysis and adaptive learning algorithms, it not only responds to changes in user emotions but also combines historical data and trend prediction models to learn and optimize based on long-term user feedback data, establishing a dynamic adjustment and improvement mechanism. This allows for dynamic adjustments to the healing landscape intervention strategy, providing users with precise intervention plans and ensuring the flexibility of the intervention plan and the long-term effectiveness of the application. If the user's emotional response does not match expectations, the intervention strategy will be automatically adjusted. For example, if the user's emotions are not effectively regulated within a predetermined time, the system will reanalyze and adjust the content of the healing landscape intervention to improve the intervention effect.

[0056] Finally, based on user clicks, views, ratings, and other behaviors and therapeutic effects, a sparse matrix is ​​constructed to mine potential landscape preferences, enabling the recommendation of videos with associated healing effects to similar users (such as patients with moderate depression who are not in remission).

[0057] Compared to existing technologies, the beneficial effects of the therapeutic landscape intervention method based on user feedback provided in this embodiment of the invention are as follows: 1) Proposes a dynamic closed-loop control mechanism for emotion data and landscape intervention: This is the first time a mechanism has been proposed that uses emotion data analysis and automatic adjustment of landscape intervention to respond in real time to changes in users' emotional states, ensuring a high degree of alignment between the landscape intervention plan and the users' emotions. This dynamic intervention model significantly improves the accuracy of intervention effects.

[0058] 2) Propose an adaptive learning and personalized optimization mechanism for emotional data: This invention uses an adaptive learning algorithm to conduct in-depth analysis of long-term user feedback data, which not only improves the effectiveness of personalized intervention, but also addresses the complexity and diversity of user emotional changes, achieving highly personalized therapeutic landscape psychological intervention.

[0059] 3) Achieving multi-dimensional user data fusion: By combining information from multiple dimensions such as emotional data, behavioral data, and environmental feedback, the system can comprehensively understand the user's emotional needs, thereby providing precise intervention measures. This method enhances the flexibility and adaptability of the intervention system.

[0060] 4) Implementing an emotion prediction model based on big data and deep learning: This invention uses big data and deep learning technologies to accurately predict and model user emotions, effectively predicting emotional fluctuations and intervening before they occur, thus preventing emotional deterioration in advance.

[0061] 5) Combine healing landscapes with remote digital psychological intervention: Provide a portable and low-cost form of psychotherapy, increase users' exposure to healing landscapes in daily situations, and break through the time and space constraints of healing in real landscape environments.

[0062] Reference Figure 2 The diagram below shows a schematic of a therapeutic landscape intervention device based on user feedback, provided in another embodiment of the present invention. The device includes: an intervention module 101, an analysis module 102, and an adjustment module 103. The intervention module 101 is used to intervene in the user to be intervened according to a preset intervention plan, and to obtain feedback data of the user to be intervened after the intervention; wherein, the preset intervention plan includes a preset therapeutic landscape video; The analysis module 102 is used to collect multidimensional user data of the user to be intervened, and to preprocess the feedback data and multidimensional user data to obtain several personal emotional indicators corresponding to the user to be intervened. The adjustment module 103 is used to construct a healing heterogeneous map based on the plurality of personal emotional indicators, and to adjust the preset intervention plan based on the healing heterogeneous map.

[0063] Furthermore, the intervention module 101 is used to intervene in the user to be intervened according to a preset intervention plan, specifically including: A preset intervention plan is pre-formulated for the user to be intervened, and the user is intervened on using the preset healing landscape video according to the preset intervention plan; wherein, the preset intervention plan includes the number of intervention weeks, the weekly intervention frequency, the daily intervention duration, the daily intervention frequency, and the preset healing landscape video.

[0064] The user-feedback-based healing landscape intervention device provided in this invention can be deployed on a terminal device or a server. The terminal device includes, but is not limited to, personal computers, smartphones, tablets, wearable devices, etc.; the server can be a single independent server or a cluster system composed of multiple servers. The user-feedback-based healing landscape intervention method provided in this invention can run on a terminal or server, acquiring multi-dimensional user data and optimizing the dynamic healing landscape intervention plan based on emotion analysis and feedback control.

[0065] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A therapeutic landscape intervention method based on user feedback, characterized in that, Includes the following steps: The intervention is carried out on the user to be intervened according to the preset intervention plan, and feedback data of the user to be intervened is obtained after the intervention; wherein, the preset intervention plan includes preset healing landscape videos; Collect multidimensional user data of the user to be intervened, and preprocess the feedback data and multidimensional user data to obtain several personal emotional indicators corresponding to the user to be intervened. A healing heterogeneous map is constructed based on the aforementioned individual emotional indicators, and the preset intervention plan is adjusted based on the healing heterogeneous map.

2. The therapeutic landscape intervention method based on user feedback as described in claim 1, characterized in that, The intervention of the user to be intervened according to the preset intervention plan specifically includes: A preset intervention plan is pre-formulated for the user to be intervened, and the user is intervened on using the preset healing landscape video according to the preset intervention plan; wherein, the preset intervention plan includes the number of intervention weeks, the weekly intervention frequency, the daily intervention duration, the daily intervention frequency, and several different types of the preset healing landscape video.

3. The therapeutic landscape intervention method based on user feedback as described in claim 1, characterized in that, The process of obtaining feedback data from the user to be intervened specifically includes: After the user to be intervened watches the preset healing landscape video, the user is asked to rate their own immediate emotions and the preset healing landscape video, and corresponding immediate emotion feedback data and healing landscape video feedback data are obtained.

4. The therapeutic landscape intervention method based on user feedback as described in claim 3, characterized in that, The collection of multidimensional user data of the user to be intervened specifically includes: Collect the basic personal information data of the user to be intervened; wherein, the basic personal information data includes gender, age, education level, screen time, and self-assessed economic and social status; Clinical assessment data of the users to be intervened are collected; wherein, the clinical assessment data includes depression scores and anxiety scores, and the clinical assessment data are collected at several preset time points; Collect the usage behavior data of the user to be intervened; wherein, the usage behavior data includes the number of clicks, viewing duration and viewing time of each preset healing landscape video by the user to be intervened, as well as the time and latitude and longitude coordinates of the user to be intervened watching the preset healing landscape video every day.

5. The therapeutic landscape intervention method based on user feedback as described in claim 4, characterized in that, The preprocessing of the feedback data and user multidimensional data yields several personal emotional indicators corresponding to the user to be intervened, specifically including: Based on the depression and anxiety scores, quantitative assessments were performed using the Hamilton Depression Rating Scale and the Hamilton Anxiety Rating Scale, respectively, to obtain corresponding depression and anxiety indicators. Based on the real-time emotional feedback data, the positive and negative emotional indices of the user to be intervened are calculated. Based on the feedback data and usage behavior data of the therapeutic landscape videos, the individual video preference index and therapeutic landscape type preference index of the user to be intervened for each of the preset therapeutic landscape videos are calculated. The intervention effect index is calculated based on the depression index, anxiety index, and the positive and negative emotion indices of the users to be intervened.

6. The therapeutic landscape intervention method based on user feedback as described in claim 5, characterized in that, The construction of a healing heterogeneous map based on the aforementioned individual emotional indicators specifically includes: A user profile is created for the user to be intervened based on the aforementioned basic personal information data. Based on the user profile, the intervention effect indicators, and the individual video preference indicators and healing landscape type preference indicators of the user to be intervened for each of the preset healing landscape videos, a personal efficacy evaluation report is generated for the user to be intervened; wherein, the personal efficacy evaluation report includes user profile analysis, dynamic change analysis of daily emotional indicators, healing landscape type preference and video preference analysis, and intervention effect analysis; Based on the individual efficacy assessment report, the healing heterogeneity map was constructed through multimodal data analysis.

7. The therapeutic landscape intervention method based on user feedback as described in claim 6, characterized in that, The adjustment of the preset intervention plan based on the healing heterogeneity map specifically includes: Based on the user profile analysis, healing landscape type preference, and video preference analysis, the preset healing landscape video type and the number of daily interventions in the preset intervention plan are adjusted. Based on the dynamic changes of the daily emotion indicators, the emotion trends of the users to be intervened are predicted by a preset emotion trend prediction model, and the number of weeks, weekly intervention frequency, and daily intervention duration of the preset intervention plan are adjusted according to the prediction results.

8. The therapeutic landscape intervention method based on user feedback as described in claim 6, characterized in that, The individual efficacy evaluation report also includes: the intervention course completion rate, the specific formula for which the intervention course completion rate is calculated is: in, and These are the total duration of the video actually watched by the user and the preset intervention target duration, respectively.

9. A therapeutic landscape intervention device based on user feedback, characterized in that, include: Intervention module, analysis module, and adjustment module; The intervention module is used to intervene in the user to be intervened according to a preset intervention plan, and to obtain feedback data from the user to be intervened after the intervention; wherein, the preset intervention plan includes a preset therapeutic landscape video; The analysis module is used to collect multidimensional user data of the user to be intervened, and to preprocess the feedback data and multidimensional user data to obtain several personal emotional indicators corresponding to the user to be intervened. The adjustment module is used to construct a healing heterogeneous map based on the several personal emotional indicators, and to adjust the preset intervention plan based on the healing heterogeneous map.

10. The therapeutic landscape intervention device based on user feedback as described in claim 9, characterized in that, The intervention module is used to intervene in the user to be intervened according to a preset intervention plan, specifically including: A preset intervention plan is pre-formulated for the user to be intervened, and the user is intervened on using the preset healing landscape video according to the preset intervention plan; wherein, the preset intervention plan includes the number of intervention weeks, the weekly intervention frequency, the daily intervention duration, the daily intervention frequency, and the preset healing landscape video.

Images