An assisted decision-making opinion recommendation method and system for mood disorders

Abstract

This application relates to a method and system for recommending auxiliary decision-making opinions for mood disorders. The method includes collecting patient user data; standardizing and scoring the user data to obtain scoring results; determining the severity level of the patient's illness based on the scoring results; and setting initial resource allocation weights for multiple treatment methods based on the severity level; acquiring the patient's psychological characteristic data and correcting the initial resource allocation weights to generate auxiliary decision-making opinions. Doctors refer to the auxiliary decision-making opinions and use personalized treatment plans to treat patients, collecting the patient's digital monitoring indicators in real time during treatment; dynamically adjusting / intervening in the personalized treatment plan based on the digital monitoring indicators and implementing it; and conducting phased evaluations of the treatment execution data according to preset efficacy evaluation criteria. This application significantly improves the efficiency and safety of mood disorder assessment by assessing disease severity through multimodal data, dynamically configuring treatment weights, and making personalized corrections.

Description

Technical Field

[0001] This application relates to the technical field of smart healthcare, and in particular to a method and system for recommending auxiliary decision-making opinions for mood disorders. Background Technology

[0002] Mood disorders, especially depression, are common and serious mental illnesses worldwide. According to the World Health Organization, approximately one billion people worldwide suffer from mental disorders. The current state of depression treatment is worrying, with less than 25% of patients receiving standardized treatment. Among patients receiving traditional oral medications or electroconvulsive therapy, less than 30% complete the full course of treatment, and the underlying cause is not fundamentally controlled.

[0003] However, current technologies still have many shortcomings in improving mood disorders. First, for populations with limited cognitive function (such as young children, those with low education levels, and elderly patients), existing methods are difficult to achieve optimal efficacy with low burden, especially those relying on anesthetic drugs, which are costly, require in-person visits and injections, leading to relatively high patient resistance. Second, patients often have some resistance to improvement, and may conceal their condition or provide false information during consultations, further increasing the difficulty of diagnosis for doctors. Third, existing improvement programs mostly focus on treatment-resistant depression, primarily using ketamine combined with fixed oral medications, limiting their role to auxiliary improvement and lacking systematic coverage and management of mild to moderate symptoms, somatization symptoms, and the process of medication tapering and discontinuation. In addition, current technologies for processing multimodal medical data are mostly limited to simple superposition or superficial comparison, lacking the ability to mine deep feature correlations between data, resulting in insufficient early prediction of improvement efficacy and side effects, a black box decision-making process, poor interpretability, and difficulty in gaining the trust of clinicians. For newly enrolled patients, the lack of individual historical data makes it difficult to quickly build an effective personalized improvement model, resulting in a cold start problem. At the same time, patients' physiological and psychological states change dynamically over time, making it difficult for fixed models to achieve long-term adaptation.

[0004] In existing technologies, for example, CN120241065A discloses a mood disorder assessment system and method. This method collects eye-tracking videos of users performing free viewing, gazing, and attention shifting tasks, extracts eye-tracking features such as fixation time, saccade count, and attention shift latency, and inputs them into a logistic regression model to output a mood disorder assessment score. However, it adopts a static, single-shot acquisition mode, which makes it difficult to continuously track and provide real-time feedback on the dynamic changes in the patient's condition, and is not suitable for the long course and relapse-prone characteristics of mood disorders. This approach relies on eye-tracking data, which cannot comprehensively reflect the complex condition of the patient. This approach is difficult to achieve long-term dynamic management and has limited guiding value for clinical treatment decisions. For example, CN112259237B discloses a depression assessment system, which uses a stimulation module to provide subjects with exogenous stimulation (neutral, negative, and positive emotional stimulation in the form of pictures and videos) and endogenous stimulation (recalling past experiences). It uses a physiological signal acquisition module to collect EEG signals, skin conductance signals, and eye movement information. It uses a physiological signal analysis module to clean and extract features from the physiological information. Finally, it uses a depression assessment module to input the extracted features into a multi-level classification model, first to determine whether it is depression, then to assess the depression subtype, and output the depression assessment result.

[0005] However, due to the long treatment period and numerous influencing factors of mood disorders, doctors still face considerable pressure in the current diagnostic or assessment process. Summary of the Invention

[0006] The purpose of this application is to provide a method and system for assessing mood disorders, which partially solves or alleviates the aforementioned shortcomings of existing technologies. It achieves accurate classification of disease severity through multimodal data collection and standardized scoring, and dynamically configures the weights of multiple means, with personal cognitive adjustment, social support, drug treatment and anesthetic treatment as the core, based on the classification results. At the same time, it makes personalized adjustments in combination with the patient's psychological characteristics. During the treatment process, it relies on real-time digital monitoring to achieve proactive warning of side effects and dynamic adjustment of the treatment plan. Through a phased assessment and closed-loop iteration mechanism, it promotes the management of the entire course of the disease until the cure endpoint of discontinuing oral drugs and anesthetic treatment is reached, thereby significantly improving the efficiency, safety and resource utilization of mood disorder assessment.

[0007] To solve the aforementioned technical problems, this application adopts the following technical solution: The first aspect of this application is to provide a method for recommending auxiliary decision-making opinions for mood disorders, comprising the following steps: S1: Collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators; wherein, S1 includes: pre-labeling each data item in the user data with a corresponding privacy level and detection difficulty level; selecting a collection strategy according to the patient's stage of treatment, the collection strategy including: privacy threshold and / or difficulty threshold; selecting the user data according to the collection strategy, wherein the privacy level of the user data is lower than the privacy threshold, or the detection difficulty level of the user data is lower than the difficulty threshold; S2: Standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set initial resource allocation weights for multiple treatment methods based on the severity level of the disease; S3: Obtain the patient's psychological characteristic data to update the initial resource allocation weights to obtain auxiliary decision-making opinions, wherein the psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine, correct the initial resource allocation weights based on the psychological characteristic data to generate at least one personalized resource allocation weight, and output the combination of the personalized resource allocation weights as auxiliary decision-making opinions.

[0008] In some embodiments, the multiple treatment methods in step S2 include at least one of anesthesia, pharmacology, personal cognitive adjustment, nutritional and exercise support, social support intervention, and physical therapy. In some embodiments, the patient self-reported data collected in step S1 is compared with objective test data to determine the consistency and reliability between the patient self-reported data and the objective test data; the objective test data includes at least auxiliary examination data, digital monitoring indicators, and BMI index in basic information; when there is a conflict between the patient self-reported data and the objective test data exceeding a preset threshold, concealment behavior is identified; when concealment behavior is identified, the initial resource allocation weights are modified, including increasing the initial weights of social support intervention and personal cognitive adjustment in the initial resource allocation weights. In some embodiments, the method further includes the step of: when concealment is identified, identifying the severity of concealment; setting a privacy threshold and / or a difficulty threshold according to the severity.

[0009] In some embodiments, the method includes the steps of: identifying the magnitude of change in personalized resource allocation weights within a set period; and triggering a threshold update when the magnitude of change is less than a set magnitude value; wherein the threshold update is an update of a privacy threshold or an update of a difficulty threshold. In some embodiments, the disease severity grading in step S2 includes a mild stage, a moderate stage, and a severe stage; in some embodiments, the configuration rules for the initial resource allocation weights are as follows: in the mild stage, personal cognitive adjustment and social support interventions are the main focus; in the moderate stage, drug treatment is initiated, and the weights of nutritional exercise support and physical therapy are increased; in the severe stage, anesthesia and drug treatment are the main focus, and social support interventions are strengthened simultaneously.

[0010] In some embodiments, in step S2, the user data is standardized and scored to obtain a scoring result, based on... The scoring results used to determine the severity of a patient's disease include at least one of the following steps: Step a: Basic information scoring, at least gender, age, and BMI are graded according to a preset grading standard and assigned corresponding scores; Step b: Medical history data scoring, at least the course of the disease, severity of core symptoms, accompanying symptoms, medication history, negative events, and family history are graded according to a preset grading standard and assigned corresponding scores; Step c: Neuropsychiatric examination scoring, at least communication status, emotional activity, and cognitive function indicators are graded according to a preset grading standard and assigned corresponding scores; Step d: Questionnaire assessment scoring, at least sleep quality, anxiety and depression symptoms, suicide risk, social support level, and cognitive function indicators are graded according to a preset grading standard and assigned corresponding scores; Step e: Ancillary examination scoring, at least organ function indicators, neuroendocrine indicators, nutritional status indicators, and drug metabolism indicators are graded according to a preset grading standard and assigned corresponding scores; Step f: Comprehensive score calculation, summing the scores assigned in steps a to e to obtain the patient's comprehensive score, and determining the severity of the disease based on the comprehensive score.

[0011] In some embodiments, in step S2, when the scoring data for all sub-items cannot be obtained, a preset core item is used. The system assigns values ​​to determine the severity level of a patient's disease and acquires data related to the scoring during treatment to correct the severity level.

[0012] In some embodiments, step S3 further includes: classifying patients into different personality types based on their cognitive level of depression and their trust in medicine; the personality types include high cognition-high trust type, high cognition-low trust type, low cognition-high trust type, and low cognition-low trust type; selecting a corresponding weight correction scheme according to the personality type to correct the initial resource allocation weights to generate personalized resource allocation weights, including: If the personality type is high cognition-high trust, increase the weight of personal cognitive adjustment and social support intervention; if the personality type is high cognition-low trust, while maintaining the weight of necessary anesthesia treatment, strengthen the weight of social support intervention and personal cognitive adjustment; if the personality type is low cognition-high trust, increase the weight of anesthesia treatment, drug treatment and social support intervention, and decrease the weight of personal cognitive adjustment; if the personality type is low cognition-low trust, increase the weight of short-term anesthesia treatment.

[0013] In some embodiments, the method further includes the steps of: labeling each data item in the user data with a privacy level and a detection difficulty level according to preset rules to obtain a preset initial collection strategy; the privacy level includes at least high privacy level data and low privacy level data, and the detection difficulty level includes at least high difficulty level data and low difficulty level data; when the patient visits for the first time, based on the preset initial collection strategy, low privacy level data and low difficulty level data are collected first.

[0014] In some embodiments, during the treatment process, the current efficacy level is obtained. If the current efficacy level reaches a preset efficacy threshold, high privacy level data and high difficulty level data are collected.

[0015] The present invention also provides an auxiliary decision-making recommendation system for mood disorders, the system comprising: The user data collection module is configured to collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators. Each data item in the user data is pre-labeled with a corresponding privacy level and detection difficulty level. A collection strategy is selected based on the patient's stage of treatment, and the collection strategy includes a privacy threshold and / or a difficulty threshold. User data is selected according to the collection strategy, wherein the privacy level of the user data is lower than the privacy threshold, or the difficulty level of the user data is lower than the difficulty threshold. An initial resource allocation module is configured to standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set initial resource allocation weights for multiple treatment methods based on the severity level of the disease. An auxiliary decision generation module is configured to acquire the patient's psychological characteristic data to update the initial resource allocation weights to obtain auxiliary decision opinions. The psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine. The initial resource allocation weights are corrected based on the psychological characteristic data to generate at least one personalized resource allocation weight, and the combination of the personalized resource allocation weights is output as auxiliary decision opinions.

[0016] Beneficial Technical Effects: Unlike conventional disease diagnosis and treatment, the evaluation of treatment plans for emotional disorders requires consideration of more detailed dimensions, which places extremely high demands on doctors' decision-making. To address this, this invention provides a decision support system that assists doctors in preliminary data analysis and processing from a computer perspective. Based on the decision support suggestions, doctors can then set treatment plans according to their practical experience.

[0017] However, there are several challenges in the computer processing: 1) There are many data dimensions to consider, making analysis very difficult; 2) Data collection sources are wide-ranging, including both objective detection indicators and some psychological or emotional indicators that are difficult to quantify. As a result, there may be biases in the collection process, and patients may even provide false information due to their own resistance, interfering with the computer's decision-making.

[0018] In this embodiment, it is preferable to use multidimensional objective indicators plus relatively easy-to-verify subjective self-reports (such as lifestyle habits) to set the initial weights of the treatment methods; then, psychological characteristic data with higher subjectivity and greater difficulty in judgment are used as fine-tuning indicators, that is, under the initial weight system, psychological characteristic data are used to make minor adjustments to one or more of the weights.

[0019] Furthermore, to improve the reliability of highly subjective and difficult-to-distinguish psychological characteristic data, this embodiment classifies and positions the psychological characteristic data based on the patient's personal cognition (such as their depth of understanding or acceptance of emotional disorders) and trust level (such as their level of trust in the doctor, or their adherence to the treatment plan), attempting to define it using indicators that are relatively easy to quantify or determine. Thus, by synergistically combining objective indicators, easily verifiable subjective self-reports, and relatively easily quantifiable psychological indicators, step-by-step weighting is achieved, enabling the computer to integrate objective and subjective multidimensional data for autonomous assisted decision-making.

[0020] In order to reduce the initial data analysis burden on doctors and to automatically eliminate interfering factors (such as patients concealing or misreporting information) to a certain extent, this embodiment adopts a recommendation scheme for making recommendations on the proportion of treatment based on subjective and objective data.

[0021] This embodiment can add an update mechanism for situations where weight adjustment stagnates. It is understood that the assessment dimensions of mood disorders are numerous and the rules are complex, making it difficult for computers to make reliable proactive judgments without the assistance of doctors' experience. Therefore, this update mechanism also serves as an additional alert for stagnation issues, providing doctors with supplementary suggestions and guiding them to focus on the causes of the stagnation.

[0022] From another perspective, the present invention can achieve the following beneficial technical effects: 1. This application establishes a standardized scoring system for multimodal data, encompassing basic information, medical history, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators. Combined with a disease severity grading mechanism, it achieves an objective, comprehensive, and quantitative assessment of the patient's condition. This overcomes the shortcomings of traditional diagnosis and treatment, such as over-reliance on physicians' subjective experience, inconsistent assessment standards, and significant differences in judgment among different physicians. It significantly improves the accuracy, consistency, and clinical reproducibility of disease grading. Furthermore, when scoring data for all sub-items is unavailable, it prioritizes collecting core symptom severity, suicide risk, key medical history information, age, and BMI for scoring and initiating initial treatment. The total score is continuously improved based on data accumulated during repeated treatments, achieving adaptive capability from incomplete data to accurate assessment.

[0023] 2. This application adopts a differentiated weighting rule based on disease severity grading. In the mild stage, the focus is on personal cognitive adjustment and social support interventions, with roughly equal weights for these three areas. In the moderate stage, drug therapy is initiated, and the weights of nutritional and exercise support and physical therapy are increased, forming a treatment pattern that emphasizes both drug and non-drug interventions. In the severe stage, the focus is on anesthesia and drug therapy, with simultaneous strengthening of social support interventions, and the combined weighting of anesthesia and drug therapy exceeds 60%. Furthermore, by incorporating psychological characteristic data such as patients' cognitive level of depression and their trust in medicine, patients are categorized into different personality types, and a corresponding weighting adjustment scheme is selected for each type. This achieves precise allocation of medical resources and personalized treatment, effectively avoiding overtreatment of mild patients and undertreatment of severe patients, and significantly improving patient treatment compliance and acceptance.

[0024] 3. This application utilizes digital monitoring indicators such as heart rate, blood pressure, electroencephalogram (EEG), and wearable device data in real time during treatment. It also constructs a side effect feature database and a real-time early warning algorithm. Employing an ultra-short-term prediction model, it identifies risk characteristics such as abnormally high EEG gamma band power and sudden drops in heart rate variability within seconds, automatically triggering audio-visual reassurance interventions or dosage adjustment suggestions. This achieves proactive prediction and intervention for acute side effects such as dissociative symptoms and blood pressure fluctuations, transforming side effect management from reactive post-treatment to proactive pre-treatment prevention, significantly improving treatment safety. Furthermore, by establishing an innovative evaluation standard encompassing seven efficacy levels—relief / cure, significant effect, effective, potentially effective, and ineffective—and combining a closed-loop iterative mechanism of phased assessment and weight updates, it dynamically calculates the weight adjustment of each treatment method based on the deviation between the current efficacy level and the target efficacy level for the next stage. This enables dynamic management of the entire disease course, from intensive treatment in the acute phase to relapse prevention in the rehabilitation phase.

[0025] 4. This application categorizes patients into four personality types—high cognition-high trust, high cognition-low trust, low cognition-high trust, and low cognition-low trust—by acquiring their cognitive level regarding depression and their trust in medicine. A corresponding weighting adjustment scheme is then selected for each type. For patients with a high cognition-high trust profile, increasing the weight of personal cognitive adjustment and social support interventions fully leverages their cognitive advantages and treatment compliance. Cognitive behavioral therapy and family support can be used to reverse the condition, reduce dependence on medication and anesthesia, and lower treatment costs and the risk of drug side effects. For patients with a high cognition-low trust profile, while maintaining the weight of necessary anesthesia to quickly establish efficacy and trust, strengthening social support interventions and personal cognitive adjustment gradually eliminates patients' distrust of medicine through objective efficacy experiences, improves treatment adherence, and avoids treatment interruptions due to patient resistance. For patients with low cognition and high trust, increasing the weight of anesthesia, medication, and social support interventions while decreasing the weight of personal cognitive adjustment can avoid poor psychotherapy outcomes due to limited cognitive abilities. This allows for the use of more direct and readily accepted medication and anesthesia, while social support compensates for the patient's lack of self-management skills. For patients with low cognition and low trust, increasing the weight of short-term anesthesia and assessing its benefit allows for rapid determination of treatment effectiveness. If no benefit is found, patients can be quickly discharged or referred, avoiding continued ineffective treatment and wasting medical resources, while also reducing the patient's unnecessary medical burden. This application, through a weighting adjustment mechanism based on patient personality type, achieves differentiated and precise treatment adaptation for patients with different psychological characteristics, significantly improving the targeting and effectiveness of treatment, reducing the risk of treatment failure due to patient cognitive biases and insufficient trust, and improving the overall treatment success rate and resource utilization efficiency. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. The elements or parts in the drawings are not necessarily drawn to scale. Obviously, the drawings described below are some embodiments of this application; for those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0027] Figure 1 This is a flowchart of an emotion disorder assessment method in one embodiment of this application.

[0028] Figure 2 This is a schematic diagram of the module structure of an emotion disorder assessment system in one embodiment of this application; Figure 3 This is a decision recommendation method in an exemplary embodiment of this application. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0030] In this document, suffixes such as “module,” “part,” or “unit” used to denote elements are used only for illustrative purposes and have no specific meaning in themselves. Therefore, “module,” “part,” or “unit” may be used interchangeably.

[0031] In this document, the terms "upper," "lower," "inner," "outer," "front," "rear," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0032] In this document, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," and "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0033] In this document, "and / or" includes any and all combinations of one or more of the listed related items.

[0034] In this article, "multiple" means two or more, that is, it includes two, three, four, five, etc.

[0035] Example 1: See Figure 3 As shown, the present invention provides a method for recommending auxiliary decision-making opinions for mood disorders, the method comprising: Collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators; The user data is standardized and scored to obtain a scoring result. The severity of the patient's disease is determined based on the scoring result, and initial resource allocation weights (hereinafter referred to as initial weights) for multiple treatment methods are set based on the severity of the disease. The patient's psychological characteristic data is acquired to update the initial resource allocation weights to obtain auxiliary decision-making opinions. The psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine. The initial resource allocation weights are corrected based on the psychological characteristic data to generate at least one personalized resource allocation weight (hereinafter referred to as personalized weight). Auxiliary decision-making opinions generated based on at least one personalized resource allocation weight are then output.

[0036] For example, in some embodiments, an initial resource allocation weight is first given for each treatment method. Subsequently, at least one of the initial resource allocation weights is modified to obtain a personalized resource allocation weight, and the remaining initial resource allocation weights are adaptively adjusted according to the personalized resource allocation weight. Finally, the multiple adjusted weights are output as auxiliary decision-making opinions.

[0037] In some embodiments, each piece of data in the user data is labeled with a privacy level and a detection difficulty level according to preset rules to obtain a preset initial collection strategy; the privacy level includes at least high privacy level data and low privacy level data, and the detection difficulty level includes at least high difficulty level data and low difficulty level data.

[0038] Preferably, different data collection strategies are selected based on the user's stage of medical treatment. Generally, the later the stage of medical treatment, the higher the privacy level or difficulty level can be.

[0039] Preferably, during the patient's first visit, based on a preset initial data collection strategy, low privacy level data and low difficulty level data are collected first.

[0040] It should be noted that the decision support method in this application is particularly applicable to existing anesthetic treatment protocols. That is to say, preferably, each treatment method includes: anesthetic treatment.

[0041] It should be noted that patients generally have a strong aversion to interventional treatments (such as anesthesia). This is especially true for patients with mood disorders, whose resistance is often intensified due to external social pressures (i.e., the generally negative perception of mental illness by the general public). Furthermore, the duration of anesthesia cannot be simply assessed based on the patient's condition (e.g., severity); if the patient's resistance is strong, anesthesia may even have the opposite effect. Therefore, this application focuses on introducing multi-dimensional intervention methods and adjusting their proportions based on changes in the patient's overall condition to reduce user resistance to anesthesia. Thus, anesthesia can be introduced or its intensity increased in real time according to changes in the patient's condition to achieve the goal of interventional treatment when necessary.

[0042] Therefore, the evaluation of treatment plans for different mood types requires consideration of more detailed dimensions, which places extremely high demands on doctors' decision-making. In response, this invention provides a decision support system that assists doctors in preliminary data analysis and processing from a computer perspective. Based on the decision support suggestions, doctors can then set up treatment plans according to their practical experience.

[0043] However, there are several challenges in the computer processing: 1) There are many data dimensions to consider, making analysis very difficult; 2) Data collection sources are wide-ranging, including objective detection indicators as well as some psychological or emotional indicators that are difficult to quantify. As a result, there may be biases in the collection process, and patients may even provide false information due to their own resistance, interfering with the computer's decision-making.

[0044] In order to reduce the initial data analysis burden on doctors and to automatically eliminate interfering factors (such as patients concealing or misreporting information) to a certain extent, this embodiment adopts a recommendation scheme for making recommendations on the proportion of treatment based on subjective and objective data.

[0045] Preferably, step S2 further includes: The patient self-reported data collected in step S1 is compared with the objective test data to determine the consistency and reliability between the patient self-reported data and the objective test data; the objective test data includes at least auxiliary examination data, digital monitoring indicators, and BMI index in basic information; if there is a conflict between the patient self-reported data and the objective test data that exceeds a preset threshold, the existence of concealment behavior is identified; when concealment behavior is identified, the initial resource allocation weight is corrected.

[0046] Furthermore, the computer can coordinate with the patient's self-reporting process based on conflicts and the data collection level. For example, when a patient is found to be concealing information, the severity of the concealment is identified (e.g., the more concealed items, the more severe the concealment; or the greater the deviation between the self-reported content and the actual content, the more severe the concealment). Based on the severity, a privacy threshold and / or a difficulty threshold are set, and correspondingly, a collection strategy is adopted to select patients whose privacy level is lower than the privacy threshold and / or whose difficulty level is lower than the difficulty threshold.

[0047] Furthermore, to prevent overly conservative data collection from delaying treatment or reducing the reliability of analysis, a threshold update mechanism is implemented. For example, the threshold update mechanism's operation process is as follows: The magnitude of change in identification weights (such as personalized weights) within a set period; When the change is less than the set value, a threshold update is triggered, such as updating the privacy threshold or updating the difficulty threshold, or increasing the privacy level or increasing the difficulty level.

[0048] For example, the magnitude of the change can be the percentage increase or decrease of the weight within a set period.

[0049] For example, when multiple weights are involved, the magnitude of the change can be the average increase or decrease of the multiple weights over a set period.

[0050] For example, one can pay particular attention to the magnitude of changes in weights. For instance, the weights to focus on could include the weight of anesthesia treatment.

[0051] For example, the weighting of considerations can also include the weighting of the patient's recommended treatment methods. Recommended treatment methods refer to those that elicit relatively low patient resistance or high cooperation during treatment; these can be set by the doctor based on feedback during the treatment process.

[0052] In this embodiment, to avoid the treatment being unable to progress due to the collection strategy remaining conservative for a long time, a threshold update mechanism will be actively activated when the weight remains unchanged for a long time.

[0053] For example, the purpose of this application is to provide dynamic interventions throughout the patient's treatment process. For instance, when the patient's symptoms are not severe and are gradually improving, the level of pharmacological intervention (such as anesthesia or oral medication) can be gradually reduced. This further alleviates the patient's emotional burden. Conversely, when the patient's symptoms are more severe and their resistance to medication is particularly strong, necessary pharmacological interventions can be appropriately increased after their condition stabilizes or shows signs of improvement.

[0054] In response, when the weighted recommendations for key treatment modalities (such as drug type interventions) remain unchanged during the process, this computer system can proactively collect and intervene in data collection to avoid the computer system being confined to erroneous perceptions due to data limitations or interference from the authenticity of the data (or, in other words, the computer being unable to perceive some patients' subtle changes due to some reasons for a long period of time), which would lead to incorrect recommendation results for doctors.

[0055] In other words, it's important to note that in typical scenarios, as treatments are gradually implemented, the patient's condition usually adjusts (e.g., gradually improves or worsens), which is reflected in the collected data. Therefore, this weight can be dynamically adjusted based on the patient's actual situation.

[0056] To address this, this embodiment adds an update mechanism for situations where weight adjustment stagnates. It is understood that the assessment of mood disorders involves multiple dimensions and complex rules, making it difficult for computers to make reliable proactive judgments without the assistance of experienced physicians. Therefore, this update mechanism also serves as an additional alert for stagnation issues, providing physicians with supplementary suggestions and guiding them to focus on the causes of the stagnation.

[0057] Specifically, in this embodiment, it is preferable to use multidimensional objective indicators plus relatively easy-to-verify subjective self-reports (such as lifestyle habits) to set the initial weights of the treatment methods; then, psychological characteristic data with a higher degree of subjectivity and greater difficulty in judgment are used as fine-tuning indicators, that is, under the initial weight system, psychological characteristic data are used to make minor adjustments to one or more of the weights.

[0058] Furthermore, to improve the reliability of highly subjective and difficult-to-distinguish psychological characteristic data, this embodiment classifies and positions the psychological characteristic data based on the patient's personal cognition (such as their depth of understanding or acceptance of emotional disorders) and trust level (such as their level of trust in the doctor, or their adherence to the treatment plan), attempting to define it using indicators that are relatively easy to quantify or determine. Thus, by synergistically combining objective indicators, easily verifiable subjective self-reports, and relatively easily quantifiable psychological indicators, step-by-step weighting is achieved, enabling the computer to integrate objective and subjective multidimensional data for autonomous assisted decision-making.

[0059] To facilitate understanding of the role of decision-making support in this invention, an exemplary application process is provided below: Figure 1 A flowchart of a mood disorder assessment method according to this application is shown, with reference to... Figure 1 The method specifically includes the following steps: S1: Collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators.

[0060] In some embodiments, basic information includes at least the patient's gender, age, BMI, exercise habits, and nutritional intake. BMI is calculated from height and weight and serves as an objective indicator to assess the patient's nutritional status and drug metabolism characteristics. Exercise habits and nutritional intake can be obtained through interviews or standardized questionnaires to assess the patient's lifestyle and its impact on the disease.

[0061] Medical history data should include at least the present illness, personal history, triggering factors, family history, and past medical history. The present illness includes the course of the illness, clinical symptoms, types of medications used, dosages, duration of use, ease of tapering off, and whether withdrawal symptoms have occurred. Personal history includes childhood environment, feeding difficulties, current negative events, and economic crises. Triggering factors include social events, changes in key interpersonal relationships, and other triggering factors. Family history includes a family history of mood disorders and related mental illnesses. Past medical history includes a history of other physical or mental illnesses.

[0062] Neuropsychiatric examination data should include assessment results in at least three dimensions: general manifestations, emotional activity, and cognitive function. General manifestations include the patient's communication status, neatness of clothing, facial expressions, and behavioral activities; emotional activity includes negative cognitive tendencies, level of self-identity, sense of belonging, and experience of security; cognitive function includes the ability to respond to language stimuli, eye contact, content of consciousness, concentration, memory, calculation ability, judgment, decision-making ability, and executive function.

[0063] The questionnaire assessment data should include at least one of the following: sleep quality assessment, anxiety and depression symptom assessment, suicide risk assessment, quality of life assessment, social support level assessment, individualized treatment assessment for mood disorders, cognitive function assessment, and treatment-related risk assessment. Specifically, sleep quality assessment uses the Insomnia Severity Index and the Pittsburgh Sleep Quality Index; anxiety and depression symptom assessment uses the Generalized Anxiety Scale, the Patient Health Questionnaire Depression Scale, the Hamilton Anxiety Scale, and the Montgomery Depression Rating Scale; suicide risk assessment uses the Beck Suicide Ideation Scale; quality of life assessment uses the European Five-Dimensional Health Scale; social support level assessment uses the Social Support Rating Scale; individualized treatment assessment for mood disorders uses the Mood Disorders Individualized Treatment Navigation Questionnaire; cognitive function assessment uses the Montreal Cognitive Assessment Scale; and treatment-related risk assessment includes fall risk assessment, drug dependence tendency assessment, drug intoxication risk assessment, and withdrawal syndrome assessment.

[0064] Ancillary examination data should include at least one of the following: organ function indicators, neuroendocrine indicators, nutritional status indicators, drug metabolism-related indicators, and imaging examination results. Organ function indicators include laboratory test results such as liver function, kidney function, and thyroid function; neuroendocrine indicators include hormone levels such as cortisol, catecholamines, melatonin, serotonin, and dopamine; nutritional status indicators include vitamin D, folic acid, homocysteine, and ferritin; drug metabolism-related indicators include drug blood concentrations and pharmacogenomic testing results, such as CYP450 enzyme phenotype; and imaging examinations include structural or functional neuroimaging assessment results such as cranial MRI, fMRI, and PET-CT.

[0065] Patient self-reported data should include at least the patient's subjective description of their symptoms, their attitude and expectations towards treatment, their level of understanding of the disease, and their level of trust in medicine. Because patients may experience stigma or resistance to treatment during their initial consultation, their self-reported data may be concealed or misrepresented. Therefore, this application will compare the patient's self-reported data with objective test data in subsequent steps to assess its reliability.

[0066] The digital monitoring indicators include at least one of the following: vital signs data, electrophysiological signal data, and behavioral activity data collected in real time during treatment. Vital signs data include heart rate, blood pressure, and blood oxygen saturation; electrophysiological signal data includes electroencephalograms (EEGs) and circadian rhythm data acquired from wearable devices such as smart bracelets and smart mattresses; behavioral activity data includes facial expressions, body movements, and limb activities captured by video or sensors during treatment. Baseline values ​​for these digital monitoring indicators are established in step S1 and continuously collected during subsequent treatment steps to monitor changes in the patient's condition and treatment effectiveness in real time.

[0067] In some embodiments, each piece of data in the user data is labeled with a privacy level and a detection difficulty level according to preset rules to obtain a preset initial collection strategy; the privacy level includes at least high privacy level data and low privacy level data, and the detection difficulty level includes at least high difficulty level data and low difficulty level data; and when the patient visits for the first time, based on the preset initial collection strategy, low privacy level data and low difficulty level data are collected first.

[0068] In one specific implementation, during the data collection process in step S1, this application further adopts a phased and tiered collection strategy to reduce patient resistance and improve data quality. Specifically, during the patient's initial visit, this application prioritizes collecting objective data with low privacy and low testing difficulty, such as height, weight, and BMI in basic information, routine laboratory tests in auxiliary examinations, and heart rate and blood pressure in digital monitoring indicators. For high privacy data, such as sensitive information like emotional activities, family relationships, and medication history, this application collects data gradually after the patient completes initial treatment and establishes trust with the doctor. For high testing difficulty data, such as costly or complex examinations like pharmacogenomics, electroencephalography, and functional imaging examinations, this application selectively collects data according to actual needs during treatment to avoid placing an excessive burden on the patient in the early stages.

[0069] S2: Standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set the initial resource allocation weights for multiple treatment methods based on the severity level of the disease.

[0070] In some embodiments, step S2 involves standardizing and scoring the user data to obtain a scoring result, and determining the severity level of the patient's disease based on the scoring result, including at least one of the following steps: Step a: Assign scores to basic information, at least classify gender, age, and BMI according to the preset level classification standard and assign corresponding scores; Step b: Assign scores to medical history data. At least the course of the disease, severity of core symptoms, accompanying symptoms, medication history, negative events, and family history indicators should be graded according to the preset grading standards and assigned corresponding scores. Step c: Neuropsychiatric examination and scoring, at least the communication status, emotional activity, and cognitive function indicators are graded according to the preset grading standard and assigned corresponding scores; Step d: Questionnaire assessment and scoring. At least sleep quality, anxiety and depression symptoms, suicide risk, social support level, and cognitive function indicators should be graded according to the preset level classification standard and assigned corresponding scores. Step e: Auxiliary examination scoring, at least organ function indicators, neuroendocrine indicators, nutritional status indicators, and drug metabolism indicators are graded according to the preset grading standard and assigned corresponding scores; Step f: Calculate the comprehensive score. Sum the scores assigned to each item in steps a to e to obtain the patient's comprehensive score, and determine the severity level of the disease based on the comprehensive score.

[0071] In one specific implementation, this application focuses on the impact of five basic information items—gender, age, BMI, exercise habits, and nutritional intake—on the treatment process of mood disorders in step S1. These indicators are all classified into three levels based on clinical evidence and assigned values ​​of 0, 1, and 2 respectively, according to their influence on treatment decisions from weakest to strongest. Specifically, for the gender indicator, males are assigned a value of 0, females a value of 1, and other genders a value of 2. For the age indicator, 18 to 45 years old is assigned a value of 0, under 18 years old a value of 1, and 45 years and older a value of 2. For the BMI indicator, 18.0 to 23.9 kg / m² is assigned a value of 0, 16 to 17.9 or 24 to 27.9 kg / m² a value of 1, and 28.0 kg / m² and above a value of 2. For exercise habits, normal exercise is assigned a value of 0, reduced exercise is assigned a value of 1, and social avoidance is assigned a value of 2. For nutritional intake, normal intake is assigned a value of 0, abnormal intake without affecting weight is assigned a value of 1, and intake leading to abnormal weight is assigned a value of 2. Based on the comprehensive evaluation of the above five indicators, this application makes the following intervention decisions: when the comprehensive value is 0, the standard dose is used; when the comprehensive value is 1, the dose is increased; when the comprehensive value is 2, the dose is decreased.

[0072] Among the five indicators mentioned above, exercise volume and nutritional intake affect BMI results. To reduce the risk of model overfitting, this application will place greater emphasis on using BMI as the basis for judgment in the actual decision-making process. The rationale for assigning values ​​to each indicator is as follows: Regarding gender, female patients generally have lower tolerance for disease symptoms, a higher urgency to solve the problem, and are more inclined to benefit from treatment as early as possible; therefore, they are given higher weights in the assignment. For patients of other genders, due to the common presence of congenital developmental abnormalities, inconsistencies between gender identity and physiological characteristics, they generally experience more severe stigma and socio-psychological pressure, and are prone to higher levels of anxiety, depression, and self-identity disorders. Therefore, in the assignment system for this group, in addition to considering physiological gender factors, it is necessary to add weight coefficients related to gender identity, social acceptance, and psychological support needs to more comprehensively reflect their treatment complexity and risk level. Regarding age, men aged 45 or older with the disease are at high risk of suicide. Regarding BMI, considering the fat accumulation effect and delayed drug release of some anesthetic drugs, to ensure the comfort and safety of patients during treatment to the greatest extent, this application will calculate the recommended dosage based on standard weight for patients with a BMI of 28.0 kg / m² or higher, where standard weight equals actual weight minus 105. Regarding the synergistic effect of indicators, although traditional anesthetic treatment protocols have considered the impact of age and BMI on drug dosage, a unified adjustment standard has not yet been established. This application emphasizes that the impact of basic patient information on dosage is not isolated and needs to be comprehensively judged in conjunction with specific indicators such as disease severity to achieve individualized dosage adjustment.

[0073] In one specific implementation, this application also focuses on the impact of the following five categories of medical history information on the treatment process of mood disorders. Present medical history includes the course of the illness, clinical symptoms, types of medications, dosages, duration of medication use, ease of tapering and discontinuation, and the occurrence of withdrawal syndrome; personal history includes childhood environment, feeding difficulties, current negative events, and economic crises; triggering factors involve social events, key interpersonal relationships, and other triggering factors; family history includes the family history of mood disorders and related mental illnesses; and past medical history includes other physical or mental illnesses. Each of the above indicators is classified into two or four levels according to its clinical impact, and assigned values ​​of 0, 1, 2, and 3 respectively, from weakest to strongest, based on their influence on treatment decisions. Specifically, the following values ​​are assigned: 0 for illness duration less than 6 months, 1 for 6-23 months, 2 for 24-59 months, and 3 for 60 months or more; 0 for no core symptoms, 1 for mild symptoms, 2 for moderate symptoms, and 3 for severe symptoms; 0 for accompanying symptoms such as hallucinations, body pain, nightmares, and insomnia, respectively assigned 0, 1, 2, and 3 for mild, moderate, and severe symptoms; 0 for no type of medication, 1 for one type, 2 for two types, and 3 for three or more types; 0 for medication duration less than 6 months, 1 for 6-12 months, 2 for 13-35 months, and 3 for 36 months or more; 0 for medication dosage below the norm, 1 for the norm, 2 for dosage above the norm, and 3 for toxicity. The value is 3; the difficulty of reducing or stopping medication is assigned as follows: easy (0), moderate (1), difficult (2), and extremely difficult (3); withdrawal syndrome is assigned as 0 (no), mild (1), moderate (2), and severe (3); childhood negative events and current negative events are assigned as 0, 1, 2, and 3 respectively (no, mild, moderate, and severe); economic crisis is assigned as 0, 1, 2, and 3 respectively (no, mild, moderate, and severe); number of related persons is assigned as 0 (no), 1 (1), 2 (2), and 3 (multiple); family history is assigned as 0 (no), 1 (present); comorbidities are assigned as 0 (no), mild (1), moderate (2), and severe (3); types of combined medications are assigned as 0 (no), 1 (1), 2 (2), and multiple (3). The comprehensive score of the above 18 indicators ranges from 0 to 52 points. The higher the score, the more severe or unstable the patient's condition, the more complex the influencing factors, and the greater the difficulty of treatment.

[0074] If the actual impact of one of the factors provided by the patient is significant, even exceeding the sum of all other factors, this application allows for an additional 10 points as a weighted adjustment to the total score. Based on the adjusted total score, this application categorizes the comprehensive impact of medical history on treatment into the following five levels: 0 to 10 points are easy, 11 to 20 points are average, 21 to 30 points are moderate, 31 to 40 points are difficult, and 41 points and above are extremely difficult. Regarding clinical decision-making and treatment adjustment recommendations, a score of 10 points or below indicates an appropriate time for medication reduction or discontinuation. Anesthesia can be used as a supportive means of comfort-enhancing medication reduction, and timely intervention when the patient experiences or anticipates significant withdrawal reactions is beneficial in maintaining the patient's confidence in medication reduction and treatment adherence. A score gradually increasing to 11 to 40 points indicates a more complex or unstable condition, and the treatment focus should gradually shift from non-pharmacological intervention to pharmacological therapy, with a corresponding increase in the frequency and intensity of anesthesia. When the score enters a stable range, especially consistently at 10 points or below, the medication reduction / discontinuation management process can be restarted, gradually achieving the clinical cure goal through a closed-loop model of assessment-treatment-reassessment. It is worth noting that the risk of relapse of depression is associated with a variety of factors, including a history of recurrent episodes, residual symptoms, a family history of mood disorders, a history of childhood abuse, and comorbidities. Among these, residual symptoms have the most significant impact. Therefore, setting the treatment goal as a cure and complete elimination of residual symptoms is of great significance in reducing the risk of relapse.

[0075] In one specific implementation, this application also focuses on the impact of the following three key dimensions in neuropsychiatric examinations on the treatment process of mood disorders. General manifestations include communication skills, neatness of attire, facial expressions, and behavioral activities; emotional activities encompass negative cognitive tendencies, self-identity levels, sense of belonging, and feelings of security; cognitive functions involve responsiveness to verbal stimuli, eye contact, content of consciousness, concentration, memory, calculation ability, judgment, decision-making ability, and executive function. Each of the above indicators is classified into secondary or tertiary categories based on its clinical significance and severity, and assigned values ​​of 0, 1, and 2 respectively, from least to most severe, according to their impact on treatment decisions. Specifically, normal communication is assigned a value of 0, decreased or increased communication is assigned a value of 1, and no communication or excessive talking is assigned a value of 2; neat attire is assigned a value of 0, and untidy attire is assigned a value of 1; normal facial expression is assigned a value of 0, indifference is assigned a value of 1, and unusually rich facial expression is assigned a value of 2; free movement is assigned a value of 0, slow movement is assigned a value of 1, and dependence on external forces is assigned a value of 2; negative cognition is assigned a value of 0 if absent and 1 if present; normal self-identity is assigned a value of 0, weak self-identity is assigned a value of 1, and excessive self-identity is assigned a value of 2; normal sense of belonging is assigned a value of 0, weak self-identity is assigned a value of 1, and none is assigned a value of 2; normal sense of security is assigned a value of 0, weak self-identity is assigned a value of 1, and none is assigned a value of 2. The scores for the 17 indicators are as follows: 0 for normal response to verbal stimuli, 1 for slow response, and 2 for no response; 0 for normal eye contact and 1 for avoidance; 0 for normal content of consciousness and 1 for abnormal content; 0 for normal attention, 1 for mild impairment, and 2 for significant impairment; 0 for normal memory, 1 for mild impairment, and 2 for significant impairment; 0 for normal calculation ability and 1 for abnormal calculation ability; 0 for normal judgment and 1 for abnormal judgment ability; 0 for normal decision-making ability and 1 for abnormal decision-making ability; and 0 for normal executive function and 1 for abnormal executive function. The maximum score for these 17 indicators is 26 points. A higher score indicates a more severe or unstable condition, more complex influencing factors, and greater treatment difficulty.

[0076] In the indicator system, dimensions such as negative cognition, self-identity, sense of belonging, and sense of security may have potential correlations with other scoring items. Care should be taken to avoid overfitting caused by these correlations during modeling. Based on the total score, this application classifies the comprehensive impact of neuropsychiatric examinations on treatment into the following four levels: 0 to 6 points are average, 7 to 12 points are moderate, 13 to 18 points are difficult, and 19 points and above are extremely difficult. Regarding clinical decision-making and treatment adjustment recommendations, if the score is 6 or below, and the score mainly comes from indicators reflecting psychosocial function and subjective experience such as negative cognition, self-identity, sense of belonging, sense of security, communication, facial expressions, eye contact, and concentration, it indicates that the condition is still under control. In this case, medication reduction and discontinuation can be gradually implemented while strengthening psychosocial support. If the score comes from indicators reflecting cognitive impairment, content of consciousness, calculation ability, judgment, and executive function, indicating neurological or organic impact, it indicates that the disease has had a substantial impact on central nervous system function. In this case, anesthesia can be used as an adjunct to comfort medication adjustments, intervening promptly when the patient experiences or anticipates significant withdrawal reactions to maintain treatment confidence and compliance. When the score gradually rises to 7 or above, it indicates that the condition is becoming more complex or unstable. The focus of treatment should gradually shift from non-pharmacological intervention to pharmacological treatment, with a corresponding increase in the frequency and intensity of anesthesia. After the score stabilizes, the structured medication reduction and discontinuation management process can be restarted. Through a closed-loop model of assessment-intervention-reassessment, the recovery of the condition can be gradually promoted, ultimately achieving the goal of clinical cure.

[0077] In one specific implementation, this application integrates multiple standardized scales and assessment tools to systematically collect quantitative data on patient symptoms, functional status, and treatment risks, focusing on the following assessment contents: sleep quality and severity, including the Insomnia Severity Index and the Pittsburgh Sleep Quality Index; anxiety and depressive symptoms, including the Generalized Anxiety Scale, the Patient Health Questionnaire Depression Scale, the Hamilton Anxiety Scale, and the Montgomery Depression Rating Scale; suicide risk, using the Beck Suicide Ideation Scale; quality of life, using the European Five-Dimensional Health Scale; social support level, using the Social Support Rating Scale; individualized treatment of mood disorders, using the Mood Disorders Individualized Treatment Navigation Questionnaire; cognitive function, using the Montreal Cognitive Assessment Scale; and treatment-related risks, including fall risk, drug dependence tendency, drug poisoning risk, and withdrawal syndrome occurrence. The above indicators are classified into two or four levels according to their clinical severity, and assigned values ​​of 0, 1, 2, and 3 respectively, from low to high, based on their influence on treatment decisions. The above 15 indicators have a maximum score of 41 points. The higher the score, the more severe or unstable the patient's condition, and the greater the difficulty of treatment. Since the assessment dimensions such as sleep, anxiety, depression, and social support often overlap and interact in clinical manifestations and mechanisms, the indicator system needs to pay attention to potential multicollinearity problems during the modeling process, and use methods such as regularization to reduce the risk of overfitting, so as to ensure the generalization ability and clinical applicability of the model.

[0078] Based on the comprehensive scoring results, this application categorizes the overall impact of questionnaires and tool assessments on treatment into the following four levels and provides corresponding decision-making recommendations. A total score of 0 to 10 indicates mild symptoms, with treatment primarily focused on psychosocial intervention and health education. This application recommends low-frequency monitoring and basic support programs. A total score of 11 to 20 indicates moderate symptoms significantly impacting function, requiring a combination of pharmacological and non-pharmacological interventions. This application recommends developing a structured treatment plan and strengthening follow-up. A total score of 21 to 30 indicates severe symptoms with multidimensional impairment, requiring priority for intensive pharmacological treatment and comprehensive psychological intervention. This application initiates high-frequency assessments and a multidisciplinary collaborative mechanism. A total score of 31 to 41 indicates extremely severe symptoms, with the patient in a high-risk state requiring emergency medical intervention and intensive management throughout the treatment process. This application prioritizes safety and recommends inpatient or intensive outpatient treatment.

[0079] In one specific implementation, this application also integrates various laboratory and imaging-based auxiliary examination data to assess the patient's physiological state, drug metabolism capacity, and potential organic causes, providing an objective basis for individualized treatment. The focus is on the following examination items: organ function and metabolic indicators, including liver function, kidney function, and thyroid function; neuroendocrine and sleep-related hormones, covering cortisol, catecholamines, melatonin, serotonin, and dopamine; nutritional status indicators, such as vitamin D, folic acid, homocysteine, and ferritin; therapeutic drug monitoring indicators, involving drug blood concentration and pharmacogenomic testing such as CYP450 enzyme phenotype; and imaging examinations, including structural or functional neuroimaging assessments such as cranial MRI, fMRI, and PET-CT. These indicators are classified into secondary or tertiary categories based on their degree of abnormality and clinical significance, and assigned values ​​of 0, 1, and 2 respectively, according to their impact on treatment decisions and risk management, from mild to severe. This application uses quantitative scoring for 11 key auxiliary examination indicators, with a maximum total score of 21 points. Treatment decisions are categorized into three levels based on the total score: Level 1 is low risk, with a total score of 0 to 7. Examination results are generally normal or only mildly abnormal. Treatment follows the standard protocol without significant adjustments to the examination results. The standard treatment plan for mood disorders is implemented, and key indicators are rechecked every 3 to 6 months. Level 2 is medium risk, with a total score of 8 to 15. Multiple moderately abnormal indicators are present. Targeted adjustments to medications, dosages, or treatment pathways are required, along with enhanced monitoring. Individualized drug and non-drug adjustments are made on top of standard treatment. Key indicators are rechecked every 1 to 3 months. Level 3 is high risk, with a total score of 16 to 21. Multiple indicators are significantly abnormal, indicating physiological disorders, organic problems, or medication safety risks. Medical conditions must be addressed first, and medications should be used to minimize the burden on organs. Treatment requires a high degree of individualization and multidisciplinary collaboration. Key indicators are rechecked every 2 to 4 weeks.

[0080] In one specific implementation, this application achieves real-time dynamic monitoring of the patient's condition and treatment response by continuously collecting vital signs, electrophysiological signals, and behavioral activity data during treatment. It focuses on three categories of digital indicators: vital sign monitoring, including heart rate, blood pressure, and blood oxygen saturation; electrophysiological signal monitoring, encompassing electroencephalography (EEG) and physiological rhythm data acquired from wearable devices such as smart bracelets and smart mattresses; and behavioral activity monitoring during treatment, involving behavioral characteristics captured by video or sensors, such as facial expressions, body movements, and limb activities. These indicators are classified into secondary or tertiary categories based on their degree of deviation from the normal range or the frequency of abnormal patterns, and assigned values ​​of 0, 1, and 2 respectively, according to their impact on treatment safety and response assessment, from least to most severe. This application uses a weighted total score of 9 digital monitoring indicators, with a maximum score of 18 points. A higher score reflects a more unstable real-time patient condition, correspondingly increasing the system risk warning level, and shifting clinical decision-making from maintenance and optimization to active intervention and risk management. A total score of 0 to 5 indicates stability, normality, or slight fluctuation. The current treatment plan should be maintained, with gradual optimization of non-pharmacological interventions such as behavioral activation and sleep rhythm adjustment. A systematic assessment should be conducted every 4 weeks, focusing on improving social function and quality of life. If the score remains at 3 or below for more than 8 weeks, a medication tapering procedure can be initiated, along with enhanced psychoeducation and self-management support. A total score of 6 to 12 indicates fluctuation, moderate deviation, and a worsening trend. Monitoring frequency should be increased to once every 2 weeks, paying attention to indicator trends and related triggers. The treatment plan should be adjusted, and psychological interventions such as cognitive restructuring and stress management should be strengthened. Medication dosage should be fine-tuned if necessary, and short-term goal setting and behavioral contracts should be introduced to enhance treatment synergy and patient participation. A total score of 13 to 18 indicates high risk, significant disturbance, and increased risk. Intensive assessment and intervention should be initiated immediately. At least weekly clinical follow-up or remote monitoring is recommended. Prioritize medication optimization such as changing medications, combination therapy, or dosage adjustment, and consider combining physical therapy or emergency psychological support. If there is a risk of suicide or severe self-neglect, multidisciplinary team intervention is necessary, and hospitalization is recommended if necessary.

[0081] In some embodiments, in step S2, the aforementioned scores are summed to obtain the patient's comprehensive score. Specifically, the scores for basic information, medical history, neuropsychiatric examination, questionnaire assessment, auxiliary examinations, and digital monitoring indicators are accumulated to obtain the patient's comprehensive score. The severity of the patient's condition is then determined based on a preset first scoring threshold range. The severity of the disease is divided into mild, moderate, and severe (severe is further subdivided into two subtypes: without psychotic symptoms and with psychotic symptoms). The comprehensive score fully reflects the severity, complexity, and diversity of influencing factors of the patient's disease. A higher score indicates a more severe or unstable condition, more complex influencing factors, and greater difficulty in treatment.

[0082] In some embodiments, when scoring data for all sub-items is unavailable, preset core items are used for scoring to determine the severity level of the patient's disease. Data related to the scoring is then acquired during treatment to correct the severity level. In some specific embodiments, in actual clinical applications, some sub-items may not be fully available. For example, patients may not complete all questionnaires or auxiliary examinations during their initial visit due to resistance or limitations; or in certain emergency situations, treatment needs to be initiated immediately, leaving insufficient time for complete data collection. To address these situations, this application adopts a scoring strategy prioritizing core items. Specifically, when scoring data for all sub-items is unavailable, the system prioritizes collecting and calculating the core items that contribute most to the assessment of disease severity. These include, but are not limited to, the severity of core symptoms in medical history data, suicide risk in questionnaire assessment, key medical history information such as disease course and medication history in medical history data, as well as age and BMI indicators in basic information. Based on the scoring results of the above core items, and according to the preset second scoring threshold range combination, the system determines the severity of the patient's condition and initiates initial treatment accordingly. The system can still determine the severity of the patient's condition and initiate initial treatment accordingly, ensuring that the patient can receive necessary medical intervention as soon as possible and avoiding delays in treatment due to incomplete data collection.

[0083] As treatment progresses and patients visit the hospital repeatedly, the system continuously accumulates new user data. New data generated during each treatment session, including updated scale scores, auxiliary examination results, digital monitoring indicators, and treatment response records, are automatically collected and incorporated into the recalculation of the comprehensive score. Through this dynamic accumulation and iterative update mechanism, the system can continuously improve the accuracy of the total score calculation, making the comprehensive score increasingly closer to the patient's true condition. Simultaneously, based on the updated comprehensive score, the system makes corresponding fine-tuning adjustments to resource allocation weights and treatment plans. As the amount of data accumulates, the system's understanding of the patient's condition deepens, and the accuracy of treatment plans gradually improves, thereby providing patients with increasingly personalized and precise treatment adjustments.

[0084] In some embodiments, the multiple treatment methods in step S2 include at least one of anesthesia, pharmacology, personal cognitive adjustment, nutritional and exercise support, social support intervention, and physical therapy.

[0085] Furthermore, the disease severity grading in step S2 includes mild, moderate, and severe stages.

[0086] And / or, the configuration rule for the initial resource configuration weights is: In the mild stage, interventions primarily focus on individual cognitive adjustment and social support. Initiate drug therapy during the moderate stage, and increase the weight of nutritional and exercise support and physical therapy; In the severe stage, the main treatments are anesthesia and medication, while social support interventions are strengthened simultaneously.

[0087] In one specific embodiment, this application further proposes a drug-cognitive-social composite intervention model in step S2. In traditional treatment pathways, it is generally recommended to continue the acute-phase medication dosage during maintenance therapy to prevent relapse, and to introduce psychotherapy after acute-phase symptom control to further reduce the risk of relapse. However, in actual clinical practice, most large general hospitals or psychiatric hospitals in my country have not yet established effective referral mechanisms and collaborative channels with professional psychological counseling service institutions. After about four weeks of treatment and initial stabilization of core symptoms, patients often only receive relatively general cognitive-behavioral guidance or general social support, and are difficult to be systematically referred to qualified psychological institutions for continuous, structured intervention. This results in a lack of true sequential integration of social support and psychotherapy. This disconnect directly leads to difficulty in maintaining stable treatment effects, and patients are prone to symptom recurrence when facing external stress or comorbid interference, thus affecting long-term treatment adherence.

[0088] Furthermore, many current psychological counseling services focus on guiding patients to analyze the triggers of their illness, particularly the impact of past experiences such as social events. However, repeatedly recalling and recounting painful experiences can disrupt already stable emotions, hindering recovery and potentially exacerbating psychological burdens. For patients with severe psychological trauma, traumatic memories are often difficult to eradicate completely through a limited number of counseling sessions; some patients may even resort to sealing away memories or avoiding mentioning them, effectively delaying the natural integration and repair process of trauma. In this process, timely integration with medication helps regulate emotional responses and improve cognitive processing abilities, thereby promoting adaptive integration of traumatic experiences, rather than relying solely on verbal analysis. Existing psychological service models generally provide insufficient support for populations with limited cognitive function, such as younger, less educated, and elderly patients. These individuals often struggle to effectively manage psychological problems through self-reflection and verbal expression, requiring specific, behavior-oriented, and practical problem-solving strategies, which current services still lack adequate provision.

[0089] To achieve a systematic integration of multiple intervention measures, including drug therapy, cognitive adjustment, social support, nutrition and exercise, and physical therapy, and to promote the overall recovery of patients more efficiently and comprehensively, this application dynamically allocates weights to each treatment component based on the patient's condition characteristics and treatment needs at different stages. Specifically, the innovative anesthetic treatment precision adaptation strategy for mood disorders proposed in this application assigns differentiated initial resource allocation weights to multiple treatment methods based on the disease severity classification determined in step S2. The disease severity classification in step S2 includes mild, moderate, and severe stages. The allocation rules for the initial resource allocation weights are as follows: in the mild stage, personal cognitive adjustment and social support interventions are the main focus; in the moderate stage, drug therapy is initiated, and the weights of nutritional and exercise support and physical therapy are increased; in the severe stage, anesthetic treatment and drug therapy are the main focus, while social support interventions are simultaneously strengthened.

[0090] The weighting of each component in this application is primarily based on the following three aspects. First, regarding the weighting of drug therapy, the clinical remission rate of antidepressants is approximately 30% to 50%, and patients with a strong aversion to medical care generally have a strong aversion to medication and concerns about its use, a tendency significantly higher than the other four treatment components. Furthermore, these medications require prescriptions from regular medical institutions, and hospital medical record systems are usually linked to medical insurance data. Related disease diagnoses may raise concerns among patients about privacy breaches, social discrimination, access to commercial insurance, and even career development. For mild to moderate patients, the organic integration of the other treatment components may lead to clinical cure; however, for severe patients, drug therapy remains indispensable. In this case, while ensuring efficacy, patient acceptance and medication safety must be fully considered. Therefore, its weight needs to be dynamically adjusted according to the severity of the disease.

[0091] Secondly, regarding the weighting of anesthesia, for severe depressive disorder with high self-harm or suicide risk, traditional treatment primarily involves modified electroconvulsive therapy (ECT) or hospitalization. However, ECT can cause side effects such as short-term memory impairment. In this protocol, ECT can be replaced by anesthesia. Through precise medication, the suicide risk can typically be significantly reduced after one to three treatments within two weeks. Based on this, combined with social support and personal cognitive correction, patients can quickly overcome the critical period and enter the consolidation phase. Specific adjustment rules for anesthesia can be found in the aforementioned core symptom-driven anesthesia decision matrix. Due to significant individual variations in the single dose of anesthetic drugs, the specific dosage needs to be dynamically adjusted based on the various indicators in the disease severity grading and efficacy evaluation criteria, with the optimization goal of maximizing efficacy and minimizing side effects. Key indicators such as age, BMI, heart rate, blood pressure, and disease severity have relatively high weights. Third, regarding the weighting of personal cognitive adjustment, nutritional and exercise support, and social support interventions, the above three types of interventions run through the entire treatment process and have a synergistic effect at each stage. However, their weights change dynamically with the severity of the disease, specifically following the initial resource allocation weighting rules defined in step S2.

[0092] In one specific embodiment, in step S2, this application further formulates differentiated phased treatment strategies and initial resource allocation weighting rules based on disease severity classification. Specifically, this application divides disease severity into mild, moderate, and severe stages, and configures differentiated treatment method weight combinations for each stage to achieve a balance between optimal efficacy and minimum patient burden.

[0093] In the mild stage, patients' attention, memory, and executive function remain largely intact, and social functioning is not significantly affected. Over 50% of patients and their families are unwilling to be diagnosed with depression, and over 70% of families choose non-pharmacological interventions. Since nutritional and exercise status are usually not significantly abnormal, and laboratory test results are also normal, the weight of nutritional and exercise interventions is relatively low. At this stage, interventions focus on personal cognitive adjustment and social support, with roughly equal weights for personal cognitive adjustment, social support, and physical therapy. This is because patients at this stage generally have milder cognitive biases and more intact social relationships, and the condition can often be reversed through family support and interpersonal psychotherapy. If the above interventions are ineffective, and the patient is unwilling or unable to accept drug treatment, non-invasive physical therapy methods with fewer side effects, such as sound, light, electrical, and magnetic stimulation, can be prioritized to improve treatment acceptance and compliance. For example, in terms of specific weighting, in the mild stage, personal cognitive adjustment accounts for 30%, social support intervention for 30%, physical therapy for 25%, nutritional and exercise support for 10%, anesthesia for 5%, and drug treatment for 0%. The above configuration reflects a treatment strategy centered on non-pharmacological interventions in the mild stage. Personal cognitive adjustment and social support interventions account for 60% of the total, physical therapy accounts for 25% as a supplementary means, and nutritional and exercise support and anesthesia are only backup options.

[0094] In the moderate stage, medication should be initiated, with increased emphasis on nutritional and exercise support and physical therapy. As core emotional and physical symptoms worsen, they can lead to insufficient nutrient intake and reduced physical activity, creating a vicious cycle of symptom exacerbation and functional decline. At this point, a structured nutritional support and individualized exercise plan should be developed based on physical examination and laboratory test results, along with simultaneous intervention to strengthen individual cognitive adjustment. Although patients at this stage often have a willingness to seek help and tend to seek psychological counseling first, their concentration and executive function are significantly affected, making it difficult to achieve a fundamental shift in cognitive and behavioral patterns solely through their own efforts. Therefore, necessary external guidance and support are required in treatment. With the emergence of self-harming thoughts or behaviors, family members often turn to professional medical help due to their inability to cope. In terms of treatment strategies, if oral medication provides a clear response, anesthesia at this stage is only used as a reinforcing or auxiliary means to quickly stabilize emotions, improve some physical symptoms, and enhance treatment compliance. Furthermore, systematic social support intervention remains indispensable, and physical therapy can be used continuously as a complementary approach. In particular, for children and adolescents with moderate to severe depression who do not respond well to psychotherapy, a family-centered systemic treatment should be initiated promptly. This model integrates parents or primary caregivers into the treatment system, promoting overall patient recovery by adjusting family interaction patterns and enhancing family support functions. Furthermore, specific evidence-based psychotherapies can be used according to the characteristics of different age groups: for children and adolescents, dialectical behavior therapy has been proven to effectively reduce non-suicidal self-harm and alleviate depressive symptoms; for adult patients, the mental chemotherapy approach can significantly reduce the recurrence frequency and severity of non-suicidal self-harm. For example, in terms of specific weighting, in the moderate stage, medication accounts for 25%, social support intervention for 20%, personal cognitive adjustment for 20%, nutritional and exercise support for 15%, physical therapy for 12%, and anesthesia for 8%. This allocation reflects a treatment strategy that emphasizes both medication and non-medication interventions in the moderate stage, with medication accounting for 25%, social support intervention and personal cognitive adjustment accounting for 40% combined, nutritional and exercise support and physical therapy accounting for 27% combined, and anesthesia used only as a reinforcing or adjunctive measure for 8%. For children and adolescents who do not respond well to psychotherapy, the weight of family therapy in social support interventions can be appropriately increased to 25% to 30%, while the weight of personal cognitive adjustment can be reduced accordingly.

[0095] In the severe stage, treatment primarily involves anesthesia and medication, while simultaneously strengthening social support interventions. Emotional and physical symptoms deteriorate significantly, nutritional and motor status is severely abnormal, manifesting as emaciation or obesity, minimal activity, severely impaired cognitive function, and near-complete loss of social function. At this stage, saving life and improving mood should be the primary goals. Since patients often struggle to cooperate with systematic nutritional and exercise interventions, only gradual, small adjustments can be made. Social relationships are often broken down, making cognitive correction and social support essential needs. These interventions should be gradually strengthened by professionals after the patient's condition has stabilized and their trust has been gained. Because most physical therapies have limited accessibility at this stage, and patients' trust is reduced due to previous ineffective treatments, the weight of physical therapy decreases significantly. In terms of specific weighting, in the severe stage, anesthesia accounts for 35%, medication for 30%, social support intervention for 25%, personal cognitive adjustment for 5%, nutritional and exercise support for 3%, and physical therapy for 2%. The above configuration reflects a treatment strategy prioritizing rapid stabilization of core symptoms in the severe stage. Anesthesia and medication account for 65% of the treatment, ensuring that suicide risk and emotional stability are controlled in the shortest possible time. Social support intervention accounts for 25% as a rigid requirement to compensate for the lack of support caused by the patient's impaired social functioning. Personal cognitive adjustment accounts for only 5%, and needs to be gradually strengthened after the patient's condition is relatively stable and trust is gained. Nutritional and exercise support accounts for only 3%, with only minor adjustments to avoid placing an additional burden on the patient. Physical therapy accounts for only 2%, due to its limited accessibility and reduced patient trust.

[0096] In some embodiments, step S2 further includes: The patient self-reported data collected in step S1 is compared with the objective test data to determine the consistency and reliability between the patient self-reported data and the objective test data; the objective test data includes at least auxiliary examination data, digital monitoring indicators, and the BMI index in the basic information. If there is a conflict between the patient's self-reported data and the objective test data that exceeds a preset threshold, then concealment behavior is identified. When underreporting is detected, the initial resource allocation weights are adjusted, including: Increase the initial weight of social support interventions and personal cognitive adjustment in the initial resource allocation weights.

[0097] In one specific implementation, this application further includes a data consistency verification mechanism in step S2. Specifically, this application compares the patient-reported data collected in step S1 with the objective test data to determine the consistency and reliability between the patient-reported data and the objective test data. The objective test data includes at least auxiliary examination data, digital monitoring indicators, and the BMI index from basic information. Among them, auxiliary examination data such as blood biochemical indicators and hormone level test results have high objectivity and accuracy; digital monitoring indicators such as heart rate, blood pressure, blood oxygen saturation, and electroencephalogram characteristics are automatically collected by the device and are not affected by the patient's subjective will; the BMI index is obtained by directly measuring height and weight and is also an objective and verifiable indicator.

[0098] By cross-referencing patient-reported data with the aforementioned objective test data, this application can effectively identify potential concealment or misrepresentation by patients. For example, patient-reported exercise habits and nutritional intake can be verified using BMI. If a patient claims sufficient exercise and regular eating habits, but their BMI indicates they are obese or underweight, there is a clear conflict between the two. Similarly, if a patient reports good medication adherence, but blood drug concentration tests show concentrations below the effective therapeutic range, it suggests the patient may have reduced or missed doses without authorization. Furthermore, if a patient reports acceptable sleep quality, but digital monitoring indicators show abnormal sleep structure, frequent awakenings, and insufficient total sleep time, it indicates a distorted perception or concealment of their sleep status.

[0099] When a conflict exceeding a preset threshold exists between the patient's self-reported data and the objective test data, this application identifies concealment behavior. The preset threshold can be set according to the specific indicator type and clinical significance, such as a difference exceeding one standard deviation between BMI and self-reported exercise volume, or a difference exceeding a preset percentage range between blood drug concentration and self-reported medication dosage. Once concealment behavior is identified, this application will adjust the initial resource allocation weights, for example, by increasing the initial weights of social support intervention and personal cognitive adjustment in the initial resource allocation weights. Through the above weight adjustment, this application can ensure the effectiveness and specificity of the treatment plan even when the patient conceals or misrepresents information, avoiding treatment deviations caused by information distortion.

[0100] In some embodiments, clinical practice has observed that improvements in symptoms such as nightmares, hallucinations, and bodily pain typically lag behind core emotional symptoms. These residual symptoms not only may be potential factors for relapse or recurrence but are also frequently overlooked during routine diagnosis and treatment. Therefore, based on the severity grading criteria for depressive episodes outlined in the "Chinese Guidelines for the Prevention and Treatment of Depressive Disorders (2025 Edition)," this application further proposes refined management strategies, emphasizing the significant role of other symptoms or additional symptoms such as nightmares, hallucinations, and bodily pain in disease progression, to promote more targeted specialized assessments and interventions in clinical practice.

[0101] Specifically, this application can classify disease severity into mild, moderate, and severe (with severe further subdivided into two subtypes: without psychotic symptoms and with psychotic symptoms) based on symptomatological characteristics, disease course, and degree of functional impairment. For example, in the mild category, the patient meets at least two core diagnostic criteria and at least two other symptoms, with or without bodily pain, and with or without nightmares, all of which persist for at least two weeks, and the patient has some difficulty continuing daily work and social activities. In the moderate category, the patient meets at least two core diagnostic criteria and at least three other symptoms, including mild hallucinations, moderate bodily pain, and nightmares, all of which persist for at least two weeks, and the patient has considerable difficulty performing work, social, or household activities. In the severe nonpsychotic symptom classification, the patient meets all three core diagnostic criteria and at least four other symptoms, including moderate to severe bodily pain and frequent nightmares. These symptoms must last for at least two weeks, but may be less than two weeks if the symptoms are exceptionally severe and have a rapid onset. The patient is unlikely to continue social, work, or household activities. In the severe psychotic symptom classification, the patient meets all three core diagnostic criteria and at least four other symptoms, including hallucinations, delusions, or stupor, moderate to severe bodily pain, and frequent nightmares. These symptoms must last for at least two weeks, but may be less than two weeks if the symptoms are exceptionally severe and have a rapid onset. The patient is unlikely to continue social, work, or household activities. The core diagnostic criteria are based on internationally accepted diagnostic criteria for depression and include items such as depressed mood, loss of interest or pleasure, decreased energy or fatigue. Other symptoms include, but are not limited to, decreased attention, indecisiveness, feelings of guilt or worthlessness, sleep disturbances, significant changes in appetite or weight, and suicidal ideation or behavior.

[0102] This application incorporates nightmares into the disease severity assessment system based on the following: Nightmares, as a clinically significant sleep disorder, are not only an important manifestation of serious mental and brain diseases but also exhibit a high epidemiological burden. Studies have shown that the incidence of nightmares can be as high as 90% in depressed patients with suicidal ideation, and multiple studies have confirmed that nightmares can serve as an important clinical predictor of suicide risk. However, due to the relatively lagging development of sleep medicine and psychology in China, the time-consuming process of dream analysis requiring a professional psychological background, coupled with the current lack of effective treatments for nightmares, there is a general lack of attention paid to dream-related issues in neuropsychiatric clinics.

[0103] The basis for including hallucinations in this application's assessment system is that frequent occurrences or increasing severity of hallucinations often indicate that the disease is entering an acute phase or is trending towards deterioration. Their content characteristics, such as command hallucinations, can be directly associated with high-risk behaviors such as suicide and aggression, requiring focused assessment and intervention. Simultaneously, hallucinations can interfere with patients' cognitive processes, emotional stability, and behavioral control, thereby leading to impairments in real-world functioning such as social avoidance and reduced occupational function. Furthermore, the severity, frequency, and risk level of hallucination content can provide objective evidence for the selection and adjustment of clinical treatment plans.

[0104] The basis for incorporating somatic pain into the assessment system in this application is that somatic pain is a highly comorbid symptom of depression, anxiety disorders, somatic symptom disorder, and post-traumatic stress disorder. Pain experience often exacerbates mood disorders, forming a vicious cycle of "pain-emotion," becoming an important marker of disease complexity and chronicity. Chronic pain involves multiple pathological mechanisms, including neurotransmitter imbalance, central sensitization, and inflammatory responses, and shares some neurobiological pathways with mental illnesses. Persistent pain severely affects patients' daily activities, social participation, and occupational function, exacerbating stigma and decreased self-efficacy. Furthermore, novel anesthetic drugs, such as esketamine, have shown unique therapeutic value in the management of refractory somatic pain, further highlighting the clinical necessity of systematically addressing somatic pain in psychiatric assessments. Through the aforementioned hierarchical management model, this application can more accurately quantify the severity of the patient's disease in step S2, thereby providing a more scientific basis for setting the initial resource allocation weights.

[0105] In some embodiments, for anesthesia treatment, this application constructs a Symptom-Based Anesthesia Decision Matrix (SAM) to guide differentiated ketamine dosing regimens during the first month of treatment for treatment-resistant depression. This decision matrix determines the initial dose and first-month treatment frequency of ketamine based on two dimensions: the patient's core symptom type and disease severity. In the core symptom type dimension, it distinguishes between two scenarios: "primarily anxiety, nightmares, and physical pain" and "primarily depression, self-harm, and suicidal ideation." In the disease severity dimension, it distinguishes between moderate and severe levels. Furthermore, it considers the presence of hallucinations to refine the dosing regimen.

[0106] For patients presenting primarily with anxiety, nightmares, and bodily pain without hallucinations, the dosage of ketamine is 0.30 mg / kg body weight for moderate severity, administered once per month; for severe severity, the dosage is 0.40 mg / kg body weight, administered twice per month. For patients presenting primarily with anxiety, nightmares, and bodily pain accompanied by hallucinations, the dosage is 0.35 mg / kg body weight for moderate severity, administered twice per month; for severe severity, the dosage is 0.40 mg / kg body weight, administered twice per month.

[0107] For patients presenting primarily with depression, self-harm, or suicidal ideation without hallucinations, the dosage of ketamine is 0.40 mg / kg body weight for moderate severity, with two treatments in the first month; for severe severity, the dosage is 0.50 mg / kg body weight, with four treatments in the first month. For patients presenting primarily with depression, self-harm, or suicidal ideation accompanied by hallucinations, the dosage is 0.40 to 0.50 mg / kg body weight for moderate severity, with two treatments in the first month; for severe severity, the dosage is 0.50 to 0.70 mg / kg body weight, with four to eight treatments in the first month.

[0108] S3: Obtain the patient's psychological characteristic data, which includes at least the patient's cognitive level of depression and trust in medicine. Adjust the initial resource allocation weights based on the psychological characteristic data to generate personalized resource allocation weights, and generate a personalized treatment plan based on the personalized resource allocation weights.

[0109] In some embodiments, step S3 further includes: Based on patients' level of understanding of depression and their level of trust in medicine, patients are classified into different personality types; these personality types include high cognition-high trust type, high cognition-low trust type, low cognition-high trust type, and low cognition-low trust type. Based on the individual type, a corresponding weight adjustment scheme is selected to adjust the initial resource configuration weights to generate personalized resource configuration weights, including: If the personality type is high cognition-high trust, then increase the weight of personal cognitive adjustment and social support intervention; If the personality type is high cognition-low trust, then while maintaining the weight of necessary anesthetic treatment, the weight of social support intervention and personal cognitive adjustment should be increased. If the personality type is low cognition-high trust, then increase the weight of anesthesia, medication and social support interventions, and decrease the weight of personal cognitive adjustment. If the personality type is low cognitive-low trust, increase the weight of anesthesia treatment in the short term and assess whether there is any benefit. If there is no benefit, quickly remove the patient from the group or refer them.

[0110] In one specific implementation, step S3 of this application further includes a weighted adjustment mechanism based on patient personality type. Specifically, this application categorizes patients into different personality types based on two dimensions: the patient's level of understanding of depression and their level of trust in medicine. The level of understanding of depression refers to the patient's comprehension of the causes, clinical manifestations, treatment methods, and prognosis of depression; the level of trust in medicine refers to the patient's trust in the medical system, doctors' advice, and medication. Based on different combinations of these two dimensions, this application categorizes patients into four personality types: high cognition-high trust, high cognition-low trust, low cognition-high trust, and low cognition-low trust.

[0111] Patients with high cognitive-trust levels have a relatively scientific understanding of depression, comprehending the nature of the illness and the necessity of treatment, and exhibiting high trust and adherence to medical advice. For these patients, this application further increases the weighting of personal cognitive adjustment and social support interventions on top of the initial resource allocation weights. This is because these patients possess strong self-reflection abilities and a willingness to learn, and through personal cognitive adjustment methods such as cognitive behavioral therapy and interpersonal psychotherapy, they can effectively identify and change negative thought patterns. Simultaneously, they have a high acceptance of social support, and social support interventions such as family therapy and peer support can play a good synergistic role. The weighting of drug therapy and anesthesia can be appropriately reduced in these patients to minimize unnecessary medical resource input and drug exposure risks.

[0112] Patients with high cognitive-low trust profiles have a high level of understanding of depression, but due to past treatment experiences, concerns about medication side effects, or distrust of the healthcare system, they have a low level of trust in medical advice. For these patients, this application maintains the weight of necessary anesthesia treatment while strengthening the weight of social support interventions and personal cognitive adjustment. This is because, although these patients have a high level of understanding, their low trust may lead them to refuse or resist medication. Therefore, it is necessary to strengthen social support interventions to build trust between doctor and patient, while personal cognitive adjustment helps patients identify and overcome sources of distrust in medicine. Anesthesia treatment needs to maintain a necessary weight in this group to quickly improve core symptoms, build confidence in treatment, and thus gradually increase the patient's trust in medicine.

[0113] Patients with low cognitive-high trust profiles have a lower level of understanding of depression and may not understand the nature of the illness or the long-term nature of treatment, but they have a high level of trust and adherence to medical advice. For these patients, this application increases the weighting of anesthesia, medication, and social support interventions, while decreasing the weighting of individual cognitive adjustment. This is because these patients have limited cognitive abilities and are unlikely to achieve good results through complex cognitive restructuring and psychological analysis. Prematurely or excessively emphasizing individual cognitive adjustment may lead to patient confusion or an increased treatment burden. Therefore, anesthesia and medication are prioritized to quickly stabilize symptoms, while structured guidance and support are provided through social support interventions. Basic cognitive education is gradually introduced after the patient's condition stabilizes, rather than in-depth individual cognitive adjustment.

[0114] Patients with low cognitive and low trust levels have a lower level of understanding of depression and a lower level of trust in medical advice, making them the most difficult group to treat. For these patients, this application increases the weighting of short-term anesthesia treatment to quickly assess whether they can benefit from it. This is because these patients have poor adherence to oral medication and low acceptance of social support and personal cognitive adjustments. Anesthesia treatment, administered under professional medical supervision, does not rely on the patient's subjective cooperation and can rapidly improve core symptoms and reduce suicide risk in a short period. If the patient demonstrates clinical benefit after short-term anesthesia treatment, such as a significant decrease in the Montgomery-Asperger's Depression Rating Scale score and a reduction in suicidal ideation, treatment will continue and social support interventions will be gradually introduced to build trust. If there is no clinical benefit, the patient will be quickly discharged or referred to other treatment facilities to avoid wasting medical resources and increasing the burden on patients due to ineffective treatment.

[0115] Through the above-mentioned individual type-based differentiated weight correction scheme, this application achieves precise and personalized adjustment of the initial resource allocation weights in step S3, which significantly improves the acceptance of treatment plans and treatment compliance of patients with different characteristics.

[0116] S4: The patient is treated using the personalized treatment plan, and digital monitoring indicators of the patient are collected in real time during the treatment process. The personalized treatment plan is dynamically adjusted / intervened and executed based on the digital monitoring indicators to generate treatment execution data containing real-time adjustment records.

[0117] In some embodiments, in step S4, the digital monitoring indicators include at least one or more of heart rate, blood pressure, blood oxygen saturation, electroencephalogram, heart rate variability, facial expression, and body movements.

[0118] And / or, dynamically adjusting / intervening in and implementing the personalized treatment plan based on the digital monitoring indicators, including: The digital monitoring indicators are input into a preset side effect prediction model to predict the probability of side effects occurring within a preset time window in the future. When the risk probability exceeds a preset safety threshold, at least one of the following operations is performed: Adjust the administration rate of anesthetic drugs or suspend administration; Trigger the sound and light soothing intervention procedure; Send warning notifications to the physician's interface.

[0119] In one specific implementation, in step S4, this application dynamically adjusts and actively intervenes in personalized treatment plans by real-time acquisition of the patient's digital monitoring indicators. The digital monitoring indicators include at least one or more of the following: heart rate, blood pressure, blood oxygen saturation, electroencephalogram (EEG), heart rate variability, facial expressions, and body movements. Specifically, heart rate and blood pressure reflect the patient's cardiovascular status and stress level; blood oxygen saturation monitors the patient's respiratory function and oxygen supply; EEG captures changes in brain electrical activity, particularly power changes in specific frequency bands such as gamma waves, which are closely related to dissociative symptoms; heart rate variability reflects the balance of the autonomic nervous system, and a decrease in its high-frequency components often indicates an enhanced stress response; facial expressions and body movements are captured by video sensors and reflect the patient's emotional state and behavioral activation level. These multimodal digital monitoring indicators are acquired in real-time at a high-frequency sampling rate, forming continuous physiological and behavioral time-series data, providing a data foundation for real-time risk assessment and dynamic intervention.

[0120] Based on the aforementioned digital monitoring indicators, the specific process for dynamically adjusting and intervening in personalized treatment plans in this application is as follows. First, the digital monitoring indicators are input into a preset side effect prediction model to predict the probability of side effects occurring within a preset time window. The side effect prediction model is constructed using machine learning algorithms, such as temporal convolutional networks or long short-term memory networks. It uses real-time collected streaming data such as EEG, heart rate, heart rate variability, and blood pressure as input, and outputs the probability of specific side effects, such as dissociative symptoms, sudden increases in blood pressure, and respiratory depression, occurring within the next tens of seconds to several minutes. The preset time window can be set according to the clinical characteristics and urgency of the side effects; for example, for acute dissociative symptoms, it can be set to 30 seconds to 2 minutes in the future; for blood pressure fluctuations, it can be set to 1 to 5 minutes in the future.

[0121] When the risk probability exceeds a preset safety threshold, this application automatically performs at least one of the following operations: First, adjust the administration rate of the anesthetic drug or suspend administration. For example, when the system predicts a risk probability of separation symptoms exceeding 80%, it automatically reduces the infusion rate of esketamine by 50% or suspends the infusion, resuming administration at a lower rate after the risk signal disappears. Second, trigger an audio-visual reassurance intervention program. For example, when the system detects an abnormal increase in the gamma band power of the patient's EEG and a sudden drop in the high-frequency component of heart rate variability, it automatically triggers the slow illumination of the soft lights in the treatment room and plays preset guiding voice reassurance content to help the patient alleviate anxiety and fear, and reduce the subjective experience of separation symptoms. Third, push a warning prompt to the physician's interface. For example, a red warning window pops up on the physician's interface, displaying the risk type, risk probability confidence level, and specific recommended measures such as "It is recommended to reduce the infusion rate by 50%" or "It is recommended to suspend administration and observe the patient's condition." The above warning prompt also includes a timestamp, a description of the risk characteristics, and recommended intervention actions, facilitating physicians to make rapid clinical decisions.

[0122] Through the aforementioned real-time dynamic adjustment and proactive intervention mechanism based on digital monitoring indicators, this application achieves a fundamental shift from passive post-event handling to proactive in-event prediction and control in step S4. This mechanism can respond before or in the early stages of side effects, curbing them in their nascent stage, significantly reducing the likelihood of patients refusing treatment due to fear of side effects, and improving treatment safety and patient acceptance. Simultaneously, the system automatically records the timestamp, physiological signal characteristics, and intervention measures for each risk event, which are used for subsequent model optimization and efficacy analysis.

[0123] S5: Perform a phased evaluation of the treatment execution data according to the preset efficacy evaluation criteria, and obtain the evaluation results.

[0124] In some embodiments, if the assessment result reaches the cure endpoint defined in the efficacy evaluation criteria, the treatment is terminated; if the assessment result does not reach the cure endpoint, the personalized resource allocation weights are updated according to the assessment result, and the updated personalized resource allocation weights are used as the basis for formulating the next stage of treatment plan. Steps S4 to S5 are repeated until the cure endpoint is reached.

[0125] The efficacy evaluation criteria are constructed based on multiple evaluation dimensions, including follow-up weeks, symptom relief rate, MADRS score, use of oral medication for mood therapy, and at least one of the number of anesthesia treatments. The multiple evaluation dimensions are combined according to preset grading thresholds to classify at least one efficacy level: relief / cure, likely relief / cure, significant effect, effective, possibly effective, possibly ineffective, and ineffective. The cure endpoint is the combination of the grading thresholds of the multiple evaluation dimensions corresponding to the relief / cure level.

[0126] In one specific implementation, this application proposes for the first time a visualized seven-level quantitative standard for the relief / cure of depression (see Table 1). This standard aims to help patients and their families clearly understand the treatment pathway, adjust treatment expectations, and thereby improve treatment adherence.

[0127] Table 1 - Efficacy Evaluation Criteria Referring to Table 1, the efficacy evaluation criteria described in this application, based on evaluation dimensions such as follow-up weeks, symptom relief rate, Montgomery-Asperger's Depression Rating Scale (MADRS) score, use of oral mood therapy medications, and number of anesthesia treatments, divide clinical efficacy into seven levels and nine sub-levels. Specifically, the relief / cure levels include Level Ia and Level Ib: Level Ia is defined as a follow-up of 25 to 48 weeks, a symptom relief rate of 90% or higher, a MADRS score of no more than 6 points at the last follow-up, the patient has discontinued all oral mood therapy medications, and the number of anesthesia treatments is 0; Level Ib is defined as a follow-up of 25 to 48 weeks, a symptom relief rate of 90% or higher, a MADRS score of no more than 6 points at the last follow-up, the patient has discontinued all oral mood therapy medications, and the number of anesthesia treatments is 2. It is highly likely that the remission / cure level will be Ia (early termination), defined as a follow-up of at least 12 weeks, a symptom relief rate of 90% or higher, a MADRS score of no more than 6 at the last follow-up, the patient has discontinued all oral medications for mood therapy, and the number of anesthesia treatments is 0. The efficacy levels are divided into Class IIa and Class IIb: Class IIa includes two scenarios. The first scenario is a follow-up of 13 to 24 weeks, with a symptom relief rate of 90% or higher, a MADRS score of no more than 6 at the last follow-up, a reduction or discontinuation of oral mood therapy medication of 75% or higher, and 2 anesthesia treatments. The second scenario is a follow-up of 13 to 24 weeks, with a symptom relief rate between 75% and 89%, a MADRS score of no more than 10 at the last follow-up, the patient has discontinued all oral mood therapy medications, and 0 anesthesia treatments. Class IIb is defined as a follow-up of 13 to 24 weeks, with a symptom relief rate between 75% and 89%, a MADRS score of no more than 10 at the last follow-up, the patient has discontinued all oral mood therapy medications, and no more than 6 anesthesia treatments. The effective levels are divided into IIIa and IIIb: IIIa is defined as a follow-up of 5 to 12 weeks, with a symptom relief rate of 75% or higher, a MADRS score of no more than 10 at the last follow-up, a reduction or discontinuation of oral medications for mood therapy of 50% or higher, and no more than 8 anesthesia treatments; IIIb is defined as a follow-up of 5 to 12 weeks, with a symptom relief rate between 50% and 74%, a MADRS score of no more than 16 at the last follow-up, a reduction or discontinuation of oral medications for mood therapy of 50% or higher, and no more than 8 anesthesia treatments. Potentially effective levels include IVa and IVb: IVa is defined as a follow-up of 0 to 4 weeks, a symptom relief rate between 50% and 74%, a MADRS score of no more than 16 at the last follow-up, a reduction or discontinuation of oral mood therapy medication of 25% or more, and 2 to 8 anesthesia treatments; IVb is defined as a follow-up of 0 to 4 weeks, a symptom relief rate between 50% and 74%, a MADRS score of no more than 16 at the last follow-up, a reduction or discontinuation of oral mood therapy medication of less than 25%, and 2 to 8 anesthesia treatments.A possible ineffectiveness grade of V is defined as a symptom relief rate of less than 50%, a MADRS score exceeding 16 at the last follow-up, a reduction or discontinuation of oral mood therapy medication of less than 25%, and 4 to 8 anesthesia treatments. An ineffectiveness grade of V is also defined as a symptom relief rate of less than 50%, a MADRS score exceeding 16 at the last follow-up, unchanged oral mood therapy medication, and 7 to 8 anesthesia treatments. It should be noted that the above-mentioned possible ineffectiveness and ineffectiveness grade (i.e., V) often lead to the termination of innovative anesthesia treatment within 12 weeks.

[0128] The core components of the depression relief / cure standard system proposed in this application include: full-process management, i.e., a follow-up period of no less than 48 weeks; patient-reported outcomes, i.e., the relief rate of depressive symptoms is calculated based on the fuzzy score (NRS) completed by the patient, relief rate = (pre-treatment score - post-treatment score) / pre-treatment score × 100%; clinician-assessed outcomes, i.e., based on the Montgomery-Asperger Depression Rating Scale (MADRS) score; and anesthesia treatment assessment, i.e., the frequency and duration of the intervention, all anesthesia treatment protocols should be implemented under professional medical supervision and strictly follow the principles of individualized treatment and safety assessment.

[0129] In some embodiments, step S5 of this application further optimizes treatment efficiency and resource utilization. This application, for the first time, constructs a seven-level assessment system for the relief and cure of mood disorders, explicitly listing the discontinuation of all oral medications and the elimination of the need for continued anesthesia as high-level cure goals, while setting a complete treatment management cycle of 48 weeks. This system provides clear path guidance and scale assessment benchmarks for the phased treatment of intelligent decision-making systems, and is an important foundation for the systematic and standardized implementation of treatment plans.

[0130] The mood disorder relief and cure assessment system proposed in this application is aligned with the efficacy evaluation criteria in step S5 above. By quantifying clinical efficacy into actionable, phased goals, it enables patients and their families to clearly understand the treatment pathway, rationally adjust treatment expectations, and thus improve treatment adherence. This system uses "discontinuation of all oral medications" as one of the core indicators of cure, meaning that patients can maintain emotional stability without relying on medication. It also uses "no need for continued anesthesia" as another core indicator, indicating that anesthesia is only used as a rapid intervention in the early and middle stages of treatment and is no longer needed after reaching the cure endpoint. The complete treatment management cycle is set at 48 weeks, covering the entire process from intensive acute treatment to relapse prevention during the recovery phase.

[0131] Based on the aforementioned phased management model for the disease, this application further optimizes the frequency of anesthesia treatment to improve the efficiency of medical resource utilization while ensuring efficacy. Specifically, in the first 24 weeks of treatment, the total number of anesthesia treatments will not exceed 22, with the following phase allocation: Weeks 1 to 4 are the intensive treatment period, with 2 to 8 anesthesia treatments to quickly stabilize the patient's core symptoms and reduce the risk of suicide; Weeks 5 to 12 are the consolidation treatment period, with 2 to 8 anesthesia treatments to gradually reduce the frequency of treatment while maintaining the stability of core symptoms and beginning to strengthen non-pharmacological interventions; Weeks 13 to 24 are the maintenance treatment period, with 2 to 6 anesthesia treatments. At this time, the core symptoms have been basically eliminated, and the additional symptoms have gradually lessened. Anesthesia treatment is only used as a consolidation measure when necessary; Weeks 25 to 48 are the rehabilitation treatment period, with 0 to 2 anesthesia treatments. The patient has entered the rehabilitation stage, and almost all symptoms have been eliminated. Anesthesia treatment is only used as needed in special circumstances; From week 49 onwards, a long-term follow-up period is entered, with treatment implemented as needed and follow-up every 1 to 3 months to determine whether further intervention is needed based on the patient's condition.

[0132] It should be noted that while current guidelines for the treatment of depression recommend electroconvulsive therapy (ECT) for treatment-resistant depression (evidence level 1A), particularly for patients with psychotic symptoms, catatonia, or high suicide risk, the standard course of treatment is 6 to 12 sessions. Meanwhile, psychological interventions such as cognitive behavioral therapy, interpersonal therapy, and behavioral activation therapy typically involve 12 to 16 sessions, often weekly. However, due to the lack of clear phased treatment goals and closed-loop control mechanisms in existing guidelines, clinicians often face difficulties in determining subsequent treatment plans and adjusting strategies when patients fail to achieve the expected therapeutic effect. For example, when a patient's symptoms do not improve satisfactorily after 6 ECT sessions, existing guidelines fail to provide clear decision-making pathways regarding whether to increase the number of treatments, whether to switch to other treatment options, and how to assess the treatment endpoint. In contrast, this application provides clinicians with clear phased treatment goals and closed-loop control mechanisms by constructing an assessment system for the relief and cure of mood disorders and combining it with clear rules for the phased allocation of anesthesia treatment frequency. In step S5, at the end of each treatment phase, this application conducts a phased evaluation of the treatment execution data according to preset efficacy evaluation criteria. If the evaluation result reaches the cure endpoint, the treatment is terminated; if the cure endpoint is not reached, the personalized resource allocation weights are updated based on the evaluation results, and the updated weights are used as the basis for formulating the next stage of treatment plan. Steps S4 to S5 are repeated. This closed-loop iterative mechanism ensures the scientific nature and traceability of the treatment process, avoiding treatment delays or resource waste caused by a lack of clear adjustment criteria.

[0133] Through the aforementioned optimization strategies, this application achieves dual optimization of treatment efficiency and resource utilization in step S5. On the one hand, by establishing clear phased goals and evaluation criteria, unnecessary treatment prolongation is avoided, thus improving treatment efficiency. On the other hand, by rationally controlling the frequency of anesthesia treatment, the input of medical resources is minimized while ensuring efficacy, reducing the economic burden and time costs for patients, and providing an efficient and economical solution for the full-course management of depression.

[0134] In some embodiments, step S5 involves updating the personalized resource configuration weights based on the evaluation results, including: Based on the current efficacy level and its corresponding follow-up weeks, the patient's current treatment stage is identified; the treatment stage includes the intensive treatment period, consolidation period, maintenance period, and recovery period. Based on the treatment stage, determine the target efficacy level for the next stage, compare the actual values ​​of each evaluation dimension corresponding to the current efficacy level with the threshold values ​​of each evaluation dimension corresponding to the target efficacy level for the next stage, and calculate the deviation value of each evaluation dimension. Based on the deviation values ​​of each evaluation dimension, the weight adjustment amount of each treatment method is calculated; The weight adjustment amount is added to the current personalized resource configuration weight to generate the updated personalized resource configuration weight.

[0135] In one specific implementation, step S5 involves updating the personalized resource allocation weights based on the evaluation results, specifically including the following sub-steps: First, based on the current efficacy level and its corresponding follow-up weeks, the patient's current treatment stage is identified. These stages include intensive treatment, consolidation, maintenance, and rehabilitation. Specifically, the intensive treatment stage corresponds to weeks 1-4 of treatment, with the goal of rapidly stabilizing the patient's core symptoms and reducing suicide risk; the consolidation stage corresponds to weeks 5-12, with the goal of reducing treatment frequency and medication dosage while maintaining stable core symptoms and strengthening non-pharmacological interventions; the maintenance stage corresponds to weeks 13-24, with the goal of gradually eliminating additional symptoms and strengthening social support while gradually reducing medication; and the rehabilitation stage corresponds to weeks 25-48, with the goal of consolidating rehabilitation effects, preventing relapse, and gradually restoring social function. By identifying the patient's current treatment stage, the system can clearly define the treatment goals and priorities for the next stage.

[0136] Secondly, based on the treatment stage, the target efficacy level for the next stage is determined. The actual values ​​of each evaluation dimension corresponding to the current efficacy level are compared with the threshold values ​​of each evaluation dimension corresponding to the target efficacy level for the next stage, and the deviation value of each evaluation dimension is calculated. The evaluation dimensions include at least one of the following: follow-up weeks, symptom relief rate, Montgomery-Asperger's Depression Rating Scale score, use of oral mood therapy medication, and number of anesthesia treatments. For example, if the patient is currently in the final stage of the consolidation phase, and the current efficacy level is effective (IIIa or IIIb), and the target efficacy level for the next stage is significantly effective (IIa or IIb), then the system compares the current actual symptom relief rate with the symptom relief rate threshold of 75% to 89% corresponding to the significantly effective level, and compares the current actual Montgomery-Asperger's Depression Rating Scale score with the score threshold of no more than 10 points corresponding to the significantly effective level, and calculates the deviation value of each dimension. A positive deviation value indicates that the dimension has reached or exceeded the target threshold, and a negative deviation value indicates that the dimension has not yet reached the target threshold and requires stronger intervention.

[0137] Next, based on the deviation values ​​of each evaluation dimension, the weight adjustment amount for each treatment method is calculated. Specifically, the system determines the direction and magnitude of the weight adjustment according to the correlation mapping relationship between the deviation values ​​of each evaluation dimension and each treatment method. For example, if the symptom relief rate deviation value is negative, meaning the target threshold has not been reached, and is mainly attributed to poor control of core emotional symptoms, the system correspondingly increases the weight adjustment amount for anesthesia and drug treatment; if the Montgomery-Asperger's Depression Rating Scale score deviation value is negative, and is mainly attributed to impaired social functioning, the system correspondingly increases the weight adjustment amount for social support intervention and personal cognitive adjustment; if the number of anesthesia treatments has reached or exceeded the target threshold but the symptom relief rate still fails to meet the target, the system correspondingly decreases the weight adjustment amount for anesthesia treatment and instead increases the weight of other non-drug intervention methods. The weight adjustment amount can be calculated using a preset adjustment coefficient matrix, where the deviation value of each evaluation dimension corresponds to a set of adjustment coefficients for each treatment method, and the weight adjustment amount for each treatment method is obtained by multiplying the deviation value by the adjustment coefficient.

[0138] Finally, the weight adjustment is added to the current personalized resource configuration weight to generate an updated personalized resource configuration weight. Specifically, the updated weight equals the current weight plus the corresponding weight adjustment, while ensuring that the sum of the weights of all treatment methods is 100%, and that no single weight is lower than a preset minimum weight threshold (e.g., 0%) or higher than a preset maximum weight threshold (e.g., 60%). The updated personalized resource configuration weight serves as the basis for formulating the next stage of the treatment plan. The system generates a new treatment plan based on this, and repeats steps S4 to S5 until the endpoint of cure is reached.

[0139] Through the aforementioned weight update mechanism, this application realizes dynamic weight adjustment based on the phased evaluation results in step S5, enabling the treatment plan to be refined and adaptively optimized as the patient's condition changes and treatment progresses, thereby ensuring that the treatment strategy always matches the patient's real-time status and improving the targeting and effectiveness of the treatment.

[0140] In some embodiments, this application further configures a full-cycle management strategy for mood disorders in step S5. Addressing the dynamic changes in clinical manifestations of mood disorders as the pathological process progresses, this application, for the first time, constructs a disease stage division system dynamically linked to treatment goals and efficacy standards. This system divides the entire treatment process into five core management stages, clearly defining the differentiated intervention priorities and assessment nodes for each stage, forming standardized matching rules between disease stages and intervention measures. These rules serve as the core basis for the intelligent decision-making system to perform stage-based assessments and weight updates in step S5, thereby achieving precise adaptation of treatment strategies and systematic advancement of disease management.

[0141] Specifically, the full-cycle management strategy proposed in this application compares the differences between traditional antidepressant drug treatment and the innovative anesthetic treatment proposed in this application at each stage. In weeks 1-4 of treatment, the traditional antidepressant drug treatment strategy is in the acute phase, with the primary goal of achieving drug efficacy; while the innovative anesthetic treatment strategy is in the intensive phase, with the primary goal of rapidly stabilizing the patient's core symptoms. In weeks 5-12, the traditional strategy enters the consolidation phase, maintaining the original drug dosage to keep core symptoms stable; the innovative anesthetic treatment strategy is also in the consolidation phase, maintaining core symptom stability while beginning to reduce treatment frequency and drug dosage, strengthening non-pharmacological interventions, and initiating the tapering process. In weeks 13-24, the traditional strategy is in the maintenance phase, with the main task being slow tapering and reducing withdrawal syndrome; the innovative anesthetic treatment strategy is also in the maintenance phase, at which point core symptoms have largely disappeared, and additional symptoms are gradually reduced. While gradually tapering medication, social support is strengthened, and anesthetic treatment is used as needed for consolidation. Between weeks 25 and 48, the traditional strategy primarily involves tapering medication, with patients often reducing or discontinuing their medication on their own. If the condition relapses, the original treatment must be quickly resumed. The innovative anesthetic treatment strategy described in this application, however, focuses on the recovery phase, where almost all symptoms have disappeared, patient cognition is rebuilt, and social function gradually recovers. This strategy reduces the frequency of social support while emphasizing relapse prevention. From week 49 onwards, the traditional strategy involves follow-up every 6 months, but intervention is often delayed. The innovative anesthetic treatment strategy described in this application involves follow-up every 1 to 3 months, implementing on-demand treatment to maintain disease stability.

[0142] It should be noted that the relapse rate of depression is as high as 50% to 85%, and most relapses occur within two years of onset. Therefore, current guidelines recommend continuing antidepressant medication for at least six months after eight weeks of complete symptom or functional recovery, with discontinuation gradually completed over six to twelve weeks. Suicide risk should be monitored throughout treatment, and follow-up appointments every four weeks are recommended. Studies have shown that intervention within one year of clinical remission for patients and their families can significantly reduce the risk of relapse. However, in practice, social support and family intervention largely rely on mental health institutions. Currently, the mental health industry faces multiple constraints, including inadequate regulations, insufficient ethical oversight, uneven regional development, a disconnect between mental health services and the medical system, and lack of medical insurance coverage. These factors often prevent traditional antidepressant treatments from achieving the desired long-term efficacy.

[0143] To address the aforementioned challenges, this application proposes integrating social support therapy into the early stages of the intervention process. Building upon the rapid symptom improvement and patient-doctor trust established by medication, the simultaneous introduction of structured social support interventions can further consolidate therapeutic effects and improve treatment adherence, thus creating conditions for earlier initiation of medication tapering. During medication tapering, the scientifically designed and appropriately applied placebo strategy can effectively mitigate withdrawal reactions, making the tapering process smoother. This integrated intervention model can be extended to 48 weeks of treatment and beyond, promoting the recovery of patients' social function and improving their psychological adaptability, ultimately achieving long-term disease stability, significantly reducing the probability of relapse, and gradually increasing the likelihood of clinical cure.

[0144] Through the aforementioned full-cycle management strategy, in step S5, this application can conduct a phased assessment of treatment execution data based on the patient's current treatment stage and pre-set efficacy evaluation criteria. The personalized resource allocation weights are then updated based on the assessment results, and the updated weights serve as the basis for formulating the next stage of treatment plans. Steps S4 to S5 are repeated until the cure endpoint is reached. This strategy achieves a complete closed loop from short-term symptom control to long-term rehabilitation management, providing a systematic solution for the full-course treatment of depression.

[0145] In some embodiments, this application proposes an intelligent decision-making system that relies on the comprehensive evaluation of multi-source heterogeneous information, including key basic patient information, medical history, severity grading of mood disorder symptoms, various specialized assessment scales, neuropsychiatric physical examinations, standardized mood assessment tools, auxiliary examination results, dynamic physiological signals, and behavioral data. Current clinical practice suffers from bottlenecks such as complex information verification, fragmented data collection processes, and limited clinical roles. Furthermore, existing technologies for processing multimodal medical data often remain at the level of simple overlay or superficial comparison, resulting in insufficient early prediction capabilities for treatment efficacy and side effects, and poor model interpretability. To overcome these difficulties, this application proposes an intelligent decision-making system that firstly performs unified rating and quantitative scoring on all collected multi-source data and transforms it into standardized language to achieve precise matching between patient treatment needs and intervention measures; simultaneously, it constructs an innovative efficacy evaluation system to help patients establish reasonable treatment expectations and integrates a disease-specific literature database to provide evidence-based support for interventions for each symptom. For core symptoms, rapid risk control is achieved through real-time monitoring and a high-risk early warning mechanism combined with potent drug intervention and psychosocial support. For additional symptoms, a synergistic treatment plan is formed by integrating cognitive-behavioral adjustments, nutritional and exercise guidance, enhanced social support, and physical therapy. "Comfortable medication reduction" support is provided during the medication tapering and discontinuation period. In terms of core technology implementation, this application employs an incremental learning algorithm to enable the system to continuously learn and optimize the model from new data. Combined with graph neural networks, deep feature extraction and correlation analysis are performed on high-dimensional time-series and structured data to model complex nonlinear relationships between different data modalities. This allows for earlier and more accurate prediction of treatment responses and side effect trends. Furthermore, the interpretability of the model is improved through visualized feature correlation graphs, ultimately breaking down data silos and achieving a technological leap from surface comparison to deep correlation mining.

[0146] In some embodiments, the intelligent decision-making system proposed in this application is based on a dynamic strategy matrix, including a core symptom-driven anesthesia decision matrix, a precise adaptation strategy for innovative anesthesia treatment of mood disorders, and guiding principles for multidisciplinary diagnosis and treatment based on mood assessment results. These strategies need to be dynamically adjusted based on real-time, multi-dimensional information such as the patient's demographic characteristics, core symptoms and degree of functional impairment, drug tolerance history, and vital signs and behavioral observation data during treatment. However, in current clinical practice, the adjustment of treatment plans, especially drug dosages, is highly dependent on physician experience, facing challenges such as complex adjustment criteria with large individual differences, strong adjustment lag, lack of real-time control over specific risks such as dissociative symptoms and blood pressure fluctuations, and insufficient dynamic management of extreme risks such as suicide. To overcome the above difficulties, this application designs an intelligent decision-making system with real-time dynamic adjustment and proactive risk intervention capabilities. The system first automatically classifies patients' clinical symptoms and risk factors. In the initial stages of treatment, it prioritizes addressing life-threatening issues such as high suicide or self-harm risks. Once acute risks stabilize, it systematically develops intervention sequences for additional symptoms. Simultaneously, it integrates a real-time suicide risk assessment and specific intervention module. Interventions are automatically triggered when assessment scores reach high-risk thresholds or specific risk factor clusters are identified, achieving coordinated risk control. Addressing patients' concerns about long-term medication and their desire to reduce dosage, this application extends the role of anesthetic treatment from short-term symptom control to assisting in the smooth reduction and discontinuation of oral medications, establishing trust and a treatment alliance through structured informed consent. After treatment begins, the system continuously collects patient feedback on efficacy and side effects and initiates adaptive learning to dynamically recommend reducing drug dosage, adding side-effect antagonists, or initiating non-pharmacological interventions. This application also specifically constructs a side effect feature database and a real-time early warning algorithm to achieve proactive prediction and management of side effects: before treatment, patients are informed of possible dissociative symptoms and their level of fear is assessed, and the drug administration rate is preset accordingly; during treatment, when the system detects an abnormal increase in power in a specific frequency band through real-time EEG signal monitoring and simultaneously detects a sudden drop in heart rate variability, it automatically determines that the state is at high risk of dissociative symptoms, immediately triggers a standardized audio-visual reassurance intervention procedure, and issues an early warning prompt to the operating physician to adjust the drug administration rate or suspend drug administration. To illustrate the real-time adjustment and proactive intervention capabilities of this application, an example is provided: During a patient's intravenous esketamine infusion treatment, the system monitors their electroencephalogram (EEG) and electrocardiogram (ECG) signals in real time. When the algorithm identifies that the power of the EEG gamma band is abnormally high in a specific lead, exceeding the threshold, and accompanied by a decrease of more than 30% in the high-frequency component of heart rate variability within one minute, it is immediately judged as a high risk of acute dissociative symptoms. The system automatically triggers the soft lights in the treatment room to slowly illuminate and plays preset guiding voice reassurance content. At the same time, a red warning window pops up on the physician's operating interface, displaying the risk type, confidence level, and recommended measures, such as suggesting a 50% reduction in the infusion rate. The system also automatically records the timestamp of the event, physiological signal characteristics, and intervention measures for subsequent model optimization and efficacy analysis.In terms of core technology implementation, this application achieves second-level safety control of the treatment process by constructing an ultra-short-term prediction model based on high temporal resolution physiological feedback signals and automated intervention logic: using real-time EEG, heart rate variability, blood pressure and other streaming data to train machine learning models such as temporal convolutional networks and long short-term memory networks that can predict the probability of side effects occurring in the next tens of seconds to several minutes. Once the output of the prediction model exceeds the safety threshold, the system automatically triggers refined intervention actions such as adjusting the infusion pump administration rate, pausing administration, or activating physical and acoustic soothing devices according to preset rules, forming a proactive safety intervention closed loop, which greatly reduces the delay of human judgment and operation; at the same time, the system modularizes effective psychological therapies for high suicide risk and non-suicidal self-harm behaviors, such as dialectical behavior therapy and mentalization basic treatment, and dynamically recommends and integrates them into the patient's overall treatment path based on real-time risk assessment results, forming a synergistic management of drug and psychological intervention.

[0147] In some embodiments, the intelligent decision-making system proposed in this application addresses the core technical challenges of cold start and long-term adaptation of personalized models. For newly enrolled patients, the lack of individual historical data makes it difficult to quickly construct effective personalized treatment models; simultaneously, patients' physiological and psychological states change dynamically over time, making it difficult for fixed models to achieve long-term adaptation. To overcome these challenges, this application develops a core symptom-driven anesthesia decision matrix as a starting scheme based on clinical experience, and, according to the staged treatment goals defined by the efficacy evaluation criteria for innovative anesthesia treatment of depression and the full-cycle management strategy for mood disorders, combined with continuously enriched patient profile information and multiple physiological, biochemical, and psychological behavioral indicators during the treatment process, achieves long-term adaptive adjustment of the model. Regarding differentiated treatment strategies, this application dynamically adjusts the weighting of each treatment component, including anesthesia, cognitive adjustment, nutrition and exercise, social support, and physical therapy, at different treatment stages: during the intensive treatment period, the weight of anesthesia is the highest; after entering the maintenance period, the weights of social support and personal cognitive adjustment significantly increase. This differentiated strategy follows the natural course of depression and the pattern of treatment response. In the early stages, potent methods are used to achieve rapid breakthroughs; in the mid-stage, the focus is on consolidating efficacy and managing withdrawal reactions; and in the long term, the emphasis is on building patients' sustainable self-management abilities and social support systems, thereby fundamentally reducing the risk of relapse. Dynamic weighting ensures precise allocation of medical resources and maximizes treatment efficiency. Regarding the treatment response and pathway adjustment mechanism, for patients whose core symptoms do not improve after four innovative anesthesia treatments within four weeks (i.e., a decrease of less than 30% in the Montgomery-Asperger's Depression Rating Scale score), the system automatically triggers an enhanced intervention module, providing timely individualized cognitive-behavioral adjustments and social support reinforcement measures. If no clinical response is observed after twelve weeks of treatment (i.e., a decrease of less than 50% in the Montgomery-Asperger's Depression Rating Scale score), innovative anesthesia treatment is discontinued, and a referral process to other institutions is initiated. For patients who receive no more than four treatments in the first four weeks and complete twelve weeks of treatment, the system conducts periodic assessments. If the initial treatment goals are achieved, such as a Montgomery-Asperger's Depression Rating Scale score of no more than 16 points and the patient enters the medication tapering phase, the patient proceeds to the next treatment cycle according to the established pathway. If the expected results are not achieved or intolerance occurs, the system initiates an exit or referral mechanism, thereby ensuring the scientific validity and safety of the treatment pathway. In terms of core technology implementation, this application is based on a transfer learning and federated learning framework. Under the premise of strictly protecting patient data privacy, it utilizes de-identified group knowledge to provide high-quality initialization for new patient models, significantly shortening the personalized modeling cycle. Simultaneously, the system incorporates an online learning mechanism that continuously tracks changes in patient treatment response patterns, enabling long-term adaptive updates of the model and ensuring that treatment recommendations always match the patient's real-time status.

[0148] In some embodiments, the intelligent decision-making system proposed in this application addresses the core technical challenge of systemic intervention for patients' cognitive biases and weak social support. In existing treatment models, patients often hold negative perceptions of medications, such as concerns about addiction or side effects, and their social support systems are typically weak. These socio-psychological factors are key obstacles leading to decreased treatment adherence and difficulty in maintaining long-term efficacy. Existing traditional treatment models often fail to adequately address these factors and lack systematic and quantifiable assessment and integration programs. To overcome these challenges, this application quantifies social support levels using standardized scales such as the Social Support Rating Scale and the Individualized Treatment Navigation Questionnaire for Mood Disorders, incorporating these assessments into the core decision-making parameter system. For patients with weak social support, the system automatically prioritizes and integrates modules such as structured psychoeducation, cognitive behavioral intervention, and family therapy, dynamically adjusting support reinforcement strategies throughout the treatment cycle. Simultaneously, the system includes a dedicated cognitive adjustment module to precisely intervene in patients' misconceptions about medications and diseases, aiming to improve their acceptance and confidence in treatment, thereby constructing a three-dimensional, systematic consolidation network encompassing medication, psychological, and social dimensions. Through the aforementioned technical solutions, this application achieves three major innovative breakthroughs: in terms of treatment modality, it upgrades from single-drug management to a dynamic integrated management system covering the entire treatment cycle; in terms of technical architecture, it completes a generational leap from offline assessment to real-time predictive regulation; and in terms of management mode, it constructs a complete intelligent closed loop of multimodal data and clinical decision-making. This system, through a data-driven real-time optimization mechanism, promotes the evolution of diagnosis and treatment models towards precision, dynamism, and systematization.

[0149] The proposed solution represents a fundamental shift from a static, singular, and drug-centric model to a dynamic, integrated, and patient-centric intelligent management system. Specifically, this application has a broader patient coverage, systematically encompassing patients throughout the entire course of their illness, from mild to severe, including those with complex additional symptoms such as pain, insomnia, and cognitive impairment. In terms of role positioning, this application identifies innovative anesthesia as a core treatment option that can be used independently or in combination with other treatments, clearly defining its crucial role in achieving "comfortable medication reduction." This involves planning and supplementing anesthesia at key points in medication reduction to help patients transition smoothly and reduce the risk of withdrawal reactions and relapse. Regarding integrated interventions, this application creatively standardizes non-pharmacological methods such as personal cognitive adjustment, nutritional and exercise support, social support interventions, and physical therapy into configurable treatment components. These components intelligently synergize with anesthesia to form an integrated prescription targeting both core and additional symptoms. This overcomes the limitations of traditional models that rely on single-drug dependence, providing a more systematic and efficient integrated treatment plan for the full course of managing mood disorders.

[0150] This application represents a generational leap from offline assessment to real-time predictive regulation. Existing technologies suffer from shortcomings such as lagging data acquisition and model updates, lack of real-time intervention capabilities, passive side effect management, and slow initiation of personalized treatment plans. In contrast, this application utilizes a streaming data processing engine to perform second-level analysis of multimodal signals, leveraging lightweight neural networks to achieve a fundamental shift from post-treatment assessment to in-process prediction and regulation. Regarding real-time dynamic regulation, this application can perform millisecond-to-second real-time analysis of physiological signals during treatment, responding before or at the initial stage of side effects, rather than waiting for retrospective assessment after treatment. In terms of proactive side effect management, this application constructs a side effect feature library and early warning algorithm. For example, it presets an initial dosage based on pre-treatment patient feedback, identifies risk characteristics in EEG and physiological signals in real-time during treatment, and automatically triggers reassurance interventions or dosage adjustment suggestions, transforming side effect management from passive post-treatment to proactive pre-treatment prevention. In terms of adaptive learning, this application adopts a federated learning framework to achieve cross-institutional knowledge sharing. Under the premise of strictly protecting patient data privacy, it uses de-identified group knowledge to provide high-quality initialization for new patient models, effectively solving the "cold start" problem of new patients lacking historical data. At the same time, through an online learning mechanism, it continuously tracks changes in patient status, realizes real-time adaptive updates and professional supervision of the model, and ensures that treatment recommendations always match the real-time status of patients, thereby promoting the evolution of diagnosis and treatment technology from static, lagging offline assessment to dynamic, real-time predictive regulation.

[0151] This application constructs a complete intelligent closed loop for multimodal data and clinical decision-making in its management model, forming a closed-loop system of multimodal data perception, intelligent model decision-making, compound intervention execution, and effect feedback optimization. At the data layer, this application integrates in real-time physiological data such as EEG, heart rate, blood pressure, and blood oxygen saturation; behavioral data such as activity level, sleep rhythm, facial expressions, and body movements; clinical assessments such as scale scores, medical records, and neuropsychiatric examination results; and patient-reported data, among other multi-source heterogeneous information. It utilizes a time-series database and feature engineering pipeline to perform time alignment, noise reduction, and feature extraction on the heterogeneous data, forming a unified patient state vector to provide standardized input for subsequent decision-making. At the decision-making layer, this application inputs the integrated data into the core model to predict efficacy and risk, and quantitatively assesses factors such as social support level, nutritional and exercise status, and cognitive function status. Based on this, the model dynamically generates personalized "treatment component weight prescriptions." For example, when data indicates weakened social support and a decline in mood scores, the system automatically increases the weight of social support interventions and recommends specific reinforcement tasks such as increasing the frequency of family therapy and initiating peer support programs. At the execution and feedback layer, treatment recommendations are implemented after review by clinicians. Some standardized instructions, such as adjusting the infusion pump's drug delivery rate, setting smart mattress parameters, and pushing health education content, can be completed automatically by the system without manual intervention. New data generated after execution, such as treatment responses, side effects, and patient feedback, are then fed back into the system to continuously optimize model parameters and decision rules, forming a self-evolving learning loop. Through the above three-layer collaborative mechanism, this application achieves fully automated closed-loop management from data collection to decision generation and execution feedback, ensuring that treatment plans can be dynamically adjusted in real time according to changes in the patient's condition, and promoting the continuous evolution of the diagnosis and treatment model towards precision, dynamism, and systematization.

[0152] Further reference Figure 2 As an implementation of the above-described method, this application provides an embodiment of an emotion disorder assessment system, which is similar to... Figure 1 Corresponding to the method embodiments shown, the system can be specifically applied to various electronic devices.

[0153] refer to Figure 2 A mood disorder assessment system, comprising: User data acquisition module 210 is configured to collect user data from patients. The user data includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators.

[0154] The initial resource configuration module 220 is configured to standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set the initial resource configuration weights of multiple treatment methods based on the severity level of the disease.

[0155] The personalized treatment plan generation module 230 is configured to acquire the patient's psychological characteristic data, which includes at least the patient's cognitive level of depression and trust in medicine. The module corrects the initial resource allocation weights based on the psychological characteristic data to generate personalized resource allocation weights, and generates a personalized treatment plan based on the personalized resource allocation weights.

[0156] The treatment execution module 240 is configured to treat the patient using the personalized treatment plan, and to collect the patient's digital monitoring indicators in real time during the treatment process, and to dynamically adjust / intervene the personalized treatment plan based on the digital monitoring indicators and execute it, so as to generate treatment execution data containing real-time adjustment records.

[0157] The efficacy evaluation module 250 is configured to perform phased evaluations of the treatment execution data according to preset efficacy evaluation criteria and obtain evaluation results. If the assessment results reach the cure endpoint defined in the efficacy evaluation criteria, then treatment shall be terminated; If the assessment result does not reach the cure endpoint, the personalized resource allocation weights are updated based on the assessment result, and the updated personalized resource allocation weights are used as the basis for formulating the next stage of treatment plan. The execution steps in the treatment execution module and efficacy evaluation module are repeatedly called until the cure endpoint is reached.

[0158] The efficacy evaluation criteria are constructed based on multiple evaluation dimensions, including follow-up weeks, symptom relief rate, MADRS score, use of oral medication for mood therapy, and at least one of the number of anesthesia treatments. The multiple evaluation dimensions are combined according to preset grading thresholds to classify at least one efficacy level: relief / cure, likely relief / cure, significant effect, effective, possibly effective, possibly ineffective, and ineffective. The cure endpoint is the combination of the grading thresholds of the multiple evaluation dimensions corresponding to the relief / cure level.

[0159] From another perspective, this embodiment provides an auxiliary decision-making recommendation system for mood disorders, the system comprising: The user data collection module is configured to collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators. Each data item in the user data is pre-labeled with a corresponding privacy level and detection difficulty level. A collection strategy is selected based on the patient's stage of treatment, and the collection strategy includes a privacy threshold and / or a difficulty threshold. User data is selected according to the collection strategy, wherein the privacy level of the user data is lower than the privacy threshold, or the difficulty level of the user data is lower than the difficulty threshold. The initial resource configuration module is configured to standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set the initial resource configuration weights of multiple treatment methods based on the severity level of the disease. The auxiliary decision generation module is configured to acquire the patient's psychological characteristic data to update the initial resource allocation weights to obtain auxiliary decision opinions. The psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine. The initial resource allocation weights are corrected based on the psychological characteristic data to generate at least one personalized resource allocation weight, and the combination of the personalized resource allocation weights is output as auxiliary decision opinions.

[0160] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the following... Figure 1 or Figure 3 The method shown.

[0161] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0162] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a computer terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0163] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.

Claims

1. A method for recommending auxiliary decision-making opinions for mood disorders, characterized in that, Includes the following steps: S1: Collect patient user data, which includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators; wherein, S1 includes: Each piece of data in the user data is pre-labeled with a corresponding privacy level and detection difficulty level; The data collection strategy is selected based on the patient's stage of admission, and the data collection strategy includes: privacy threshold and / or difficulty threshold; The user data is selected according to the collection strategy, wherein the privacy level of the user data is lower than the privacy threshold, or the inspection difficulty level of the user data is lower than the difficulty threshold; S2: Standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set the initial resource allocation weights for multiple treatment methods based on the severity level of the disease. S3: Obtain the patient's psychological characteristic data to update the initial resource allocation weights to obtain auxiliary decision-making opinions. The psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine. The initial resource allocation weights are corrected based on the psychological characteristic data to generate at least one personalized resource allocation weight. The personalized resource allocation weights are then combined and output as auxiliary decision-making opinions.

2. The method according to claim 1, characterized in that, The multiple treatment methods in step S2 include at least one of anesthesia, pharmacology, personal cognitive adjustment, nutritional and exercise support, social support intervention, and physical therapy.

3. The method according to claim 2, characterized in that, The patient self-reported data collected in step S1 is compared with the objective test data to determine the consistency and reliability between the patient self-reported data and the objective test data; the objective test data includes at least auxiliary examination data, digital monitoring indicators, and the BMI index in the basic information. If there is a conflict between the patient's self-reported data and the objective test data that exceeds a preset threshold, then concealment behavior is identified. When underreporting is detected, the initial resource allocation weights are adjusted, including: Increase the initial weight of social support interventions and personal cognitive adjustment in the initial resource allocation weights.

4. The method according to claim 3, characterized in that, It also includes the following steps: When a patient is identified as concealing information, the severity of the concealment should be determined. Set privacy thresholds and / or difficulty thresholds based on the severity level.

5. The method according to claim 1, characterized in that, Including the following steps: Identify the magnitude of change in personalized resource allocation weights within a set period; When the change is less than a set value, a threshold update is triggered; wherein, the threshold update is to update the privacy threshold or update the difficulty threshold.

6. The method according to claim 2, characterized in that: The disease severity grading in step S2 includes mild, moderate, and severe stages; And / or, the configuration rule for the initial resource configuration weights is: In the mild stage, interventions primarily focus on individual cognitive adjustment and social support. Initiate drug therapy during the moderate stage, and increase the weight of nutritional and exercise support and physical therapy; In the severe stage, the main treatments are anesthesia and medication, while social support interventions are strengthened simultaneously.

7. The method according to claim 1, characterized in that: In step S2, the user data is standardized and scored to obtain a scoring result. Based on the scoring result, the severity level of the patient's disease is determined, including at least one of the following steps: Step a: Assign scores to basic information, at least classify gender, age, and BMI according to the preset level classification standard and assign corresponding scores; Step b: Assign scores to medical history data. At least the course of the disease, severity of core symptoms, accompanying symptoms, medication history, negative events, and family history indicators should be graded according to the preset grading standards and assigned corresponding scores. Step c: Neuropsychiatric examination and scoring, at least the communication status, emotional activity, and cognitive function indicators are graded according to the preset grading standard and assigned corresponding scores; Step d: Questionnaire assessment and scoring. At least sleep quality, anxiety and depression symptoms, suicide risk, social support level, and cognitive function indicators should be graded according to the preset level classification standard and assigned corresponding scores. Step e: Auxiliary examination scoring, at least organ function indicators, neuroendocrine indicators, nutritional status indicators, and drug metabolism indicators are graded according to the preset grading standard and assigned corresponding scores; Step f: Calculate the comprehensive score. Sum the scores assigned to each item in steps a to e to obtain the patient's comprehensive score, and determine the severity level of the disease based on the comprehensive score. And / or, In step S2, when it is not possible to obtain the scoring data for all sub-items, the preset core items are used to assign values ​​to determine the severity level of the patient's disease. During the treatment process, data related to the scoring are obtained to correct the severity level of the disease.

8. The method according to claim 2, characterized in that, Step S3 also includes: Based on patients' level of understanding of depression and their level of trust in medicine, patients are classified into different personality types; these personality types include high cognition-high trust type, high cognition-low trust type, low cognition-high trust type, and low cognition-low trust type. Based on the individual type, a corresponding weight adjustment scheme is selected to adjust the initial resource configuration weights to generate personalized resource configuration weights, including: If the personality type is high cognition-high trust, then increase the weight of personal cognitive adjustment and social support intervention; If the personality type is high cognition-low trust, then while maintaining the weight of necessary anesthetic treatment, the weight of social support intervention and personal cognitive adjustment should be increased. If the personality type is low cognition-high trust, then increase the weight of anesthesia, medication and social support interventions, and decrease the weight of personal cognitive adjustment. If the personality type is low cognition-low trust, then the weight of short-term anesthesia treatment is increased.

9. The method according to any one of claims 1-8, characterized in that, It also includes the following steps: Each piece of data in the user data is labeled with a privacy level and a detection difficulty level according to preset rules to obtain a preset initial collection strategy; the privacy level includes at least high privacy level data and low privacy level data, and the detection difficulty level includes at least high difficulty level data and low difficulty level data; During a patient's first visit, based on a pre-set initial data collection strategy, data with low privacy levels and low difficulty levels are collected first. And / or, During treatment, the current efficacy level is obtained. If the current efficacy level reaches the preset efficacy threshold, high privacy level data and high difficulty level data are collected.

10. A decision-making recommendation system for mood disorders, characterized in that, The system includes: The user data collection module is configured to collect user data from patients. The user data includes at least one of the following: basic information, medical history data, neuropsychiatric examination data, questionnaire assessment data, auxiliary examination data, patient self-report data, and digital monitoring indicators. Among them, each piece of data in the user data is pre-labeled with a corresponding privacy level and detection difficulty level; The data collection strategy is selected based on the patient's stage of admission, and the data collection strategy includes: privacy threshold and / or difficulty threshold; The user data is selected according to the collection strategy, wherein the privacy level of the user data is lower than the privacy threshold, or the difficulty level of the user data is lower than the difficulty threshold; The initial resource configuration module is configured to standardize and score the user data to obtain a scoring result, determine the severity level of the patient's disease based on the scoring result, and set the initial resource configuration weights of multiple treatment methods based on the severity level of the disease. The auxiliary decision generation module is configured to acquire the patient's psychological characteristic data to update the initial resource allocation weights to obtain auxiliary decision opinions. The psychological characteristic data includes at least the patient's cognitive level of depression and trust in medicine. The initial resource allocation weights are corrected based on the psychological characteristic data to generate at least one personalized resource allocation weight, and the combination of the personalized resource allocation weights is output as auxiliary decision opinions.

Images