Intelligent negentropy dry prediction algorithm based on trace-German-kernel-sense-gift semantic feedback

Abstract

The invention discloses an intelligent negentropy dry prediction algorithm based on Tao-German-kernel-sense-gift semantic feedback, and belongs to the technical field of medical intelligent control. According to the algorithm, multi-modal life data of a patient is continuously collected, life information entropy indexes are calculated to quantify the health state, value dimensions (Tao, Degree, kernel, meaning and gift) of entropy deviation are analyzed in combination with a DIKWP semantic knowledge graph, and a candidate intervention scheme set is generated accordingly. According to the algorithm, the scheme is filtered and optimized according to the five-dimensional moral criterion, and it is ensured that the selected intervention measures conform to the physiological law, the medical ethics and the willingness of the patient. Finally, the control module coordinates information field intervention, energy field intervention, behavior feedback intervention and other means to carry out negentropy correction on the patient, entropy change is monitored to dynamically adjust the scheme, and closed-loop negative feedback control is formed. The system can reverse the disease development trend in real time and maintain the human health steady state, realizes the conversion from passive treatment to active prevention, and has significant advantages in improving the chronic disease management effect and guaranteeing the medical decision ethics.

Description

Technical Field

[0001] This invention pertains to intelligent control technology in the field of medical and health care, and specifically relates to an intelligent negative entropy intervention algorithm based on semantic feedback of Tao-De-Ren-Yi-Li. Background Technology

[0002] With the development of wearable devices and biosensor technologies, people can acquire a large amount of multimodal data reflecting their vital signs in real time. However, in current technologies, the utilization of this data is mainly limited to passive monitoring and threshold alarms, lacking comprehensive quantitative indicators of health status and proactive intervention mechanisms. For example, conventional health monitoring systems will issue alarms when single parameters such as heart rate and blood pressure exceed the normal range, but often fail to detect trends indicating the body is heading towards disease. By the time patients develop obvious symptoms and begin treatment ("treating the disease as it has already occurred"), the optimal intervention time has often been missed. Therefore, the current medical model urgently needs to shift from passive response to proactive prevention ("treating disease before it occurs"), in order to correct and intervene in the early stages of disease.

[0003] The concept of entropy in information theory provides a tool for measuring the degree of disorder in a system. Life science theory states that living organisms maintain order and balance in their internal environment by continuously acquiring "negative entropy" (negative entropy, representing ordered energy or information) from the external environment. When an organism is healthy, its various physiological and biochemical processes maintain dynamic homeostasis, and information entropy is at a low level. When the organism is affected by disease or stress, internal regulation becomes unbalanced, and disorder increases, meaning that life information entropy rises. In traditional medical practice, there is a lack of methods to directly evaluate health status using changes in entropy values, but this concept is gradually attracting attention. Some specialized equipment has attempted to use entropy for clinical monitoring, such as using EEG signal entropy values ​​to assess a patient's state of consciousness during anesthesia depth monitoring. However, these applications only target specific physiological signals and have not yet formed a feedback intervention system for overall health status.

[0004] On the other hand, artificial intelligence (AI) technology is increasingly being applied in the medical field, such as using machine learning to analyze medical big data and assist in diagnostic decision-making. However, most current medical AI systems focus on providing diagnostic references or simple decision support, and are still in their infancy when it comes to real-time control and intervention. One important reason for this is the complexity of medical interventions: the optimal intervention plan for each patient may vary due to individual differences, and medical decisions need to consider multiple factors such as ethics, safety, and individual wishes. If decisions are made solely based on algorithm output, they may contradict clinical ethics (such as overtreatment or neglecting patient feelings). Existing automated control systems rarely incorporate humanistic and ethical factors into the decision-making process, making them difficult to directly apply in medical scenarios. This, to some extent, limits the application of AI in proactive medical intervention.

[0005] In summary, existing technologies have the following shortcomings: First, they lack a unified quantitative evaluation index for the overall health status of living organisms, making it impossible to detect subtle deviations from homeostasis in a timely manner; second, there are no mature closed-loop control mechanisms to automatically correct deviations and achieve dynamic health maintenance; third, existing automated decision-making systems lack consideration for medical ethics and humanistic care, which may lead to interventions that are not reasonable or acceptable to patients. Therefore, it is necessary to provide a new technical solution that can continuously assess the level of vital information entropy, detect signs of homeostasis imbalance in advance, and actively input "negative entropy" through multimodal means to correct the disorder. Simultaneously, this solution should incorporate feedback from ethical norms and humanistic semantics, ensuring that interventions comply with ethical principles and patient interests while guaranteeing technical feasibility, thereby achieving truly intelligent and safe medical negative feedback control. Summary of the Invention

[0006] The main objective of this invention is to overcome the shortcomings of existing health monitoring and intervention methods by providing an intelligent negative entropy intervention algorithm based on the semantic feedback of Tao, De, Ren, Yi, and Li. This algorithm can dynamically assess the entropy of life information and actively input negative entropy into the body through multimodal intervention, thereby reversing disease progression and maintaining health homeostasis. Furthermore, this invention introduces a five-dimensional humanistic semantic feedback mechanism of Tao, De, Ren, Yi, and Li to filter and optimize candidate intervention schemes based on value criteria, ensuring that the selected schemes are not only technically effective but also ethically sound and meet individual needs, thus constructing a safe and reliable closed-loop negative entropy control system.

[0007] The present invention provides an intelligent negative entropy intervention algorithm, which specifically includes the following main modules and steps:

[0008] Multimodal data acquisition and entropy assessment: Through wearable sensors and smart terminals, continuous collection of human physiological parameters (such as heart rate, blood pressure, blood oxygen, body temperature, etc.) and behavioral / emotional data (such as sleep duration, activity level, and vocal emotional characteristics) is performed. Information entropy is calculated from the collected multimodal data to obtain the current life entropy state index, which is then compared with the individual's health baseline and a standard homeostasis model to assess the degree of entropy increase in the body.

[0009] Semantic Deviation Diagnosis: This involves mapping the entropy state index to a humanistic semantic space to diagnose deviations in the life system across the "Five Constant Virtues" value dimensions. Specifically, a pre-constructed knowledge graph (such as the DIKWP semantic graph) is used to analyze the value imbalance dimensions corresponding to abnormal entropy values. For example, if elevated entropy is primarily manifested as physiological rhythm disorder, it is judged as an imbalance at the level of "propriety" (disorder); if it is accompanied by low mood or abnormal social behavior, it corresponds to an imbalance at the level of "benevolence" (lack of humanistic care), etc. This step determines which moral / value dimensions the organism's disorder is mainly concentrated in, providing a basis for developing intervention plans.

[0010] Intervention plan generation: Based on the diagnostic results, feasible intervention measures corresponding to the current deviation are retrieved from the knowledge base and experience rule base to generate a set of candidate intervention plans. The knowledge base contains numerous health homeostasis paths and intervention experiences, such as the transfer paths between "health homeostasis" nodes pre-labeled in the DIKWP semantic graph. The algorithm selects relevant path nodes based on the deviation type and deduces several possible intervention plans. Each plan includes one or more specific measures (e.g., a plan to adjust the work and rest routine when the "propriety" dimension is imbalanced, or a plan for psychological counseling or music therapy when the "benevolence" dimension is imbalanced), and includes an estimated entropy reduction effect.

[0011] Ethical Criterion Filtering: The generated candidate solution set is filtered for ethical and safety considerations according to five principles: Tao, De, Ren, Yi, and Li. Solutions that may violate Li (physiological laws or ethical regulations) are eliminated to ensure the safety and compliance of the intervention; solutions that violate Yi (fairness and impartiality) are eliminated to ensure that the patient's interests are prioritized without bias; solutions lacking Ren (care and respect) are eliminated to avoid causing psychological discomfort to the patient or infringing upon their wishes; and solutions that violate De (moderation and appropriateness) are eliminated to exclude measures that involve excessive intervention or significant side effects. Through multi-layered filtering, technically feasible and ethically acceptable alternatives are retained.

[0012] Optimal intervention decision-making: Among the intervention options filtered through ethical criteria, the expected negative entropy effect of each option and its match with the patient's individual preferences are comprehensively evaluated to select the optimal intervention. Optimization criteria include: appropriate and not excessive intervention intensity, effectively reducing entropy; a balance between intervention speed and robustness, i.e., correcting the disorder as quickly as possible while acting on the body as gently as possible; and consideration of the patient's wishes and comfort. For example, among several effective options, the one that is more easily accepted by the patient and has fewer side effects is selected.

[0013] Intervention Implementation and Feedback Monitoring: The selected intervention plan is implemented through the corresponding execution module. Intervention methods can be at the information field level (e.g., verbal guidance, psychological counseling), the energy field level (e.g., electrical stimulation, drug administration), or the behavioral feedback level (e.g., respiratory rhythm training, exercise guidance), and may be implemented individually or in combination. During intervention implementation, sensors continuously monitor changes in the patient's vital signs and entropy values ​​to assess the intervention effect in real time.

[0014] Closed-loop feedback and dynamic adjustment: The system determines whether the intervention has achieved the expected effect based on monitoring feedback results. If the entropy value has returned to the normal steady-state range, the control loop ends; otherwise, it automatically enters the next loop. The algorithm adjusts the intervention plan based on feedback: on the one hand, it can fine-tune the parameters of the current plan (e.g., increase the dose or frequency of a certain intervention stimulus); on the other hand, if the current plan is ineffective, it considers using alternative plans (e.g., switching or adding other intervention measures). This process iterates repeatedly until the life entropy value returns to the healthy steady-state threshold, achieving dynamic correction and stable maintenance of the body's state.

[0015] In summary, this invention constructs a complete closed-loop negative entropy feedback control system: using the deviation of life information entropy as the control input, a decision controller integrating humanistic semantic weights outputs multimodal intervention commands to the patient, and feedback is generated through continuous sensing and monitoring. This system can correct deviations like the human physiological homeostasis system, except that the "controller" here incorporates higher-level intelligence and ethical considerations, thereby achieving safe and effective health state regulation.

[0016] Compared with existing technologies, the intelligent negative entropy intervention algorithm provided by this invention has significant beneficial effects:

[0017] Proactive prevention and timely intervention: The algorithm continuously monitors the patient's condition and can intervene immediately once it detects signs of sub-health such as an abnormal increase in entropy value, nipping the disease risk in the bud and significantly reducing the rate of disease deterioration.

[0018] Personalized adaptive regulation: The intervention plan is dynamically adjusted based on the real-time feedback of each individual, which avoids adverse reactions caused by over-intervention and prevents poor efficacy caused by under-intervention, thereby maintaining individual homeostasis more accurately.

[0019] Integrating humanistic ethics ensures reliable decision-making: The five-dimensional principles of morality (harmony), virtue (moderation), benevolence (care), righteousness (justice), and propriety (norm) are introduced to constrain AI decision-making, ensuring that the resulting intervention plan conforms to medical ethics and the patient's wishes, and reducing the possibility of over-medicalization or inappropriate measures caused by mechanical decision-making.

[0020] Enhancing Chronic Disease Management: Through continuous negative entropy input, this invention helps improve long-term indicators and quality of life for patients with chronic diseases. For example, rehabilitation training plans can be flexibly adjusted based on entropy trends, promoting the restoration of homeostasis and reducing recurrent attacks.

[0021] Assisting physicians in decision-making and reducing their workload: This algorithm can serve as an intelligent assistant for medical personnel, providing real-time recommendations for treatment plans. With this invention, physicians can identify potential patient problems earlier and take action, improving the scientific rigor of clinical decision-making while reducing their workload in monitoring and treatment planning.

[0022] Innovation and Applicability: This invention is the first to use the information entropy index for the quantification of health status and combines it with humanistic semantics for multimodal intervention and control, thus compensating for the shortcomings of traditional medical feedback mechanisms. The algorithm can be deployed on wearable devices, mobile medical platforms, or hospital monitoring systems, and is compatible with existing medical infrastructure, showing broad application prospects. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the system structure of the intelligent negative entropy intervention algorithm based on the semantic feedback of Tao-De-Ren-Yi-Li.

[0024] Figure 2 This is a flowchart of an intelligent negative entropy intervention algorithm based on semantic feedback of Tao, De, Ren, Yi, and Li. Detailed Implementation

[0025] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, those skilled in the art should understand that various modifications or equivalent substitutions can be made to the implementation methods without departing from the spirit of the present invention, and the scope of protection of the present invention is not limited to the following specific examples.

[0026] like Figure 1 As shown, the intelligent negative entropy intervention system of the present invention mainly includes: a multimodal sensor (1), a negative entropy intervention control module (2), an intervention execution device (3), and a controlled object, namely, a patient (4). The sensor (1) is used to collect various vital signs and behavioral data of the patient and send the data to the control module (2). The control module (2) performs information entropy calculation and semantic analysis on the received data and generates intervention decision instructions. The intervention execution device (3) implements corresponding intervention measures (such as physical stimulation or information guidance) on the patient (4) according to the decision instructions. After receiving the intervention, the patient (4)'s physiological or psychological state changes, which is reflected in the new sensor data and fed back to the control module, thus forming a closed-loop control. The entire system architecture ensures the cyclical process of data acquisition—decision processing—intervention execution—feedback monitoring, and can continuously regulate the body's state through negative feedback.

[0027] like Figure 2 As shown, the process of the intelligent negative entropy intervention algorithm based on the present invention includes the following steps:

[0028] Data Acquisition and Entropy Assessment: Real-time data is acquired through various sensors worn on the patient, including physiological signals (such as ECG, blood pressure, blood sugar, body temperature, etc.) and behavioral and psychological signals (such as activity level, sleep records, speech tone and content, etc.). The control module (2) preprocesses and extracts features from this data and calculates the current life information entropy value. For example, the entropy value can be calculated using the complexity of heart rate variability, or the "behavioral entropy" can be estimated by analyzing the fluctuations in daily behavior patterns. Then, the entropy value is compared with the pre-established patient personal baseline and the standard steady-state entropy value range of the general healthy population to determine the current life entropy state index. If the entropy value is significantly higher than the healthy steady-state level, it means that the body is in a state of disorder or stress; if the entropy value continues to rise, it indicates that the patient's condition is developing in an unfavorable direction and needs to be taken seriously.

[0029] Semantic Deviation Diagnosis: Based on changes in the state of life entropy, the control module, combined with a medical knowledge graph and semantic analysis model, diagnoses the causes and nature of entropy increase. This step utilizes the "Five Constant Virtues" (Tao, De, Ren, Yi, Li) as an analytical framework: the algorithm determines which value levels the current entropy abnormality primarily reflects as imbalance. For example, if sensor data indicates that the patient's work and rest are extremely irregular and their physiological rhythms are disordered, it can be determined that the "Li" dimension (order and norms) is deviating from normal—possibly due to irregular lifestyle habits leading to disorder in the body's internal environment. Similarly, if language sentiment analysis results show that the patient has recently experienced low mood and reduced social behavior, then the corresponding "Ren" dimension (emotional care) is lacking, indicating a mental health problem. Likewise, short-term overexertion of the body (such as overwork) leading to a deterioration in physiological indicators can be considered an imbalance in the "De" dimension (moderation and self-discipline); unfair access to medical resources can be considered a problem in the "Yi" dimension; and severe overall physical and mental imbalance violates the harmony of "Tao." Through such semantic diagnosis, the algorithm can locate the deep-seated factors behind entropy increase, including both physiological causes and psychosocial factors. This lays the foundation for developing targeted intervention measures in the future.

[0030] Intervention Plan Generation: After clarifying the nature and main dimensions of the patient's state deviation, the system enters the intervention plan deduction stage. The control module accesses the built-in knowledge base and rule base (such as a health management knowledge graph, clinical experience database, etc.) to retrieve intervention strategies corresponding to the current problem. This knowledge base can be in the form of a DIKWP semantic graph, where nodes represent various health states and intervention measures, and edges represent pathways from one state to another. For example, for a problem of imbalance in the "etiquette" dimension (disordered daily routine), the knowledge base may contain nodes such as "regular daily routine" and "acupuncture and meridian conditioning," pointing to the steady-state path of restoring order; for an imbalance in the "benevolence" dimension (psychological depression), the graph may contain intervention nodes such as "psychological counseling" and "music therapy." The algorithm combines the patient's current specific situation (such as the severity of entropy deviation, past medical history, etc.) to select several relevant paths from the graph to form a set of candidate intervention plans. Each plan consists of one or more specific measures, and its expected effect on reducing entropy is estimated through a model. For example, for mild sleep disturbances, a solution might be "reducing screen time one hour before bedtime and incorporating relaxation meditation," which is expected to lower entropy to a certain extent. For severe psychological stress, a solution might include "weekly psychological counseling + music relaxation therapy," which is expected to lower entropy even more. The set of candidate solutions generated in this step provides a basis for subsequent screening.

[0031] Ethical Criterion Filtering: After generating candidate solutions, the system does not directly select the solution with the highest entropy reduction effect. Instead, it first reviews and filters the solutions based on five principles: Tao, De, Ren, Yi, and Li. This step ensures that the intervention measures recommended by the algorithm are acceptable from a safety, ethical, and humanistic perspective. Specifically:

[0032] Etiquette guidelines: Eliminate any plans that violate medical norms, ethical standards, or laws and regulations. For example, do not use folk remedies whose safety has not been verified, and do not conduct interventions that violate clinical guidelines, ensuring that the plan is compliant and reliable.

[0033] Fairness in justice: Eliminate solutions that may be unfair to the patient group, ensure that the interests of patients are put first, and avoid unfair resource allocation or discriminatory decisions caused by algorithmic bias.

[0034] Compassionate care: Eliminate methods that are too harsh or rigid, potentially causing discomfort to patients. Prioritize gentle, humane interventions that respect the patient's dignity and feelings. For example, for psychological problems, tend to use guidance rather than blame or admonishment.

[0035] The virtue of moderation: Avoiding over-medicalization and over-intervention. Eliminating solutions that clearly exceed actual needs or may cause significant side effects, and advocating for minimizing unnecessary interventions to the body when the problem can be effectively resolved. For example, avoiding invasive surgery or high-intensity medications unnecessarily, and prioritizing conservative treatments.

[0036] The harmony of the Tao: The solution that ultimately remains should be one that restores the system's balance with minimal necessary intervention. It pursues holistic harmony of body and mind, rather than treating symptoms piecemeal. In other words, the chosen solution should comprehensively consider the patient's overall condition, achieving the restoration of overall homeostasis with minimal intervention.

[0037] After the above multi-layered filtering process, only solutions that simultaneously meet both technical effectiveness and ethical appropriateness will proceed to the next step. This step constrains AI decision-making with human values, preventing situations where the pursuit of optimal entropy metrics violates medical ethics, and ensuring the reliability and acceptability of the decision results.

[0038] Solution Optimization: After obtaining a set of ethically compliant candidate solutions, the system comprehensively evaluates them to select the best solution. The comprehensive evaluation considers factors such as the expected entropy reduction, the time required for intervention, patient subjective acceptability, implementation costs, and alignment with the patient's prior preferences. Typically, solutions with significant and rapid entropy reduction effects, while minimizing patient interference, are selected. For example, if two solutions can reduce entropy by a similar magnitude, but one is faster or more readily accepted by the patient, that solution is preferred. If necessary, two or more complementary solutions can be combined for implementation to quickly and smoothly bring the entropy back to a steady state. The optimization decision process may also refer to historical data from similar cases to identify interventions that have yielded better results in similar situations. The final output is the intervention plan to be implemented in the current cycle.

[0039] Program Implementation and Monitoring: The system will issue the selected intervention plan to the corresponding execution device (3) or prompt medical staff / patients to cooperate in its implementation. Depending on the content of the plan, the implementation methods include:

[0040] Information field intervention: Providing patients with language guidance, psychological counseling, music / video therapy, etc. through multimedia terminals. These interventions mainly act on the patient's psychological and cognitive levels, typically improving deviations in the areas of "benevolence" or "righteousness".

[0041] Energy field intervention: This involves applying physical or chemical stimulation to the patient using medical devices, such as low-frequency electrical stimulators to relax muscles, intelligent drug pumps for precise drug delivery, and traditional Chinese medicine meridian therapy devices. This type of intervention primarily corrects physiological imbalances, corresponding to adjustments at the level of "propriety" or "virtue."

[0042] Behavioral feedback intervention: This involves guiding or assisting patients in specific behavioral training, such as breathing rhythm training, posture adjustment, and motor rehabilitation exercises. Sometimes, it is used in conjunction with wearable assistive devices (such as exoskeletons) to ensure that patients achieve the required movement standards. This type of intervention helps to rebuild the overall harmony and dynamic balance of the "Tao" (the Way).

[0043] During the intervention, the sensor (1) continuously collects real-time response data from the patient. The control module (2) monitors the process while it is being implemented, and assesses the effectiveness and degree of effectiveness of the current intervention by comparing the new entropy value with the previous value. If the intervention involves a long period of time (such as a rehabilitation training cycle), the system will also periodically record the trend of entropy value changes for periodic assessment.

[0044] Feedback Adjustment: By monitoring data, the system determines whether the current intervention has brought the patient's condition back to a steady state. If the entropy value has decreased and remained near the target threshold, it indicates that the intervention has achieved the expected effect, and the current control cycle can end, requiring only routine monitoring. However, if the entropy value still does not reach the ideal range after a period of intervention, or although it has decreased, it is still higher than the baseline, indicating that the strategy needs to be adjusted. At this time, the system will automatically perform feedback adjustment, i.e., enter the next cycle:

[0045] First, based on the feedback from the previous round of intervention, the model is updated and calibrated (for example, correcting the entropy reduction effect estimate of a certain solution in similar situations in the knowledge base).

[0046] Secondly, the intervention program should be adjusted accordingly. If monitoring shows that the patient responds well to one measure but poorly to another, the former should be increased in proportion or intensity, while the latter should be reduced or replaced. If a previously rejected measure is found to be more suitable for a new situation (e.g., changes in patient tolerance), it can be reconsidered.

[0047] Implement the new optimization plan again and continue to monitor and evaluate it.

[0048] This process repeats until the patient's entropy value stabilizes at a healthy homeostatic level. The entire process is analogous to negative feedback regulation in classical cybernetics, except the regulator is an intelligent algorithm, the target is the entropy of life information, and the means are multimodal comprehensive interventions. In each iteration, the system continuously approaches the optimal intervention strategy, ultimately achieving precise control over the complex human body system.

[0049] A 45-year-old male patient experienced high work stress and irregular lifestyle. Wearable devices detected a significant decrease in his daytime heart rate variability, intermittent sleep at night, and a slightly elevated blood pressure trend. The algorithm calculated that his life entropy state index was 15% higher than his usual baseline. Semantic diagnosis results showed that this entropy increase was mainly reflected in the deviation of the "propriety" dimension—that is, the disruption of physiological rhythm and life order, accompanied by a mild deviation of the "benevolence" dimension (emotional irritability). The system retrieved relevant intervention measures from the knowledge base, and the generated candidate solutions included: (a) lifestyle intervention: going to bed 1 hour earlier, exercising early in the morning, and using a blue light filter to improve sleep quality; (b) traditional Chinese medicine rhythm regulation: using intelligent acupuncture devices to stimulate specific meridians to regulate autonomic nerve function; (c) psychological stress reduction counseling: 10 minutes of meditation training before bed every night and group psychological counseling once a week.

[0050] After being filtered by ethical guidelines, all the above solutions were deemed safe, compliant, and patient-friendly. Considering both entropy reduction and patient acceptability, the algorithm optimally selected a combination of solutions (a) and (b): a comprehensive intervention combining a regular sleep schedule with meridian regulation. The system pushed specific sleep plans to the patient via a mobile app (e.g., fixed bedtime and wake-up times, dietary and exercise recommendations), and a smart acupuncture device performed gentle electrical stimulation on acupoints such as Zusanli for 10 minutes each night before bedtime according to a pre-set program. After the intervention, sensor monitoring showed that the patient's nocturnal heart rate variability gradually increased, the number of sleep interruptions decreased, and morning blood pressure returned to the normal range. Three days later, the retested life entropy index decreased by 10%, returning to the patient's normal fluctuation range. Simultaneously, the patient subjectively reported improvements in energy and mood. Due to the improved health status, the system gradually reduced the frequency of acupuncture stimulation as planned, while continuing to monitor sleep schedule and physiological indicators. Throughout the process, the algorithm successfully pulled the patient from a sub-healthy state of increased entropy back to a stable state, achieving a preventative effect of "treating disease before it occurs."

[0051] A 60-year-old female patient entered the recovery period after surgery. She exhibited mild depressive symptoms and recently showed decreased adherence to her active rehabilitation training. Sensor data indicated low daytime activity, high nighttime heart rate, and shallow sleep, with her vital entropy index increasing by 20% compared to discharge. Semantic deviation diagnosis pointed to two aspects: first, an imbalance in the "benevolence" dimension, indicating low mood and lack of motivation; second, a deviation in the "propriety" dimension, indicating poor adherence to daily routines and rehabilitation training plans. In response, the algorithm retrieved the following intervention plans from the knowledge base: (a) Psychological care and motivation stimulation: A virtual nursing assistant communicates with the patient daily, providing positive encouragement and playing the patient's favorite light music to aid relaxation; (b) Personalized rehabilitation training plan: Adjusting exercise intensity, starting with light exercise that the patient can tolerate and gradually increasing, and using a gamified training app to enhance engagement; (c) Sleep environment optimization: Using smart aromatherapy and light music in the evening to create a relaxing atmosphere and improve nighttime sleep.

[0052] The above-mentioned solutions, after ethical review, are all mild and feasible, reflecting care and restraint for the patient. After comprehensive evaluation, the system decided to adopt a combined approach of (a) + (b). In practice, the smart nursing app sends encouraging messages every morning and guides the patient to complete simple exercises (such as stretching), gradually increasing the intensity. A virtual nursing assistant communicates with the patient via voice, encouraging them to persevere in rehabilitation and providing psychological support or contacting a real doctor when the patient is feeling down. Simultaneously, wearable devices monitor the patient's exercise data and physiological indicators, feeding this information back to the algorithm. In the initial days, the patient's mood and activity remained low, but with continued care and gradual training, after a week, their average daily steps increased by 50%, nighttime heart rate decreased, sleep quality improved, and entropy state index decreased by 15%. At this point, the system upgraded the rehabilitation plan, introducing slightly more intense training programs while continuing to monitor the patient's emotional responses. The entire rehabilitation process, dynamically adjusted by the algorithm, achieved a balance between exercise intensity and the patient's physical and mental state. This avoided the physical burden that overtraining might cause (reflecting the principle of "moral restraint") while simultaneously enhancing the patient's willingness to participate in rehabilitation through continuous humanistic care (reflecting the principle of "benevolence"). Ultimately, the patient's various physiological indicators tended to normalize, their psychological state significantly improved, and their life information entropy returned to a healthy homeostatic level. This embodiment demonstrates the application value of the algorithm of this invention in rehabilitation management: through a negative entropy intervention loop, individualized adjustments to rehabilitation strategies promote comprehensive patient recovery.

Claims

1. An intelligent negative entropy intervention algorithm based on Dao-De-Jing-Yi-Li semantic feedback, characterized in that, The algorithm comprises the following steps: (1) Data collection and entropy state evaluation: Collecting multi-modal life data of the patient, including physiological parameters and behavior / emotion information, calculating the current life information entropy value, and comparing it with the preset health baseline to evaluate the degree of entropy increase; (2) Semantic deviation diagnosis: Mapping the entropy value state to the semantic space of the five value dimensions of Dao, De, Ren, Yi and Li, and diagnosing the deviation of the life state from the corresponding main value level; (3) Intervention scheme generation: Based on the diagnosis result, retrieve the candidate intervention measures corresponding to the deviation from the knowledge base, generate multiple candidate intervention schemes and estimate their expected entropy value reduction effect; (4) Ethical criteria filtering: According to the preset ethical criteria of Dao, De, Ren, Yi and Li, filter the candidate intervention schemes, eliminate the schemes that violate the physiological rules, medical ethics, fairness principles or do not meet the requirements of humanistic care and restraint, and obtain the scheme set that meets the ethics and is technically feasible; (5) Scheme optimization: Comprehensive evaluation of the filtered scheme set, and selection of the optimal intervention scheme according to the entropy value reduction effect, intervention intensity appropriateness and patient preference; (6) Scheme implementation and monitoring: Implementing the selected intervention scheme, intervening the patient through information field, energy field or behavior intervention device, and monitoring the life information entropy change in the intervention process by using sensors; (7) Feedback adjustment: According to the monitoring feedback, judge whether the entropy value returns to the steady state, if not, adjust the intervention parameters or replace the candidate scheme and return to step (3) to execute again, forming a closed loop negative feedback control until the entropy value returns to the steady state range.

2. The intelligent entropy intervention algorithm of claim 1, wherein, The multi-modal life data includes physiological signal data and psychological / behavioral data, the physiological signal data at least contains one or more of heart rate, blood pressure, respiratory rate and body temperature, and the psychological / behavioral data includes sleep duration, activity amount, voice tone or emotion score information.

3. The intelligent entropy-negative intervention algorithm of claim 1, wherein, The semantic deviation diagnosis utilizes the pre-constructed health semantic graph to analyze and locate the entropy value anomaly, and corresponds the entropy increase state to at least one deviation in the "Five Constants" value dimension, wherein: the Li dimension deviation represents the disorder of physiological rhythm or behavior standard, the Yi dimension deviation represents the problem of decision fairness or interest balance, the Ren dimension deviation represents the lack of psychological emotion or social care, the De dimension deviation represents the lack of self-discipline or excessive consumption, and the Dao dimension deviation represents the overall physical and mental system disorder.

4. The intelligent entropy-negative intervention algorithm of claim 1, wherein, The intervention scheme generation step utilizes the DIKWP semantic knowledge graph to deduce the health steady state path, the knowledge graph contains various health state nodes and corresponding intervention measure nodes, and the algorithm selects the corresponding steady state path node according to the diagnosed deviation type, generates candidate intervention schemes and calculates the expected entropy value reduction value of each scheme.

5. The intelligent entropy-negative intervention algorithm of claim 1, wherein, The intervention scheme includes at least three types of information field intervention, energy field intervention and behavior feedback intervention: information field intervention acts on the cognitive and emotional level of the patient through language guidance, psychological counseling or multimedia therapy, energy field intervention acts on the physiological level of the patient through physical stimulation or drug control, and behavior feedback intervention acts on the behavior habit level of the patient through training device or guidance scheme.

6. The intelligent entropy-negative intervention algorithm of claim 1, wherein, The moral code filter screens the intervention scheme according to the following rules: priority is given to the scheme that meets the physiological norms of propriety and medical ethics, does not harm the rights of patients (benevolence), respects the will of patients and is gentle in approach (benevolence), the intervention intensity is moderate to avoid excessive medical treatment (virtue), and the scheme with minimal intervention and the ability to restore overall harmony and stability (Tao) is the final candidate.

7. The intelligent entropy-negative intervention algorithm of claim 1, wherein, The feedback adjustment adopts a closed-loop negative feedback control strategy, automatically adjusts the parameters or replaces the scheme for the intervention that does not achieve the expected effect, and performs monitoring again, and iterates until the life information entropy is restored to the preset stable state threshold range.

8. A smart entropic intervention system for implementing the algorithm of any one of claims 1 to 7, the system comprising: At least one set of multi-modal sensors, control processing modules and intervention execution devices, characterized in that: The multi-modal sensors are used to collect physiological parameters and behavior data of the patient, and send the data to the control processing module; The control processing module includes an entropy value calculation unit, a semantic analysis unit, an intervention decision unit and a feedback control unit, which is used to calculate the life entropy state and the value dimension deviation according to the received data, generate candidate intervention schemes and apply the preset Tao, virtue, benevolence, righteousness and propriety criteria to screen and optimize the scheme, and determine the optimal intervention scheme to generate control instructions; The intervention execution device implements corresponding intervention measures on the patient according to the control instructions, and the multi-modal sensors collect feedback data after the intervention and transmit it back to the control processing module in real time, so that the control processing module automatically adjusts the intervention scheme parameters according to the feedback, forming a closed-loop negative entropy intervention control loop.

9. A computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the intelligent negative entropy intervention algorithm of claim 1.

10. The intelligent entropic negativity intervention system of claim 8, wherein: The multi-modal sensors include wearable physiological parameter sensors and voice emotion analysis modules for collecting physical sign signals and language / emotion information of the patient, respectively; the intervention execution device includes an electrical stimulation therapy module, a drug administration module, and a human-computer interaction terminal for providing language guidance or music therapy, to respectively realize physical intervention, drug intervention and information intervention on the physiological level and the psychological level of the patient.

Images