Auxiliary evaluation system for future scene guidance
By designing a psychological assessment system for guiding future scenarios, the problems of existing systems being unable to automatically adjust and lacking dynamic risk calculation are solved. This enables dynamic and process-oriented assessment of high-frequency and multimodal scenarios, improving the effectiveness and accuracy of the assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- DONGHUA MEDICAL TECH CO LTD
- Filing Date
- 2025-12-09
- Publication Date
- 2026-05-01
AI Technical Summary
Existing psychological assessment systems cannot automatically adjust to subsequent scenarios based on the context and phased assessment results. They lack dynamic risk calculation and multimodal interfaces, making it difficult to meet the characteristics of future scenario-guided psychological assessment processes.
An auxiliary evaluation system for future scenario guidance was designed, including an evaluation management module, an evaluation control module, and an evaluation process module. By defining evaluation stages and risk probability calculation methods, an evaluation plan is generated and dynamically adjusted during the evaluation process, supporting high-frequency and multimodal evaluations.
It enables the dynamic and process-oriented psychological assessment process guided by future scenarios, improves the effectiveness and accuracy of the assessment, overcomes the obstacles of high-frequency and multimodal assessment, and avoids the "one-size-fits-all" approach in traditional interventions.
Smart Images
Figure CN121964068A_ABST
Abstract
Description
An auxiliary evaluation system for guiding future scenarios Technical Field
[0001] This invention relates to the fields of information technology and psychological assessment, and more specifically, to an auxiliary assessment system for guiding future scenarios. Background Technology
[0002] Future-scenario guided psychological assessment is a method that prioritizes allowing clients to experience the future first. The entire assessment process is divided into multiple stages, generating assessment risks based on the results of each stage, and adjusting the conversation process and steps accordingly. Therefore, future-scenario guided psychological assessment is characterized by its dynamic, high-frequency, multi-modal, and process-oriented nature.
[0003] Currently used psychological assessment systems are static questionnaire streams. They can interact with users step by step according to the assessment items and calculate the user's assessment results. However, the assessment items and interaction steps are fixed in advance and cannot be automatically adjusted for subsequent scenarios based on the scenario and the assessment results at each stage. They also lack dynamic risk calculation and multimodal interfaces, making it difficult to embed them into high-frequency, process-oriented future scenarios.
[0004] Therefore, there is a need for an auxiliary assessment system for guiding future scenarios that meets the characteristics of the psychological assessment process for guiding future scenarios while minimizing physician intervention. Summary of the Invention
[0005] To achieve the above objectives, this application provides an auxiliary assessment system for guiding future scenarios, comprising: an assessment management module, an assessment control module, and an assessment process module; wherein, the assessment management module is used to define assessment stages, assessment standards, and methods for calculating risk probabilities; the execution steps include: defining the assessment session scenarios, assessment content, and assessment stages; determining assessment standards, scoring dimensions, and methods for calculating risk probabilities based on the assessment standards; and setting an assessment plan based on the risk probabilities; wherein, the session scenarios include: initial session, multi-stage scenario sessions, and short test stages; the risk probabilities are categorized as low, medium, and high; the assessment control module is used to execute the assessment plan and initiate assessment events to the assessment process module; the assessment process module is used to respond to assessment events, provide assessment content to the user terminal according to the assessment stages, calculate the risk probabilities of the assessment stages based on user feedback, and generate assessment results and an assessment plan.
[0006] Before the auxiliary assessment system is used, the assessment management module constructs an assessment requirement library. The assessment requirement library includes assessment content, the scenarios corresponding to the assessment content, and assessment control information. The assessment content includes the type of assessment question, the question content, and the question ID. The question types include inducement, sense of presence, and motivation. The scenarios corresponding to the assessment content consist of people, place, time, target action, reward result, emotion, and reminder cues. The assessment control information includes the control information of the assessment process, specifically including the hierarchy of the initial session parameters and the parameter content corresponding to the hierarchy.
[0007] The elements of the initial session parameters include: time distance, specificity, session structure, and whether to add a CB strategy. The initial session parameters are divided into three levels: Level 1, Level 2, and Level 3. Level 1 corresponds to the smallest time distance, indicating that a re-evaluation should be conducted on the same day or the next day. Level 2 corresponds to a time distance indicating that a re-evaluation should be conducted within the week. Level 2 indicates that an evaluation should be conducted after one week. Specificity refers to the strength of the script detail level. Session structure refers to the combination of sessions in the scenario phase.
[0008] Furthermore, the evaluation management module also stores user data corresponding to the evaluation content, including: user's returned content and user's response information; among which, user's response information includes consistency between words and actions, nonverbal engagement, distraction frequency, and delay preference.
[0009] The initial session is the first stage for each user to begin the evaluation; the multi-stage scenario session includes multiple stages, each stage containing at least a combination of people, place, time, target action, reward result, emotion, and reminder cues; the short test stage sets the delay preference baseline, defines the delay items and their corresponding durations, and the evaluation content of the delay items includes stickers, free time, and activity time; the evaluation stage also includes a flexible selection stage as a micro-unit of cognitive behavioral strategies.
[0010] The scoring dimensions include: delayed selection rate, subjective rating, motivation rating, incentive, consistency of speech and behavior, nonverbal engagement, and distraction frequency rating; the algorithm for the probability p of insufficient future orientation risk is expressed as follows: Where DLR is the delayed choice ratio score, PRES is the subjective score, MOT is the motivation score, TRIG is the incentive average score, CONS is the verbal-behavioral consistency score, NVERB is the nonverbal engagement score, and DIST is the distraction frequency score; a1 to a7 are the weights of each score, and i is the auxiliary parameter.
[0011] Furthermore, the content of the evaluation plan is determined by the initial session parameters, specificity content, session structure, and CB policy; the evaluation plan also includes the start time, which is determined based on the time distance and the generation time of the evaluation plan.
[0012] The evaluation control module executes evaluation plan caching and evaluation control. The evaluation plan caching is implemented by a job queue, which includes user identifiers and evaluation plans. After the evaluation management module generates an evaluation plan, it is stored in the job queue and retrieved in the evaluation control step. The evaluation control is implemented by a time-based task process or a scheduled task program. It compares the current time with the start time in the evaluation plan, retrieves the evaluation plan from the job queue, generates an evaluation event, and starts the evaluation process.
[0013] Furthermore, the assessment process module performs the following steps: initiate the assessment and generate assessment content; determine the stage to be assessed based on the assessment content generated when the assessment is initiated; interact with the assessment management module to extract assessment content based on the stage to be assessed; after interacting with the user and doctor terminals, obtain the score values and risk probabilities of each dimension; generate assessment results and assessment plans based on the obtained score values and risk probabilities of each dimension; determine whether there are any incomplete assessment stages; if there are, continue to determine the assessment stage; otherwise, end the assessment.
[0014] The evaluation results include generating an evaluation report. The indicators considered when generating the evaluation report include: Current Delayed Selection Ratio (DLR_now): the number of questions selected "More Later" in this N-question short test ÷ N, with a value between 0 and 1; Previous Delayed Selection Ratio (DLR_prev): the DLR_now of the previous session; Baseline Delayed Selection Ratio (DLR_base): the number of questions selected "More Later" in the baseline N-question test on the first day ÷ N, with a value between 0 and 1; Delayed Selection Ratio Difference (ΔDLR): ΔDLR = DLR_now - DLR_prev. If no DLR_prev was generated, this indicator is not used; Presence (PRES) and Motivation (MOT): scores ranging from 1 to 5 points respectively; Accuracy (GNG): the accuracy rate of the Go / No-Go suppression task, scored as a percentage. These indicators are used to determine whether the user's status is improving, regressing, or remaining flat. If ΔDLR ≥ 1 / 3, it is considered an improving status; if ΔDLR < 1 / 3, it is considered a non-improving status. If the score is ≤−1 / 3, it is considered a regression; otherwise, it is considered a flat state. If the PRES or MOT score improves by ≥0.5 points compared to the previous score, it is considered to have "subjective progress".
[0015] According to the present invention, the assessment process guided by future scenarios is reproducible and traceable, effectively solving the assessment obstacles caused by the high frequency and multimodal characteristics of such psychological assessment processes. Furthermore, through the risk probability design in this invention, risks are divided into three levels: low, medium, and high, and mapped to different initial parameters and different assessment plan contents. This avoids the "one-size-fits-all" approach in traditional interventions, achieving a dynamic and process-oriented approach to long-term assessments, and improving the effectiveness and accuracy of user psychological assessments. Attached Figure Description
[0016] Figure 1 is a schematic diagram of the structure of an auxiliary evaluation system for guiding future scenarios provided according to an embodiment of the present invention. Detailed Implementation
[0017] The specific implementation of the present invention will now be described in detail with reference to the accompanying drawings.
[0018] Figure 1 provides a schematic diagram of the structure of the auxiliary assessment system for future scenario guidance. As shown, it includes: P100 assessment management module, P110 assessment control module, and P120 assessment process module, as follows: The P100 assessment management module is used to define the assessment stage, define the assessment criteria, and define the calculation method for risk probability. Before the auxiliary assessment system is used, an assessment requirement library is built, which includes assessment content, the scenario corresponding to the assessment content, and assessment control information.
[0019] The assessment content includes the type of questions, the content of the questions, and the question ID.
[0020] Specific problem types include at least the triggers (including: phone / game distractions, immediate preferences, emotionally driven impulses, and little thought about the future), presence, and motivation; the scenarios corresponding to the assessment content consist of: people, place, time, target action, reward result, emotions, and cueing cues (objects or words).
[0021] The assessment control information includes control information for the assessment process, specifically the hierarchy of initial session parameters and the corresponding parameter content for each level. Initial session parameters include: time interval, specificity, session structure, and whether to add a CB (Content Decision) strategy. The initial session parameter hierarchy includes levels one, two, and three; level one corresponds to the shortest time interval, requiring a reassessment on the same day or the next day; level two corresponds to a reassessment within the week; and level two indicates an assessment after one week. Specificity refers to the intensity of script detail levels (location details / objects / social feedback, etc.). Session structure refers to the combination of sessions across the three scenario phases (all three phases are executed, or a lightweight session); a lightweight session refers to executing only phases one and two, with the question content corresponding to the chosen motivation being a reduced-load version.
[0022] The assessment management module also stores user data corresponding to the assessment content. The user data corresponding to the assessment content includes: the user's returned content and the user's response information (such as: completion level, facial expression, posture, speech rate and other status information). The user's returned content can be returned directly by the user or by the doctor. The user's response information is obtained from the user's response process, and can also be filled in by the doctor or returned by a third-party system after video capture.
[0023] The user's response information includes consistency between words and actions (whether the user has filled in all six slots as prompted and repeated the prompt words), nonverbal engagement (including anchor scores such as eye contact, nodding, posture, and speech rate stability), distraction frequency (number of times the user is distracted, plays on their phone, and goes off-topic), and delay preference (i.e., the preference for choosing "immediately a small amount" and "later more").
[0024] Based on the assessment requirements library, the assessment management module performs the following steps based on the characteristics of future scenario guidance: Step S101: Define the assessment session scenario, assessment content, and assessment stage: The assessment stages involved in future scenario-guided assessment include the initial session and multi-stage scenario sessions; 1) The initial session is the first stage for each user to begin the assessment, during which a questionnaire and self-assessment are conducted; In this step, the question types and the corresponding number of questions for the questionnaire are determined, which are used to extract the corresponding assessment content based on the number during the assessment process and output it to the user end and the doctor end; At the same time, the type of user response information to be considered and the corresponding content are determined, which can be used to extract the response content from the user end or the doctor end to obtain the corresponding score during the subsequent assessment process; Further, the delay preference is determined to obtain the impulsivity level.
[0025] 2) Multi-stage scenario sessions: This step defines the scenario sessions and session durations for each stage, which will be used to generate evaluation content during the subsequent evaluation process.
[0026] A multi-stage scenario conversation comprises multiple stages, each stage containing at least a combination of elements corresponding to the emotion, such as characters, location, time, target action, reward result, emotion, and cueing. For example, it could be defined as three stages with a total duration of 30 minutes: stage one is a combination of time and location, stage two is a combination of target action and cueing, and stage three is a combination of reward result and emotion, with characters and scenarios randomly added to each of the three groups.
[0027] This step defines the phases, the corresponding problem types, and the number of problems for each type.
[0028] 3) Set short-test phases within, between, and after each phase, i.e. set a baseline for delay preference, define delay items and their corresponding durations; the evaluation content of delay items includes stickers, free time, and activity time; for example, after the first session, specify 30 minutes of free time as the delay item.
[0029] In addition, the CB strategy phase is defined as a flexible choice phase, which is a micro-unit of cognitive-behavioral strategy. In the evaluation of this phase, the self-reward of the evaluation user is realized.
[0030] Step S102: Determine the evaluation criteria, scoring dimensions, and the method for calculating the risk probability based on the evaluation criteria; First, determine the evaluation criteria, that is, the scores corresponding to each question; Determine the scoring dimensions and the weights corresponding to each scoring dimension. The scoring dimensions are the induction of questions, user performance, and delay preferences, and the average score or total score is calculated to participate in the subsequent calculation of the risk probability calculation dimension; The scoring dimensions determined in this step include: the proportion of delayed choices, subjective scoring, motivation scoring, incentives, consistency of words and deeds, non-verbal participation, and distraction frequency scoring; Among them, the dimensions from the questions include: subjective judgment, motivation, incentives; The dimensions from user performance include: consistency of words and deeds, language participation, distraction frequency.
[0031] Calculate the scoring values for each dimension through the scores or average scores of the questions corresponding to each dimension; Before participating in the subsequent calculation, normalize each scoring value.
[0032] In this step, use the form of logistic regression to define the algorithm for the risk probability p of insufficient future orientation, expressed as: , where DLR is the score of the proportion of delayed choices, PRES is the subjective score, MOT is the motivation score, TRIG is the average score of incentives, CONS is the score of consistency of words and deeds, NVERB is the score of non-verbal participation, and DIST is the score of distraction frequency; a1 to a7 are the weights of each score, and i is an auxiliary parameter.
[0033] For example, first use the default parameters: i = 0, a1 = -1.2, a2 = -0.6, a3 = -0.4, a4 = 0.6, a5 = -0.5, a6 = -0.4, a7 = 0.6, and these can be reset according to the sample situation in subsequent applications.
[0034] The risk probability p is used to judge the risk of insufficient future orientation. In this step, the judgment threshold of the risk level can be set. Generally speaking, the higher the sense of presence or motivation, the higher the proportion of delayed choices, and the better the external observation consistency and participation, the lower the risk probability p; The stronger the incentive and the more distractions, the higher the risk probability p. The judgment threshold of the risk level is set as: p < 0.33 low risk; 0.33 < p < 0.66 medium risk; p ≥ 0.66 high risk, and it supports adjustment according to the sample distribution.
[0035] Step S103: Set the evaluation plan according to the risk probability in the evaluation stage; In this step, determine the next-stage evaluation plan according to the evaluation stage and the corresponding risk probability in the evaluation stage. The content of the evaluation plan includes: start time, evaluation stage, delay preference, and CB strategy script; The specific content of the evaluation plan is reflected in the starting session parameters, specific content, session structure, and CB strategy; The start time is determined according to the time distance and the generation time of the evaluation plan.
[0036] The specific determination rules are shown in the table below:
[0037] During the evaluation process, the evaluation management module calls the evaluation management module to generate an evaluation plan, which is then stored in the evaluation control module. The evaluation control module initiates the evaluation in the evaluation process module based on the start time or other start events.
[0038] The P110 evaluation control module executes the evaluation plan and initiates evaluation events to the evaluation process module. As shown in Figure 1, the evaluation control module includes S111 evaluation plan cache and S112 evaluation control. The evaluation plan cache is implemented by a job queue, which includes user identifiers and evaluation plans. After the evaluation management module generates the evaluation plan, it is stored in the job queue and retrieved during the evaluation control step. The evaluation control is implemented by a time-based task process or a scheduled task program. It compares the current time with the start time in the evaluation plan. When the start time matches the current time, it retrieves the evaluation plan from the job queue, generates an evaluation event, and starts the evaluation process. When the evaluation process starts, an evaluation event is sent to the evaluation process module.
[0039] The P120 assessment process module is used to interact with both the user and doctor terminals, respond to assessment events, provide assessment content to the user terminal according to the assessment stage, calculate the risk probability of the assessment stage based on the user terminal feedback, and generate assessment results and assessment plans.
[0040] The sources of assessment events include assessment events generated by the assessment control module, or a new assessment initiated by a doctor / user.
[0041] Specifically, the specific assessment process is shown in Figure 1: Step S121: Start the assessment: The assessment process disclosed in this invention can be initiated by the doctor or the user, or the assessment control module can continue the existing assessment process into a new assessment stage according to the assessment plan.
[0042] 1) If starting a new assessment: initialize the assessment content; 2) If continuing an existing assessment process, load the assessment data already completed in the current assessment process and obtain the assessment content to be completed next from the assessment event information.
[0043] Specifically, the evaluation includes initial session parameters, specificity content, session structure, and CB policy.
[0044] Step S122: Determine the evaluation stage: Based on the evaluation content generated when the evaluation is started, determine the stage to be evaluated; if it is starting a new evaluation, that is, the evaluation content is the initialization content, then the stage to be evaluated is the first session stage; if it is continuing an existing evaluation process, determine which stages need to be evaluated in the evaluation plan, and generate the stages to be evaluated based on the stages completed in the previous evaluation.
[0045] Step S123: Interact with the evaluation management module to extract evaluation content; in this step, obtain the content to be evaluated according to the stage to be evaluated, including the initial session parameters, specificity content, session structure and CB policy.
[0046] After acquiring the content to be evaluated, the corresponding scenario conversation for the loading stage is loaded, and interactive content is provided to the user and doctor. This includes scenario modeling, generating scenario forms that need to be filled out by the user or by the doctor, with content including: people, location, time, target action, reward / result, 2 to 3 words of emotion and reminder cues.
[0047] In this step, based on the corresponding items in the content to be evaluated, the corresponding scenario and evaluation content are retrieved from the evaluation management module. The evaluation content is then pushed to the doctor's and user's terminals according to the scenario configuration, guiding the user and doctor to conduct the evaluation.
[0048] Step S124: After interacting with the user and doctor, obtain the risk probability of the assessment stage; based on the interaction with the user and doctor in step S123, obtain the score value of each item in the assessment of this stage, and calculate the score value of each dimension.
[0049] Step S125: Generate assessment results and assessment plan based on the obtained scores of each dimension; in this step, based on the probability p of insufficient future guidance risk obtained from the completed assessment content and the assessment values of each assessment dimension, determine whether it is necessary to plan the next step of the assessment.
[0050] An evaluation plan is generated based on the judgment results. The evaluation plan includes the start time of the next stage of evaluation, the specific evaluation content, and other information. The evaluation plan is then sent to the P110 evaluation control module.
[0051] When generating the assessment results in this step, the assessment report is output to both the user and the doctor's end; when generating the assessment report, subjective and objective indicators are combined to generate the recommendations and trend analysis.
[0052] The subjective and objective indicators to be considered when generating the evaluation report include: 1) Current delayed selection ratio DLR_now: the number of questions selected "more later" in this N-question short test ÷ N, with a value of 0~1; 2) Previous delayed selection ratio DLR_prev: the DLR_now of the previous session; 3) Baseline delayed selection ratio DLR_base: the number of questions selected "more later" in the baseline N-question test on the first day ÷ N, with a value of 0~1; 4) Delayed selection ratio difference ΔDLR: ΔDLR = DLR_now - DLR_prev. If no DLR_prev is generated, this indicator is not used; 5) Presence perception PRES and Motivation Motion MOT: scores range from 1 to 5 points respectively; 6) Accuracy GNG: the accuracy rate of the Go / No-Go inhibition task, with a score of percentage (%).
[0053] Based on the above indicators, we can determine whether a user's status is progressing, regressing, or remaining flat. If ΔDLR ≥ 1 / 3, it is considered a progressing state; if ΔDLR ≤ −1 / 3, it is considered a regressing state; otherwise, it is considered a flat state.
[0054] If the PRES or MOT score improves by ≥ 0.5 points compared to the previous score, it is considered to have "subjective improvement".
[0055] On the other hand, an evaluation plan can be generated based on the current ΔDLR / PRES / MOT (and optional GNG), which is the starting session parameter for the next evaluation.
[0056] Step S126: Determine if there are any unfinished evaluation stages. If there are unfinished evaluation stages, continue to step S122 to continue determining the evaluation stages and extracting evaluation content for the evaluation process; otherwise, end the evaluation and complete this round of evaluation.
[0057] This invention targets psychological assessment scenarios for Future Scenario Guidance (EFT), employing a closed-loop "assessment-measurement-planning" approach. Unlike common EFT exercises that only provide scenario visualization, this scheme immediately conducts a short-term delayed preference test after each guidance session, using quantitative indicators such as the Delayed Choice Ratio (DLR) to provide feedback on the effectiveness of that session. Based on this, it triggers adaptive rules with fixed thresholds (automatic adjustments to time distance, specificity, reward salience, and session length), making the assessment process reproducible and traceable, effectively overcoming assessment obstacles caused by high frequency and multiple modalities. Specifically, by designing risk probabilities, risks are divided into low / medium / high levels and mapped to different initial parameters and different assessment plan schemes, avoiding the "one-size-fits-all" approach of traditional interventions. By setting future assessment schemes based on the current assessment, it achieves a dynamic and process-oriented assessment process over a long period, improving the effectiveness and accuracy of user psychological assessments.
[0058] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the present invention is not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.
Claims
1. An auxiliary evaluation system for guiding future scenarios, characterized in that, include: The system comprises an assessment management module, an assessment control module, and an assessment process module. The assessment management module defines assessment stages, assessment criteria, and methods for calculating risk probabilities. Execution steps include: defining assessment session scenarios, assessment content, and assessment stages; determining assessment criteria, scoring dimensions, and methods for calculating risk probabilities based on the assessment criteria; and setting an assessment plan based on the risk probabilities. The session scenarios include: initial session, multi-stage scenario sessions, and short test stages. The risk probabilities are categorized as low, medium, and high. The assessment control module executes the assessment plan and initiates assessment events to the assessment process module. The assessment process module responds to assessment events, provides assessment content to the user based on the assessment stage, calculates the risk probabilities for each assessment stage based on user feedback, and generates assessment results and an assessment plan.
2. The auxiliary evaluation system according to claim 1, characterized in that, Before the auxiliary assessment system is used, the assessment management module constructs an assessment requirement library. The assessment requirement library includes assessment content, the scenarios corresponding to the assessment content, and assessment control information. The assessment content includes the assessment question type and question content, and question ID. The question type includes inducement, sense of presence, and motivation. The scenario corresponding to the assessment content consists of people, place, time, target action, reward result, emotion, and reminder cues. The assessment control information includes control information for the assessment process, specifically including the hierarchy of the initial session parameters and the parameter content corresponding to the hierarchy.
3. The auxiliary evaluation system according to claim 2, characterized in that, The elements of the starting session parameters include: time distance, specificity, session structure, and whether to add a CB strategy; wherein, the starting session parameters are divided into three levels: Level 1, Level 2, and Level 3; Level 1 corresponds to the smallest time distance, indicating that a re-evaluation should be performed on the same day or the next day; Level 2 corresponds to a time distance indicating that a re-evaluation should be performed within the week; Level 2 indicates that an evaluation should be performed after one week; the specificity refers to the strength of the script detail level; the session structure refers to the combination of sessions in the scenario stage.
4. The auxiliary evaluation system according to claim 1, characterized in that, The evaluation management module also stores user data corresponding to the evaluation content, which includes: the user's returned content and the user's response information; among which, the user's response information includes consistency between words and actions, nonverbal engagement, distraction frequency, and delay preference.
5. The auxiliary evaluation system according to claim 1, characterized in that, The initial session is the first stage in which each user begins the evaluation; the multi-stage scenario session includes multiple stages, each stage containing at least a combination of people, place, time, target action, reward result, emotion, and reminder cues; the short test stage sets a baseline for delay preference, defines delay items and their corresponding durations, and the evaluation content of delay items includes stickers, free time, and activity time; the evaluation stage also includes a flexible selection stage as a micro-unit of cognitive behavioral strategies.
6. The auxiliary evaluation system according to claim 1, characterized in that, The scoring dimensions include: delayed selection ratio, subjective rating, motivation rating, incentive, consistency of speech and behavior, nonverbal engagement, and distraction frequency rating; the algorithm for the probability p of insufficient future orientation risk is expressed as follows: Where DLR is the delayed choice ratio score, PRES is the subjective score, MOT is the motivation score, TRIG is the incentive average score, CONS is the verbal-behavioral consistency score, NVERB is the nonverbal engagement score, and DIST is the distraction frequency score; a1 to a7 are the weights of each score, and i is the auxiliary parameter.
7. The auxiliary evaluation system according to claim 3, characterized in that, The content of the evaluation plan is determined by the initial session parameters, specificity content, session structure, and CB policy; the evaluation plan also includes a start time, which is determined based on the time distance and the generation time of the evaluation plan.
8. The auxiliary evaluation system according to claim 1, characterized in that, The evaluation control module executes evaluation plan caching and evaluation control; wherein, the evaluation plan caching is implemented by a job queue, which includes user identifiers and evaluation plans; after the evaluation management module generates an evaluation plan, the evaluation plan is stored in the job queue and retrieved in the evaluation control step; the evaluation control is implemented by a time task process or a scheduled task program, which compares the current time with the start time in the evaluation plan, retrieves the evaluation plan from the job queue, generates an evaluation event, and starts the evaluation process.
9. The auxiliary evaluation system according to claim 1, characterized in that, The evaluation process module performs the following steps: initiating the evaluation and generating evaluation content; determining the stage to be evaluated based on the evaluation content generated when the evaluation is initiated; and interacting with the evaluation management module to extract the evaluation content based on the stage to be evaluated. After interacting with users and doctors, obtain the scores and risk probabilities for each dimension; Based on the obtained scores and risk probabilities for each dimension, generate assessment results and assessment plans; determine if there are any incomplete assessment stages. If there are, continue to determine the assessment stages; otherwise, end the assessment.
10. The auxiliary evaluation system according to claim 9, characterized in that, The generated evaluation results include the generation of an evaluation report. The indicators considered when generating the evaluation report include: Current Delayed Selection Ratio (DLR_now): the number of questions selected "More Later" in this N-question short test ÷ N, with a value ranging from 0 to 1; Previous Delayed Selection Ratio (DLR_prev): the DLR_now of the previous session; Baseline Delayed Selection Ratio (DLR_base): the number of questions selected "More Later" in the baseline N-question test on the first day ÷ N, with a value ranging from 0 to 1; Delayed Selection Ratio Difference (ΔDLR): ΔDLR = DLR_now - DLR_prev. If no DLR_prev was generated, this indicator is not used; Presence (PRES) and Motivation (MOT): scores ranging from 1 to 5 points respectively; Accuracy (GNG): the accuracy rate of the Go / No-Go suppression task, scored as a percentage. These indicators can be used to determine whether the user's status is improving, regressing, or remaining flat. If ΔDLR ≥ 1 / 3, it is considered an improving status; if ΔDLR ≤ 1 / 3, it is considered a regressing status. If the score is -1 / 3, it is considered a regression; otherwise, it is considered a flat state. If the PRES or MOT score improves by ≥ 0.5 points compared to the previous score, it is considered to have "subjective progress".